Cedar Springs Hospital, Inc. UHS of Delaware, Inc., dba Cedar Springs Hospital 20-0887

Alicia Truman, Esq., and Beau Ellis, Esq., Department of Labor, Office of Solicitor, Denver, Colorado

After learning of significant allegations through the Occupational Safety and Health Administration’s (“OSHA’s”) Whistleblower Protection Program, Compliance Officer (“CO”) Aimee Stark opened an investigation of a behavioral health facility located at 2315 Southgate Road, Colorado Springs, CO (the “Worksite”). (Ex. R-UHS-8.) The multi-month investigation identified numerous injuries from the hazard of workplace violence. The Secretary cited Cedar Springs Hospital, Inc. (“Cedar Springs”) and an affiliate, UHS of Delaware, Inc. (“UHS-DE” and collectively with Cedar Springs, “Respondents”) for: (1) failing to provide employment and a place of employment free from the recognized hazard of workplace violence under the general duty clause of the Occupational Safety and Health Act of 1970, 29 U.S.C. §§ 651-678 (the “OSH Act”), and (2) failing to provide an authorized government representative records within four business hours as required by 29 C.F.R. § 1904.40(a).

Respondents timely contested the Citation, bringing the matter before the Occupational Safety and Health Review Commission (“Commission”). A hearing commenced on February 7, 2022, and concluded on February 17, 2022. All parties filed briefs after the hearing.

The Citation’s first assertion is that employees were exposed to physical attacks from aggressive patients in violation of section 5(a)(1) of the OSH Act, the provision commonly known as the general duty clause. 29 U.S.C. § 654(a)(1). This provision requires employers to “furnish to each of his employees employment and a place of employment which are free from recognized hazards that are causing or are likely to cause death or serious physical harm to his employees.” 29 U.S.C. § 654(a)(1). The Citation alleges that employees suffered serious injuries from workplace violence, including concussions, sprains, bruising, and other injuries to the head and torso. The Citation’s second contention is that Respondents failed to provide the Worksite’s injury and illness records to an authorized representative within a reasonable time from the request.

The key issues in dispute are: (1) the appropriateness of sanctions for Respondents’ failure to preserve electronically stored information (“ESI”), (2) whether the Secretary established the alleged violations, and (3) whether Cedar Springs and UHS-DE should be considered a single-employer for the cited violations.

For the reasons below, the Secretary’s renewed Motion for Sanctions for Respondents’ Destruction of Relevant Recorded Video Evidence is GRANTED in part, and DENIED in part. Citation 1, Item 1 is AFFIRMED as to Cedar Springs and UHS-DE as a serious violation of the OSH Act, and a $13,494 penalty is assessed. Citation 2, Item 1, is AFFIRMED as to Cedar Springs and UHS-DE as an other-than-serious violation of 29 C.F.R. § 1904.40(a), and a $1,928 penalty is assessed.

I.Jurisdiction

Cedar Springs and UHS-DE are both employers under the OSH Act.¹ Each business affects interstate commerce within the meaning of 29 U.S.C. § 659(c).² “The Citation and Notification of Penalty underlying this proceeding was issued on May 15, 2020.” (Stip. 5.) In response, “Respondents timely filed their Notices of Contest on June 4 and 5, 2020.” (Stip. 6.) The parties agree that the Commission “has jurisdiction in this proceeding pursuant to § 10(c) of the [OSH Act].”³ Based upon the record, including the parties’ stipulations and acknowledgments, the undersigned concludes that the Commission has jurisdiction over the parties and the subject matter of these proceedings.

II.Factual Background

“Cedar Springs is a wholly owned subsidiary of Psychiatric Solutions Hospital, Inc.” (Stip. 18.) Psychiatric Solutions Hospital, Inc., in turn, is itself a wholly owned subsidiary of Universal Health Services, Inc. (“UHS”), a large publicly traded company with no direct employees.⁴ UHS has other wholly owned subsidiaries, including UHS-DE. (Stips. 1-2, 17-18; Meloni Dep. 17, 25; UHS-DE Br. 12.) Although Cedar Springs and UHS-DE share the same ultimate corporate parent, each is a separate corporate entity.⁵

Employees of both entities worked at the Worksite.⁶ Cedar Springs’ Chief Executive Officer (“CEO”) and its Chief Financial Officer (“CFO”) were employed by UHS-DE. (Stips. 21-22.) The CEO directly supervised Cedar Springs’ employees.⁷ However, “physicians, nurses, Mental Health Specialists (MHSs), and other direct care providers on the staff of Cedar Springs Hospital are not employees of UHS-DE.” (Stip. 51.)

The Worksite is a behavioral health hospital that treats up to 110 adult, adolescent, and child patients in acute and residential settings.⁸ Most patients are admitted involuntarily while suffering from serious conditions. (Tr. 2311; CS Br. 10.) Patients arrive through referrals from other hospitals, by coming in directly and requesting admission, or are brought in by police. (Tr. 89-90.)

The Worksite has several separate buildings. (Exs. C-3 at 19-27, C-5, C-9.) New patients often arrive at the Evergreen building where the Admissions and Referrals Department (“A&R”) begins the admission process. (Tr. 89-90.) A&R is open 24 hours daily, and admissions can occur anytime. (Tr. 92.) There are no patient care units in the Evergreen building. However, group therapy and outpatient therapy can occur there. (Tr. 89; Ex. C-9; CS Br. 12.)

Once admitted, a worker brings the patient to a care unit in one of three buildings: Aspen, Juniper, or Peak House. In Aspen, there are four patient care units, each with up to 12 patients: (1) Hickory Hall for high-acuity adults, (2) Cottonwood Cove for high-acuity adolescents, (3) Treehouse for high-acuity adolescents and children, and (4) Maple for high acuity adolescents and children. (Tr. 99-100, 2311; Ex. C-2b, C-3 at 30; CS Br. 12-13.) The Juniper building has three patient care units: (1) Birchwood for 12 lower acuity adults, (2) Ponderosa for 12 lower acuity adults, and (3) Mountain View Place (“MVP”), a residential unit for up to 24 adolescents. (Tr. 92-96, Ex. C-5; CS Br. 12.) Peak House is for patients voluntarily seeking chemical dependency treatment. (Tr. 101, 2312; CS Br. 13.) The Worksite includes another building, referred to as the Commons, which has a school that minors attend while receiving care. (Tr. 101; Ex. C-5 at 7; CS Br. 13.)

Each care unit has a nurses’ station. During the inspection period, MVP’s nurses’ station was open, with no enclosure or doors. (Tr. 94, 992-93; Ex. C-3 at 24.) The nurses’ stations on the Ponderosa and Birchwood units had partial barriers extending approximately six feet high. (Tr. 95-97; Ex. C-3 at 27.) However, there was a half door that patients could reach over and unlock. (Tr. 96-98; Ex. C-3 at 27). The nurses’ stations on Hickory, Cottonwood, and Maple had barriers that extended approximately six feet above the counter, with a sliding Plexiglas window. (Tr. 100; Ex. C-3 at 30.) Those windows were commonly left open. (Tr. 1488-89; Ex. C-107.) Even when closed, patients sometimes could remove the screw that was supposed to keep the window closed, or the windows would break. (Tr. 1490-91.) On one side, below the plexiglass, there is a ledge, and further below are desks. (Tr. 1484; Exs. C-3, C-107.)

Patients sometimes reached into nurses’ stations to grab pens, pencils, and paperclips. (Tr. 94-95, 276-77, 487, 628, 859, 960, 1487-89, 1851.) Patients were also able to enter the nurses’ stations. (Tr. 859-60, 864-65, 992-93, 1489, 2323; Exs. C-54, C-58.) On December 7, 2019, two patients climbed over the nurses’ station in the Treehouse unit. (Ex. C-57 at 4; Tr. 507-8.) A few days later, a patient repeatedly climbed up and jumped over the barrier into Cottonwood’s nurses’ station. (Ex. C-58; Tr. 864-65.) Patients climbed and jumped over the counters and partial enclosures, including those with the six-foot partial barrier. (Tr. 404, 486-87, 859-62, 864-65, 959, 1487-89, 2323; Exs. C-53, C-54.). It was not unusual for patients to find ways into the nurses’ station.⁹ Patients jumped on the desk, bit, hit, and kicked staff at the nurses’ stations. (Exs. C-20 at 1-5, 11-12, 23, 26-33, 35-36, 45-46, 48, 51; C-64; Tr. 1850-51.)

The following month, there continued to be multiple incidents at nurses’ stations. On January 6, 2020, an MHS’s finger was lacerated, and his knee was hurt when he tried to address a patient entering the nurses’ station. (Tr. 623-26; Exs. C-13, C-64 at 27.) In another incident, a patient jumped on an MHS after climbing over the nurses’ station’s barrier. (Tr. 957-58; Ex. C-14 at 1, 3.) Other staff were kicked or assaulted when they tried to escort patients out of, or away from, the nurses’ station. (Exs. C-19, C-20, C-64 at 1, 61, 73.)

Approximately 300 people work at the Worksite in various capacities. (Tr. 87.) Physicians, nurses, social workers, teachers, therapists, and MHSs provide direct care for the patients. (Tr. 87; Ex. C-3 at 11.) MHSs work very closely with the patients. They manage the care environment, observe patients, accompany them to meals and recreation, and provide activities. (Tr. 2431-32; Exs. RCS-7, RCS-29, RCS-32; UHS-DE Br. 26.) MHSs, like nurses, work eight-hour shifts: 7:00 a.m. to 3:00 p.m., 3:00 p.m. to 11:00 p.m., or 11:00 p.m. to 7:00 a.m. (Ex. C-68.)

Respondents set the number of staff typically assigned to each unit based on the number of patients. (Tr. 834-35, 1656, 2375-76; Ex. C-68.) Their staffing grid identifies the number of MHSs per shift. Id. The number varies with the patient census and the time of day. Id. The grid calls for fewer MHSs on the overnight shift. (Tr. 838; Ex. C-68.) The grid does not set out adjustments for patient acuity, the number of patients on assault precautions, or planned admissions. (Tr. 394, 839, 1210, 1656; Exs. C-3 at 43, C-68.) It is a strictly mathematical arrangement based on the patient census at a given time. (Tr. 1630, 1656.) In addition to the MHSs, the grid calls for one nurse per unit for all shifts, regardless of the census. (Tr. 836-39, 2375-76; Exs. C-3 at 43, C-68.)

Besides nurses and MHSs, other workers could sometimes assist during workplace violence incidents. On weekdays, some therapists and non-direct care staff may be able to stop their typical duties and assist in an emergency. There were fewer workers outside of the day shift and on weekends. (Tr. 395-96, 511, 758, 840, 1408-9, 2386, 2375, 2558, 2596; Exs. C-3 at 11, C-43.) During the evening and overnight shifts and throughout the weekends, staff was limited to those working in the units and only two or three others. (Tr. 840, 2375, 2596.) Typically, there were no therapists or administrative staff overnight.¹⁰ When staff took breaks, there was only one nursing supervisor to fill in for them. (Tr. 511, 2386-87, 2596-97; Exs. C-43, C-56, C-57.) On weekends, there was no one in hospital management at the Worksite. (Tr. 395-96, 758, 2558.) Housekeepers only worked in the mornings, and maintenance workers were off-site. (Tr. 395-96, 2558.)

How many people were typically available to respond to emergency requests is unclear. Workers might be engaged in other tasks or unable to hear the codes from their location. (Tr. 1419; Ex. C-3 at 43.) At times, patient care units operated with less staff than the levels set out in the staffing grid. (Tr. 394-95, 605-6, 688, 835, 839-40, 1009, 1227, 2375-77, 2453, 2590-91, 2595-96, 2598-99; Exs. C-9, C-33, C-34, C-51, C-52, C-55, C-57, C-68, C-73.) The grid was a guideline instead of the minimum or maximum staff level. (Ex. C-68.) The Worksite’s staffing policy required one nurse per unit, except for MVP. (Tr. 2556; Ex. RCS-47.) For MVP, a nurse did not have to present full-time. (Ex. RCS-47.) The policy did not require a specific number of MHSs or other workers. (Tr. 2556-57; Ex. RCS-47.)

Dawn Wilkosz was a charge nurse during the investigation.¹¹ She stated that her units were “very rarely” staffed to the amount called for by the grid. (Tr. 1512.) Latricia Degenhardt worked as a day-shift nursing supervisor.¹² It was not unusual for someone to serve simultaneously as the nursing supervisor and the only nurse for a particular unit. (Tr. 504-5, 2376; Exs. C-52, C-57.) She discussed a shift when only one-third of the staff the grid called for was working. (Tr. 392-95; Exs. C-2B at 8, C-33.) The Director of Nursing was dismissive when repeatedly informed of short-staffed shifts. (Exs. 2b, C-30, C-33.) Kathleen Nolan, a current employee and supervisor, agreed that there would be times when the units had fewer staff than the grid’s lowest ratios. (Tr. 2376-77.)

The Worksite’s staffing records were not always accurate and could be misleading. For example, new employees were once counted as being in the units caring for patients while in training outside the units. (Tr. 458-59; Exs. C-9 at 2, C-30, C-57 at 3.) The records do not identify how many people were available to respond to violence. (Tr. 458-59, 1212-18; Exs. C-3 at 42, C-9, C-30, C-57 at 3, C-70 at 1-7.) The Secretary’s expert, Howard Forman, M.D., was accepted as an expert in the clinical management of psychiatric patients and had extensive experience working in psychiatric settings. (Tr. 1183.; Ex. C-91.) He explained that the staffing grid is not particularly relevant as a standalone document. (Tr. 1216-17; Ex. C-92 at 13-16.) Staffing at the Worksite did not consistently match the grid, and the number of workers the grid called for was not adequately adjusted for patient acuity. Id.

Even when staffed consistent with the grid, sometimes there was not enough staff to maintain safety. (Tr. 397, 400, 472-73, 488, 790, 988, 1036-37, 1210-18, 1941, 2610; Exs. C-2b, RCS-41 at 17, Exs. C-37, C-69, C-70, C-92 at 11.) Melissa Drawdy worked as the charge nurse during the time of OSHA’s investigation. (Tr. 751.) Based on her Worksite experience, one nurse and one MHS for a unit with twelve to fourteen patients was inadequate to maintain staff safety. (Tr. 790.)

The Director of Nursing could increase the grid’s staffing level for acuity. (Tr. 490; CS Br. 26-32.) Staff could be added if patients had more significant needs or a higher than baseline level of aggressive behavior. (Tr. 159-61, 838-39; Ex. C-68.) However, patients needing more frequent observation or being placed on assault precautions did not trigger staffing increases automatically. (Tr. 490, 842, 988.) Managers could not increase staffing levels directly, and workers had different understandings of acuity. (Tr. 842, 988, 1629-30, 2313-14, 2372-73; Exs. R-CS 47, R-CS 91.) There was no clear policy regarding when acuity staffing would be provided. (Tr. 490, 839, 842, 988, 1629-30, 2557, 2372-73; Exs. RCS-47, RCS-91.) The acuity request form used to get the required approval from the Director of Nursing’s approval did not specify when acuity staffing would be provided.¹³

In practice, increased staffing for acuity rarely happened. (Tr. 472-73, 489, 688, 790-91, 988, 1217-18, 1451; Ex. C-92 at 11, 13-17.) Even approved requests would not necessarily increase the staff level. (Tr. 393-94, 490, 605-6, 688, 867-67, 988, 1514-15, 2560-61; Exs. C-58, C-69, C-70, C-71.) As an example, sometimes doctors determined that patients needed direct 1:1 observation, but supervisors were not given enough staff to increase the total number of workers. (Tr. 489-90, 493-95, 868, 1218-19; Ex. C-58.) Instead, an MHS would be reassigned from the whole unit to remain with one patient. This left the other MHS responsible for all the other patients in the unit. (Tr. 160-61, 397-98, 490, 1514.)

It was not unusual for workers to be alone with patients. (Tr. 472-73, 484-88, 620, 791, 841-42, 962, 1394, 1424, 1452, 1515-16, 2557, 2559, 2561-62, 2596, 2614; Exs. C-9, C-34, C-51, C-54, C-92 at 13, 15.) For instance, therapists often worked alone in offices with patients. (Tr. 1413-15, 1425.) Sometimes, a therapist would run a group session for patients without any additional staff present. (Tr. 1414-15.) Staff were alone in units when others responded to codes, took breaks, or assisted with patient admissions. (Tr. 487-88, 511, 841, 962-63, 1022, 1515-16, 2386; Exs. C-51, C-79 at 7, C-92 at 15.)

To give an example, new patients had to undergo a body scan.¹⁴ This was a safety measure and part of patient care. (Tr. 467-68.) When OSHA’s investigation commenced, the process occurred in the units after admission with two staff members. (Tr. 2377, 2387, 2871; Ex. RCS-30.) On the overnight shift, a worker from another unit was often called over to perform the task since each unit only had one MHS. (Tr. 2377, 2387-88.) The responding person’s assigned unit would have only one worker during the scan. (Tr. 2377.)

Aggression or other behaviors sometimes lead to patients being restricted to the care unit. In such situations, the MHS would typically accompany the other patients to the gym or cafeteria while the nurse remained in the unit with the other patient or patients. (Tr. 485, 487-88, 791, 841-42, 2561-62; Ex. C-9, C-54, C-92.)

Ms. Nolan indicated that if there were few staff in a unit, the supervisor would sit directly in that unit. (Tr. 2596.) She did not explain whether there was a policy requiring this or if it was just her practice. Id. At night, there was only one supervisor. Id. The supervisor typically provided coverage in the various units when the nurses went on break and assisted in admitting new patients. (Tr. 2377, 2386-87; Exs. C-56, RCS-30.) Thus, even at baseline, the overnight supervisor already covered units at various points throughout each shift. (Tr. 962, 1515, 2377, 2386-87, 2596-98; Exs. C-43, C-46.) Ms. Degenhardt corroborated that the nurse supervisor had to provide coverage when unit nurses took breaks. (Tr. 504-5; Exs. C-43, C-55, C-56.)

Respondents were aware of injuries occurring when staff were alone in care units. In 2018, the Colorado Department of Public Health and Environment (“CDPHE”) found that workers were left alone and injuries were occurring. (Tr. 172; Ex. C-9.) The staffing levels described in the 2018 CDPHE citation were consistent with those during the OSHA inspection. (Tr. 172; Exs. C-9, C-36.)

Workers discussed attempting to manage aggressive patients by themselves. A therapist was alone when a patient attacked her. (Tr. 1413-14, 1425-29; Ex. C-65.) She had no radio or means to call a code. (Tr. 1414, 1432-33.) She received medical treatment for the bites incurred when she attempted to restrain the patient single-handedly.¹⁵ In another incident, on December 8, 2019, a single worker needed to restrain an aggressive patient when a fight broke out.¹⁶ The following month, an MHS was alone when a patient climbed over the nurses’ station. (Tr. 961.) The patient jumped on the MHS and struck him. (Tr. 961, 963.) The MHS needed to initiate a therapeutic hold by himself to prevent being struck again. Id.

A worker was alone the following month when a patient started behaving aggressively. (Tr. 1658-61, 2390; Ex. C-108.) As the worker backed away, the patient followed her down a hallway and grabbed her. (Ex. C-108.) A housekeeper came to assist, but the two workers could not stop the aggression. (Tr. 1658-61; Ex. C-108.) The patient kicked and hit the workers, causing them to fall to the floor. Id. The two workers struggled with the patient, unable to stop the aggression until additional workers came to the unit.¹⁷ Patients congregated and lingered around the incident as there was initially not enough staff to direct them away. (Tr. 1660-61, 2360-61; Ex. C-108.)

Police were needed to assist with aggression and elopements. (Tr. 414, 469, 518-19, 760-61, 776, 779, 995; Ex. C-2b.) They also sometimes brought potential patients to the Worksite. (Tr. 85, 90, 466, 468, 507, 519, 995, 1206; Ex. C-57 at 3.) Ms. Degenhardt instructed staff to call the police when a patient, after attempting to elope, began “bashing” a worker’s head onto the concrete outside of the care unit. (Tr. 414-15.) She also called the police in January 2020 regarding an extremely aggressive patient. She did not believe there was a way to safely attempt to hold him, even if multiple workers were available. She feared for the unit’s safety and did not feel she had adequate tools to handle the patient. (Tr. 518-19; Ex. C-59 at 2.)

While staff felt it was sometimes necessary to contact the police to facilitate worker safety, senior leadership did not support calling the police for assistance. (Tr. 416, 557.) Respondents maintain that staff “were free to contact the police if they wish to press charges against a patient.” (CS Br. 43.) The testimony cited in support of this claim relates to contacting police after an incident, as opposed to calling police to assist during acts of violence. (CS Br. 43; Tr. 557.) In that context, workers were not supported in contacting the police for assistance. (Tr. 416-17, 557, 762; Ex. C-2b at 12.) Ms. Degenhardt’s demeanor in making the statements strengthens this finding. After an incident, there was a specified waiting period and other steps workers had to take before it was permissible to press charges. (Tr. 416; Ex. C-2b at 12.) Witnesses discussed the lack of clarity on the permissibility of contacting the police. One nurse was unclear whether workers were permitted or encouraged to call the police. (Tr. 762.) She was unaware of any written policy on contacting the police. (Tr. 814-15.) Another described a situation where the police were called but were not permitted to enter because their presence had to be cleared by management. (Tr. 1041.)

There was no security to contact instead of the police. Workers perform security-related tasks, such as screening for contraband and responding to violence, but no one is designated as being responsible just for security and staff safety. (Tr. 263-64, 277, 360, 417, 618, 714, 1457, 1510, 1676, 2425; Exs. RCS-15, RCS-30, RCS-31, RCS-72.) Respondents encouraged staff to assess the contents of patient belongings either during the shift the patient first arrived at the unit or the following shift. (Tr. 360, 1570.) However, workers had many competing tasks, and the search occurred when staff was available. (Tr. 360.) During OSHA’s investigation, appropriate and timely searches did not consistently occur. (Tr. 361-62, 498, 1312, 1570, 1644; Exs. C-39, C-92 at 12, 22-23, C-94 at 24, 34-36.) Dangerous materials were brought into the units and remained there either because the initial search was inadequate or was not completed timely. (Tr. 498, 1644; Exs. C-39, C-92 at 12, 22-23, C-94 at 24, 34-36.)

The Worksite had a Workplace Violence Prevention Plan (“WVPP”) on its intranet. (Tr. 1822-23; Ex. C-22.) The Risk Management Director (Cristina Kolln) has responsibility for the WVPP and said it was in place when the inspection commenced. (Tr. 1817, 1822, 1854; Ex. C-22 at 1, 8; CS Br. 41; UHS-DE Br. 22.) The Worksite also had an employee handbook that noted workplace violence but did not specifically discuss patient-on-staff violence. (Tr. 1055, 1105; Ex. C-81 at 55-56.)

Aspects of the WVPP were either not followed or not implemented. The WVPP calls for the Worksite to report, record, and monitor events of patient aggression through, among other things, Healthcare Peer Review (“HPR”) Reports. (Ex. C-22 at 1.) HPR Reports were created electronically in the Worksite’s MIDAS database.¹⁸ The information collected focused on patients. (Stip. 38; Tr. 1824-25.) Worker injuries were not entered into it. (Stip. 38; Tr. 1824-25; Ex. C-2a.) The same section of the WVPP also called for reports and monitoring of events by the Performance Improvement Committee (“PIC”) and the Patient Safety Council (“PSC”). (Ex. C-22 at 1-2; UHS-DE Br. 24-25.) However, specific incidents of patient aggression toward staff were not typically discussed in those committee meetings.¹⁹ The WVPP also called for an annual evaluation of the Worksite’s violence prevention plans, but there is no evidence of such reviews before the Citation. (Tr. 1059, 1854, 1870; Ex. C-22 at 7.) In multiple respects, Ms. Kolln was unsure what the WVPP required or was referring to. (Tr. 1854-56.) Other workers did not recall seeing the WVPP or did not know it was in place. (Tr. 148-49, 695-96, 797, 1452-54.)

Although sometimes there was only one employee in a unit, additional assistance could be obtained to deal with aggressive patients. (CS Br. 36-39; Tr. 2793; Ex. C-3 at 42-43.) Staff could call a Staff Assistance Code to obtain “early intervention” to deal with an aggressive patient. Id. A trainer confirmed there was such a code, but no direct care employees discussed using such codes in practice, and it was not included in the list of codes in the annual re-training packet for direct care workers.²⁰

Rather than staff assistance calls, workers typically used the Code Green process. (Tr. 121; Ex. C-3 at 42-43.) These codes were used after verbal de-escalation and other techniques failed. (Tr. 2794; Ex. C-9 at 1.) To commence a Code Green, workers are trained to use the one phone in the unit to ask available staff to come to a specific location to deal with aggression or violence. (Tr. 142, 2392, 2888; Ex. RCS-5, RCS-18 at 5.) Although not in the training packet, workers also used radios to commence the process if they could access them. (Tr. 121, 141-42, 2888.) If staff could not access a telephone or radio, they could yell, “Code Green.” (Tr. 138, 144.) Sometimes, specific staff were designated as Code Green responders. (Tr. 1737, 1796.) Typically, an MHS rather than a nurse was designated. (Tr. 1737-38, 1797, 2388.) Such an assignment did not preclude assigning the employee tasks that would prevent their ability to respond to a Code Green promptly. (Tr. 1796-97; Ex. C-3 at 42.) Respondents subsequently revised their process and began using a De-escalation and Response Team to respond to emergencies. (Tr. 2847, 2883; Ex. C-94 at 38.)

Phones were located only at the nurses’ stations, and staff typically did not carry radios in the care units. (Tr. 101-3, 690-91, 720, 793-94, 1648; Ex. RCS-14.) Staff would yell for help if they could not reach the unit phone. (Tr. 138, 401, 691, 962-63, 1516, 1649, 2882.)

Therapists did not have radios. (Tr. 1414, 1649.) Laurie Martin, a therapist, indicated she did not know how to use the radios available to other staff. (Tr. 1414.) The lack of radios made it difficult for therapists to respond to codes. The overhead pages for Code Greens were challenging to hear from therapists’ offices, and they only heard the calls if another staff member happened to have a radio and was nearby. (Tr. 1419.)

Workers in the A&R unit also lacked consistent access to radios. (Tr. 1648-49.) The A&R department only had one radio. (Tr. 137-39.) When a worker had to escort a patient from A&R to a care unit, they could not always carry a radio. (Tr. 101, 138-39, 1648-49, 1672-73.) Within the A&R department, one of the two assessment rooms had a panic button that could be used to request assistance from staff. (Tr. 103.) The button did not alert the police, and there were no security personnel. There were no similar buttons in the care units. (Tr. 103, 1507-8, 2392.)

Respondents had a policy on using, storing, and monitoring the two-way radios. (Tr. 2888‑89; Ex. RCS-14.) However, the policy was out of date. (Tr. 2889.) Although carrying a radio when escorting patients off a unit was required, at times, there were not enough functioning radios for every staff member to have one when they needed to leave the unit. (Tr. 102, 694, 793-94, 1648.) Radios were often missing, uncharged, or broken. (Tr. 102, 506, 508-9, 694-95, 1630; Exs. C-42, C-57.) Issues with radios complicated the response to codes. (Tr. 506, 508-9, 694-95, 794; Exs. C-42, C-57 at 4.) Workers informed supervisors and senior management of issues with radios and unit phones. (Tr. 506, 509; Exs. C-42, C-43, C-50, C-57.)

Respondents trained staff to observe patients and institute various precautions. (Tr. 536, 742, 2436-38, 2440, 2455, 2857; Exs. RCS-7, RCS-8, RCS-21.) An MHS observed each patient at least once every fifteen minutes.²¹ A doctor must order the frequency of the observations. (Tr. 2433-34, 2651.) The MHSs assigned to conduct the observations record their findings on a form (sometimes called a tracker or check log). (Tr. 616-19, 705, 2328, 2433; Exs. RCS-7, RCS-32, RCS-40.) The information recorded was “a very quick view” of what “might be going on” with the patients. (Tr. 2441.) No completed forms were part of the record. When assigned to perform these checks, the MHS remained in one unit and did not respond to Code Greens elsewhere. (Tr. 121-22, 177-78, 180, 1514, 2373-74; Exs. C-92 at 14; RCS-7 at 1, RCS-32.) The observation checks were supposed to ensure “we always know where everybody is.” (Tr. 2329, 2433.) The care team also can use the checks to assess whether patients were sleeping, appropriately engaged, or isolating themselves. (Tr. 2329, 2439.) They were for the patient’s safety. (Tr. 2328, 2433.)

Respondents had a written policy about possible precautions staff could implement to protect patients. (Tr. 1630, 2341-42, 2434-35, 2439, 2553-54; Exs. RCS-8, RCS-21, RCS-40.) The Patient Precautions Policy defined assault and other precautions. (Ex. RCS-8 at 3.) A doctor could order assault precautions after assaultive behavior, including threats and gestures. (Tr. 839, 2435-36, 2438-39; Ex. RCS-8.) At the investigation’s start, the policy did not instruct staff about what they should do differently for the patients on assault precautions compared to those who were not. (Tr. 839, 842, 2552; Ex. RCS-8.) Assault precautions did not automatically result in additional staffing. (Tr. 839, 842.) The policy detailed specific actions to alert staff to other issues even if they did not have the paper medical file. (Ex. RCS-8.) Colored armbands and magnets on bedroom doors were used for those on fall, seizure, or infectious control precautions. (Tr. 2553-55; Ex. RCS-8.) In contrast, workers were not always aware of assault precautions. (Tr. 1251-53; Ex. C-92 at 22.)

Respondents trained workers in de-escalation and restraint techniques and provided certain re-training. (Tr. 2880; Exs. RCS-17, RCS-18.) During the investigation, workers underwent a training program called Satori Alternatives for Managing Aggression (“SAMA”). (Tr. 2885.) SAMA training occurred over two days with demonstrations and participation. (Tr. 2856-57; Ex. RCS-22.) Workers were “recertified” on SAMA every six months. (Tr. 2859; Ex. RCS-18.) SAMA techniques generally require at least two people. (Tr. 640, 2853.) As part of the training, workers had to acknowledge, in writing, that “[s]ingle person restraints with patients are prohibited unless emergency criteria are met (life-threatening situation or small child).” (Ex. RCS-22 at 39.) Despite this limitation, workers felt they often had no choice but to initiate holds alone. (Tr. 623-25, 640-41, 963.)

In 2021, a different de-escalation and restraint training program from the Crisis Prevention Institute (“CPI”) was implemented. The CPI approach to restraints precludes solo restraints and generally requires more people than the SAMA approach. (Tr. 2853, 2890.) Mr. Mauro indicated that he did not decide to switch from SAMA to CPI. (Tr. 2844, 2884-85.) He described the change as a “corporate requirement.” (Tr. 2884-85.)

Many patients exhibit no signs of aggression and never engage in assaultive behavior. (CS Br. 11.) Still, highly aggressive patients were receiving treatment during the inspection period. Id. at 18. Patient 23 was an example of an assaultive patient.²² Over four days, the patient needed to be restrained 14 times. (Ex. C-64.) Several of those restraints occurred after assaults on staff. (Ex. C-64.) He jumped through the nurse’s station window and hit and kicked the staff who tried to escort him from the area. (Tr. 925; Ex. C-64 at 1.) The following morning, he choked a staff member. (Tr. 926; Ex. C-64 at 6.) Later in the day, he punched an MHS in the chest and body. (Tr. 636-40; Ex. C-64 at 12-15.) In the evening, he fought again with staff, necessitating another restraint. (Tr. 649-50; Ex. C-64 at 18-20.) In a subsequent incident, he climbed up the front of the nurses’ station and entered by forcing open the window. (Tr. 623-25, 651-52; Ex. C- 64 at 26.) An MHS was injured as he tried to restrain the patient. (Tr. 623-25, 651-52; Ex. C- 64 at 27.)

The next day, multiple incidents with Patient 23 continued to result in staff injuries. In the morning, the patient threw hot chocolate at staff, repeatedly hit an MHS, and kicked a housekeeper who tried to intervene to stop the assault. (Ex. C-64 at 37-38, 41.) In the afternoon, he hit a worker, and then in the evening, he punched and head-butted a lead MHS. Id. at 42, 47-50. At night, he climbed over the nurses’ station again and jumped on the lead MHS. (Tr. 965-68, 975-76; Ex. C-64 at 55.) By this point, the patient had been moved to a different unit, but the lead MHS did not observe any other changes to the management of the patient’s aggression. (Tr. 971.) About an hour after his third climb into the nurses’ station, he punched a worker. (Ex. C-64 at 56.) His behavior continued into the next day. By about 8:00 a.m., he had been restrained four more times. Id. at 61, 63, 66, 73. This included two more incidents at the nurses’ station and staff assaults. Id.

Like the lead MHS, another MHS involved with multiple incidents with Patient 23 did not observe any measures to address the patient’s aggressive behavior adequately. (Tr. 671.) “Master Treatment Plan Update” forms were completed after each restraint of Patient 23. (Ex. C-64.) The updates were for patients, not staff safety, and MHSs did not review them. (Tr. 649, 1434.) Numerous Master Treatment Plan Updates for Patient 23 were not finalized until 24 days after the restraints occurred. (Ex. C-64 at 2, 11, 25, 35, 40, 46, 52, 54, 67, 75, 76.) Dr. Gray, the Medical Director, signed eleven updates, each related to a different restraint of Patient 23, weeks after the restraints occurred. (Tr. 2656.) He offered various theories but did not know why he signed them long after the restraints occurred. (Tr. 2653-56.)

Respondents believe that reducing restraints reduces the risk of employee injury. (Tr. 229-30, 2525, 2851; Ex. RCS-41.) The Worksite had a restraint reduction team. The team was largely non-direct care workers, even though 90% of those injured by workplace violence were in direct care roles. (Ex. C-3 at 11.) Before the inspection commenced, the only direct care worker on the team was a trainer.²³ The Risk Management Director claimed that the restraint reduction team met at least monthly at the start of OSHA’s investigation. (Tr. 2787, 2828.) However, the records do not show any meetings during the inspection period, and only four meetings in 2019. (Tr. 2828; Ex. RCS-41.)

Respondents had a written procedure for reporting aggression.²⁴ The policy called for reporting staff “injuries” as well as “safety concerns” and “unusual incidents” of “sufficient significance.” (Ex. RCS-3 at 1-2.) The policy does not elaborate on what injury means or whether the safety concern must be ongoing to be reportable. (Ex. RCS-3.) Immediate notification of the Administrator On Call was required for patient injuries but not worker injuries. (Tr. 2823-24; Ex. RCS-3 at 2.) The policy does not refer to the OSH Act, including its injury recording obligations. (Ex. RCS-3.)

Respondents also had a one-page flyer that directed injured employees to call a hotline to have their injury assessed for the need for medical treatment and possibly to answer questions about the “accident.” (Tr. 2729; Ex. RCS-23.) The number on the flyer was for Sedgwick, a third-party contractor Respondents retained.²⁵

Respondents’ policy indicates that the “Safety Committee” will review trends and act to decrease or eliminate safety hazards. (Ex. RCS-3 at 3.) There is no evidence of a “Safety Committee.” (Tr. 154-55, 1634-35; Ex. C-3 at 34-35, 38.) The Safety Solutions Team stopped meeting nearly a year before the inspection commenced, and there was no committee focused on employee safety and health concerns. (Exs. RCS-84 at 8, C-3 at 34.)

The rate of patient-to-staff aggression is noted in the minutes of a different committee, the Performance Improvement Committee (PIC). (Tr. 2802-5; Ex. RCS-84.) This rate is based on the number of injuries to staff from patient aggression per 1,000 patient days. (Tr. 2802-4.) The Worksite has approximately 2,000 patient days per month. (Tr. 2805.) In 2019, the rate of such aggression ranged from 1.57 to 7.10.²⁶ The PIC minutes do not note specific staff injuries or discuss incidents that led to such injuries. (Ex. RCS-84.) Similarly, the restraint reduction team’s minutes note patient aggression toward staff but make no mention of worker injuries.²⁷

Although not used to track or report worker injuries, a nurse was supposed to complete a “packet” of forms after a restraint or seclusion. (Tr. 418-19, 560, 630, 965, 1469, 2603-4; Exs. C-64, C-65, C-66.) The information collected focused on the patient and their perception of events and triggers. (Tr. 418-19, 1434; Exs. C-3 at 41, C-64, C-65 at 5.) There was a space to indicate the staff involved in the debriefing, but it was not consistently completed. (Tr. 651, 927, 966-67, 1471-73; Ex. C-64.) The same is true for a section requesting information on what measures would permit better management of future incidents. (Ex. C-64 at 9, 23, 36, 59, C-66 at 6.) The information was to be used to update the patient’s Master Treatment Plan, not worker safety. (Tr. 1433-34; Ex. C-3 at 41.) This documentation typically did not include worker injuries “because there were so many assaults” and because employee injuries did not “belong” in the medical file where the information was ultimately placed. (Tr. 419, 560, 630-31; Ex. C-2b.) The packets were not aggregated, so reviewing all reported information about assaultive incidents would require referring to individual patient records. (Exs. C-2b at 13, C-3 at 48-49, C-94 at 11, 20.) MHSs were not required to view the patient medical files, and it could take a while for the post-restraint packet to be added to the file. (Tr. 560, 1253, 1431, 2554; Ex. C-92 at 22.) Likewise, MHSs did not review Master Treatment Plan updates. (Tr. 649, 2656.) It could take time to complete Master Treatment Plan updates and insert them into the individual paper medical file. (Tr. 649, 2656; Ex. C-64.)

Respondents contend that management audits the incident reports to ensure they are completed. (CS Br. 44 citing Tr. 321 and Ex. RCS-84.) Neither the PIC meeting minutes nor the CO’s testimony sufficiently support that contention. The CO rejected Cedar Springs counsel’s suggestion that incident reports were audited daily and limited her testimony to expressing her belief that the reports were audited periodically. (Tr. 321.) Further, the CO’s testimony was about the restraint and seclusion packets, which generally did not include worker injuries. (Tr. 560.) The record does not reflect auditing of the Sedgwick Reports or other reports made under the Worksite’s incident reporting policy (Exhibit RCS-3.).

Post-incident debriefings occur when patient behavior leads to them being restrained or put into seclusion. (Tr. 182-83; Ex. C-3 at 44-45.) As noted, the debriefings focus on patient feedback and patient safety. (Tr. 182-83, 1434; Exs. C-3 at 45-46, C-64, 65, C-66.) They are not about worker injuries. (Tr. 560.) A nurse typically completes a debriefing form within thirty minutes of the restraint or seclusion. (Tr. 630, 2344.) The Director of Nursing reviewed them before the debriefing forms were eventually placed in the patient’s medical file. (Tr. 560, 630-31, 2523.) Ms. Nolan sometimes completes the debriefings rather than those directly involved. (Tr. 2344.) She might only be able to gather one person to conduct the debriefing because staff would need to return to their respective units or have other duties. (Tr. 2344; Ex. RCS-29 at 4, 19-22.) Workers discussed violent incidents they were involved in but for which they did not participate in any post-incident debriefing or investigation. (Tr. 183, 643, 651, 672-73, 966-67, 983, 992, 1420, 1438, 1471-73, 1495, 1507, 1510; Exs. C-3 at 45.)

The debriefing forms in the record support the testimony about inconsistent debriefing practices. (Exs. C-64, C-65 at 5, C-66 at 5.) Most of the one-page forms have multiple blanks. (Exs. C-64 at 5, 9, 15, 25, 36, 39, 59, C-65 at 5, C-66 at 5, C-79 at 5.) Sometimes, the debriefing forms do not identify who participated in the debriefing. (Ex. C-64 at 5, 9, 23.) When people were listed, often there was a disconnect between those involved in the restraint and those who participated in the debriefing. (Ex. C-64 at 15, 23, 59, C-65 at 5, C-66 at 5.) For example, sometimes Ms. Nolan was an observer or participant but was not involved in the debriefing.²⁸

Respondents also obtained incident information by reviewing events captured on the Worksite’s video surveillance system. The Risk Management Director would be informed of restraints or seclusions and look for video of the incidents. (Tr. 1834, 1898-99, 1903-6, 2762-63; Ex. C-75.) Workers frequently did not have an opportunity to review video footage of incidents during which they were injured by patient aggression or violence. Therapists, nurses, and MHSs indicated they never saw videos of incidents of workplace violence they experienced or saw. (Tr. 672-73, 1437, 1476-77, 1507, 1510, 1541-42.) Witnesses discussed violent incidents they believed were caught on camera but for which they never saw any video footage. (Tr. 672-73, 682, 771, 972-73, 979-80, 983, 991-92, 1437, 1476-77, 1541-42.) Other documents support the workers’ testimony that reviewing camera footage of incidents was not consistently done. (Exs. C-79 at 8; RCS-41 at 5.)

The majority of patients arrive via ambulance and are brought directly to a care unit. (Tr. 2310-11, 2419-20.) At the start of OSHA’s investigation, there were no metal detectors in care units. (Tr. 258-59, 466, 468-69, 715, 1501, 1569, 2310-11, 2428, 2430; Exs. C-38, C-39 at 1.)

The process was different for individuals voluntarily coming to the A&R department. Upon arrival, the person would speak with a receptionist and then wait in the lobby with their possessions. (Tr. 90-91.) There was no search before the person was brought into the assessment rooms. If the person was admitted after the assessment, a metal detector wand was supposed to be used over their clothes and the exterior of their belongings. (Tr. 91; Ex. RCS-15.) Staff would do a “cursory” search of the patients and their bags with a metal detecting wand. (Tr. 176, 260, 359, 1204, 1312, 1629.) Neither patients nor their belongings were thoroughly searched in the A&R department. (Tr. 176, 1204, 1569, 1672-73; Ex. C-39.) One of the A&R workers would then walk the patients across the grounds to the care unit. (Tr. 91, 896.) Staff or patients would carry the belongings to the building with the care units. (Tr. 91, 896, 1204, 2551, 1672-73; Ex. C-39.)

Regardless of how patients were admitted, the Worksite’s policy was to inventory patient possessions after the patient got to the care unit. (Tr. 2892; Exs. RCS-30 at 2, RCS-31.) Workers did not do an immediate inventory but tried to get to it at some point during the shift. (Tr. 1570, 2321, 2377.) Patients’ belongings were sometimes kept in the nurses’ stations for long periods before being searched. (Tr. 360, 497-98, 897, 1313, 1377, 1488-89, 1502, 1570, 1629, 1644, 1672, 2320-22, 2377, 2892.) While most of the nurses’ stations were typically locked, the station in one unit did not lock. (Tr. 2378.) No one knows the contents of those belongings kept behind the nurses’ station until they are searched. (Tr. 2548-49; Ex. C-39.)

Patients themselves were typically searched more quickly for contraband and weapons. (Tr. 467-68, 715, 2319-20; Ex. RCS-30 at 2.) At the investigation’s start, patients would undergo a body scan after arriving in the unit. (Tr. 715, 2319-20, 2871; Ex. RCS-30 at 2.) Respondents changed their procedures after the investigation commenced. (Tr. 2425.) The belongings are now secured before the body scan and entry into the care unit. Id.

Respondents raise several unsupported claims about the CO and the appropriateness of OSHA’s investigation. They misrepresent the record and make sweeping allegations of “bias” and “misconduct” without grounding them in any particular legal theory or appropriate support. (CS Br. 13-14.)

An employee contacted OSHA’s Whistleblower Office with concerns about hazards at the Worksite. (Tr. 64; Ex. C-2a.) That office then referred the complaint for further investigation. Id. Ms. Stark received the referral and held an opening conference with the Nursing and Risk Management Directors on November 18, 2019. (Tr. 71-72; Ex. C-3 at 4.) She returned on subsequent dates to tour the Worksite and to speak with workers. (Tr. 73-74; Ex. C-3.) She interviewed management, including the CEO, three directors, and approximately fourteen other workers.²⁹ Ms. Stark had previously inspected a behavioral health hospital and had conducted other inspections involving workplace violence allegations. (Tr. 64.)

OSHA spent months gathering evidence, reviewing documents, and talking with employees before issuing the Citations. (Tr. 73-75; Exs. C-2a, C-3, C-4.) Additional information was also sought via subpoena. (Tr. 81-83; Exs. C-6, C-7, C-8.) The requested documents included incident reports, information related to debriefings, intake screenings, behavior plans, and seclusion and restraint packets. (Tr. 81-82; Exs. C-6, C-7, C-8.) The CO reviewed the OSHA 300 logs and logs of injuries employees reported via a hotline. (Tr. 107-19.) She concluded that employees in direct contact with patients suffered serious workplace violence-related injuries. (Tr. 106-7, 115-23.) These included a high number of injuries serious enough to require lengthy days away from work and restricted duty. (Tr. 110-15; Ex. C-3.) After the site visit, interviews, and document review, the CO believed there was a violation of the general duty clause for the hazard of workplace violence and recommended the citation’s issuance. (Tr. 104-5, 123.)

Some documents were not produced to OSHA within four hours of the request, some requests were refused, and Respondents did not preserve some responsive information. (Tr. 198-201; Exs. C-6, C-7.) After multiple requests and a subpoena, Respondents failed to timely provide all requested illness and injury records. (Tr. 190-94; Exs. C-3, C-4, C-7.) OSHA also did not receive video footage of all incidents of patient-to-staff aggression as demanded in the subpoena. (Tr. 82-85; Exs. C-6, C-8.) Videos related to several such incidents were destroyed and unavailable during discovery. (Tr. 82-85.)

Respondents cite no caselaw in support of their misconduct claims. (CS Br. 13-14.) Further, they mischaracterize the record when discussing the CO’s interaction with one worker. Id. at n.8. The transcript contradicts their argument. If the events counsel described had occurred, that “would be concerning.” (Tr. 452.) But no witness or document supported counsel’s theory. There was no finding that the conduct counsel theorized occurred. (Tr. 428, 457.) Further, there is no contention that the emails provided to OSHA were protected from disclosure to the agency. Although the emails were not disclosed in discovery, Respondents do not assert that they were unresponsive or irrelevant to the Secretary’s valid discovery requests. Nor do they claim any privilege protected them from discovery. (Tr. 456-57.) Respondents’ misconduct claims do not withstand scrutiny and are dismissed with prejudice. There was ample evidence to support the Citations.

“At the time of the inspection, Cedar Springs had surveillance cameras located in the common areas of the facility and units. No cameras [were] located in bathrooms or patients’ rooms.” (Stip. 46.) The cameras monitor common areas, including hallways, nurses’ stations, the lobby, group rooms, the gym, and the cafeteria. (Tr. 103-4, 1414, 1898, 1905-6, 2771; Stip. 46.) The video is in color but does not include sound. (Tr. 1717; Exs. C-106, C-106, C-108.) The preserved clips are sufficiently clear to see various parts of the units and individual faces. (Exs. C-106, C-107, C-108.)

Video can be watched in real-time, but that was not often done. (Tr. 104, 1073, 1206, 1898; Ex. C-3 at 42.) Instead, it is recorded and automatically saved for thirty days so that it can be reviewed later. (Tr. 1073, 1898-99; Resp. Opp’n to Sanctions Mot. 3.) At or before thirty days, the Risk Management Director can save the video before it is deleted. (Tr. 1898-99, 2763-67; Resp. Opp’n to Sanctions Mot. 3.)

Video is routinely reviewed and saved for various reasons. (Tr. 883-84, 916, 1476-77, 1832, 1834-35, 2762-66, 2771; Exs. C-14 at 3, C-19, C-20, C-22, C-75; RCS-41 at 4, 6.) Senior managers conducted camera audits by reviewing videos from different shifts at least once a week. (Ex. C-75.) Respondents claim these were a primary focus area to address staff safety. (CS Br. 45 n. 41.) In practice, the audits focused on patient safety, ensuring workers perform tasks, and confirming overtime was needed. (Tr. 510, 698, 1476-77, 1831-32, 2771; Exs. C-43, C-75.) The audits considered whether the rounds were timely, whether workers entered rooms and used a flashlight, and whether they completed the related documentation. (Ex. C-75 at 2.)

Separate from the audits, video of all restraints and seclusions was supposed to be reviewed under the Worksite’s WVPP. (Tr. 883, 1834-35, 1899, 1903, 1906, 2761-62; Exs. C-19, C-20, C-22 at 3, C-44, C-48, C-75 at 6-7, RCS-41.) During OSHA’s investigation, management also reviewed camera footage of other incidents and for other reasons. (Tr. 899, 910-16, 1913-14; Exs. C-14 at 3, C-19, C-20, C-39 at 2, C-44, C-48.) For instance, camera footage was reviewed to assess the accuracy of a worker’s claim about how an injury occurred. (Tr. 913-16; Ex. C-48.) Management also reviewed video related to an instance when a patient successfully brought a knife into the care units. (Tr. 899; Ex. C-39 at 2.)

OSHA commenced its investigation on November 18, 2019. Two days later, on November 20, 2019, it served a subpoena. (Tr. 83.) The subpoena was consistent with an evidence preservation letter OSHA sent to Respondents on November 21, 2019. (Sanctions Mot. 2.) The letter directed Respondents to “not remove, intentionally alter, destroy, rearrange, or otherwise tamper with any surveillance video footage or recordings” at the Worksite. (Sanctions Mot. Ex. 2 at 1-2.) It instructed Respondents to “take all steps necessary to identify, retain, and preserve all ‘potentially relevant information’” to OSHA’s investigation. Id. at 2. On January 10, 2020, the CO reiterated OSHA’s concern that video evidence be maintained during the investigation. (Sanctions Mot. Ex. 3.) Litigation formally commenced on June 4, 2020, when Cedar Springs submitted its Notices of Contest.³⁰

During discovery, the Secretary learned of deleted videos. On December 2, 2021, she moved for Sanctions for the lost electronically stored information (ESI). The Motion was denied without prejudice in the undersigned’s Order on Complainant’s Motion for Sanctions for Respondents’ Destruction of Relevant Recorded Video Evidence (“Order”).

As the Order permitted, the Secretary renewed her request for sanctions after the hearing. (Sec’y Br. 130-34.) The Secretary relies on witness testimony and other records to support the spoliation claims. In their brief, Respondents refer to the “claimed destruction” of videos of workplace violence and claim that any destruction did not prejudice the Secretary. (CS Br. 13, n. 7.) They do not address the records that refer to videos that were not produced. Id.

As the Order stated, Respondents had a duty to preserve video evidence as of November 20, 2019, at the latest. (Order 3, 6.) See Scalia v. UHS of Fuller, Inc. & UHS of Delaware, No. 1:19-mc-91541-FDS, Document No. 42 (D. Mass. Jan. 6, 2020) (Magistrate Judith Gail Dein entering an order directing UHS-DE to preserve “any existing video surveillance footage,” which U.S. District Court Judge F. Dennis Saylor subsequently adopted); Phillips Elecs. N. Am. Corp. v. BC Tech., 773 F. Supp. 2d 1149, 1167 (D. Utah 2011) (party had a duty to prevent ESI from being inadvertently or negligently overwritten). The duty to preserve continues throughout the litigation. A party cannot continue its otherwise routine procedure of destroying evidence. Emily Sharp Rains v. Westminster College, et al., No. 2:20-cv-00520, 2023 WL 2894506, at *1 (D. Utah Apr. 11, 2023); Peskoff v. Faber, 251 F.R.D. 59, 62 (D.D.C. 2008) (threat of litigation should have prompted the defendant to “deactivate network maintenance tools that automatically delete electronically stored information”).

Federal Rule of Civil Procedure Rule 37(e) “authorizes and specifies measures a court may employ if information that should have been preserved is lost, and specifies the findings necessary to justify these measures.” Advisory committee’s note to the 2015 amendment. Under Rule 37(e), relief is available when ESI should have been preserved in anticipation of litigation and the party failed to take reasonable steps to preserve it. Rule 37(e) only applies if the ESI cannot be restored or replaced through additional discovery. The rule then bifurcates the appropriate response to the loss depending upon whether the party who failed to take reasonable steps acted with the “intent to deprive” its opponent of the information’s use in litigation. See Equal Employment Opportunity Comm’n v. JetStream Ground Servs., Inc., 878 F.3d 960, 965-66 (10th Cir. 2017) (relating the requirements of Rule 37(e)(2) to another context); In re Google Play Store Antitrust Litig., No. 21-md-02981-JD, 2023 WL 2673109, at *1-2, 7-10 (N.D. Cal. Mar. 28, 2023) (finding sanctions appropriate when the party did not take reasonable steps to preserve ESI after litigation was reasonably anticipated); Integrated Commc’ns & Tech., Inc. v. Hewlett-Packard Fin. Servs., Co., No. 16-0386-LTS, 2020 WL 4698535 (D. Mass. Aug. 13, 2020) (applying First Circuit precedent and imposing sanctions under Rule 37(e)(1) when a party did not backup emails or take other steps to preserve the information).

The Third Circuit addressed Rule 37(e)’s applicability and upheld sanctions a judge issued under Rule 37(e)(2). GN Netcom, Inc. v. Plantronics, 930 F.3d 76, 82-83 (3d Cir. 2019) (upholding judge’s decision to impose Rule 37(e)(2) sanctions short of the requested default). Similarly, in an unpublished opinion, the D.C. Circuit upheld the rule’s applicability to ESI destruction. Ball v. George Washington Univ., 798 F. App’x 654, 654-56 (D.C. Cir. 2020) (unpublished) (upholding judge’s decision not to issue sanctions after the destruction of ESI).

Neither the Commission nor the Tenth Circuit directly discussed Rule 37(e) in binding precedent. However, Commission ALJs have.³¹ For example, Judge Phillips applied Rule 37(e) and awarded sanctions for ESI destruction. UHS-DE/Suncoast, 2023 WL 2388069, at *40-43. He imposed curative measures and sanctions for lost video from patient care units at a behavioral health facility. Id. at *16-43. The Commission affirmed the ALJ’s decision without discussing his award of sanctions. 2023 WL 2388069, at *1, 10.

As for the Tenth Circuit, its pre-2015 precedent largely tracks the provisions of Rule 37(e)(2) in that it requires a finding of “bad faith” related to the spoliation. The Tenth Circuit’s post-2015 position on imposing sanctions under Rule 37(e) is not yet fully crystallized. See Helget v. City of Hays, Kansas, 844 F.3d 1216, 1226 n.7 (10th Cir. 2017) (noting in a different context that the 2015 amendment provided guidance on sanctions for the destruction of electronic evidence); Jetstream, 878 F.3d at 965-66 (relating the requirements of Rule 37(e)(2) to another context). In particular, the Tenth Circuit has not addressed a sanction request under Rule 37(e)(1), which does not require a showing of intent to deprive. While there may not be direct Tenth Circuit precedent applying Rule 37(e), numerous district courts have consistently applied the rule to analyze the loss or destruction of ESI. The undersigned will do so as well.

Rule 37(e) requires a multi-step inquiry. First, did the party fail to take reasonable steps to preserve the ESI? GN Netcom, 930 F.3d at, 82-83. Second, can the lost ESI be restored or replaced? Id. Third, was there prejudice from the loss? Id. If so, what actions will redress that prejudice? Id. Fourth, did a party act with the intent to deprive another of the information’s use in litigation? Id.

Respondents failed to take reasonable steps to preserve the ESI. Their approach resulted in lost ESI that cannot be restored or replaced. The loss prejudiced the Secretary. Respondents also intentionally failed to preserve ESI with an intent to deprive the Secretary of the ability to use such evidence at the hearing.

The Secretary showed that Respondents’ process to preserve the subpoenaed videos of the hazard was neither adequate nor reasonable. Even without an ongoing OSHA investigation, Respondents’ policies required reviewing video of restraints and seclusions. (Tr. 883-84, 1834, 1903, 1906, 2762-63; Exs. C-22 at 3, C-44, C-75.) The Risk Management Director (Kolln) reviews incident reports each day. (Tr. 1837-38, 1904-5.) If the reports indicate a patient was restrained, she looks for and then watches video of the event. (Tr. 1903, 1906.) Such video reviews are supposed to be documented. (Ex. C-75 at 6.)

Ms. Kolln’s testimony is consistent with the HPR incident logs, which make repeated references to reviews of videos from incidents.³² The logs set forth several incidents for which video should have been in Respondents’ possession as of November 20, 2019, or later. (Exs. C-19 at 24-27; C-20 at 23, 28, 36.) Ms. Kolln confirmed that there was a video of an incident if that is indicated in the HPR log. (Tr. 1907-9; Ex. C-20.) However, she could not recall if she saved the videos referred to in the logs. Id. See Franklin v. Stephenson, No. 20-cv-0576 MIS-JFR, 2022 WL 6225303, at *8 (D.N.M. Feb. 16, 2022) (noting that even though a party reviewed the video for a different purpose, the failure to retain the watched video demonstrated that the party did not take reasonable steps to preserve the evidence); Mueller v. Swift, No. 15-CV-1974-WJM-KLM, 2017 WL 3058027, at *5 (D. Colo. July 18, 2017) (awarding sanctions when party was “unjustifiably careless” in handling ESI).

After OSHA’s investigation commenced, Ms. Kolln was instructed to retain video of patient-to-staff aggression. (Tr. 1899.) At some point, she stopped doing that and only attempted to retain video of incidents in which workers were injured. (Tr. 1899, 2821-22.) She could not explain when or why that change occurred. Id. Under either criterion for saving, i.e., employees being injured or not, videos of many events were not produced. Workers reported injuries from patient aggression on January 3rd, 6th, and 7th. (Tr. 622, 672, 909, 913, 961, 979-80, 1900-2; Exs. C-13, C-14, C-44, C-64.) Ms. Kolln and the workers involved agreed that the incidents should have been visible on camera. (Tr. 672-73, 682, 1474, 1476, 1900-3, 1906.) The January 7, 2020 incident report specifies that camera footage was available. (Ex. C-14 at 3.)

Despite this, Ms. Kolln could not recall if she reviewed or saved video related to those incidents. (Tr. 1900-2, 1914; Ex. C-44.) Overall, Ms. Kolln could not specify how many videos she reviewed or tried to locate from when litigation was first anticipated through the close of discovery. (Tr. 1899-1914.) She did not keep a list of videos she watched or complete the required form after her reviews.³³ She did not note any videos she saved after she was instructed to preserve all videos of incidents involving patient aggression against staff. (Tr. 1899.) Similarly, she did not document which incidents she looked for video but did not find any. (Tr. 1899, 1902, 1905-9, 2822.) Without such records, it’s unsurprising that she could not identify the incidents for which she reviewed camera footage. (Tr. 1905-7, 1914.) See In re Google, 2023 WL 2673109, at *8-9 (failure to check whether employees were appropriately preserving text messages was evidence of a failure to take reasonable steps).

The amount of video produced is far smaller than the number of incidents for which the log incident indicates there was video. (Tr. 112, 1907-9; Ex. C-20.) Most incidents occurred in areas with cameras, including the nurses’ stations, common areas, and hallways, suggesting video would be available based on the numerous cameras and their locations. (Tr. 1900-1, 1905-6, Exs. C-19, C-20.) Ms. Kolln offered various theories as to why footage might not be available. However, again, she provided no specifics on instances where she tried to view footage, but the cameras did not capture anything. (Tr. 1899-14.) See Graham v. City of Lone Grove, Okla., No. 19-CV-000298-JFH, 2022 WL 2276337, at *5 (E.D. Okla. June 23, 2022) (finding that the failure to check to see if video was saved constituted failure to take reasonable steps), appeal dismissed, sub nom. Graham v. Hensley, No. 22-7032, 2022 WL 18107698 (10th Cir. Dec. 16, 2022) (parties settled); UHS-DE/Suncoast, 2022 WL 2388069, at *22-23, 32-35. Initially, Ms. Kolln indicated that a camera might not function on occasion. (Tr. 1901-3.) She later suggested there might be an issue with a camera two or three times a month. (Tr. 1912-13.) The incident logs sometimes note that one camera was not working but others captured the same incident. (Ex. C-19 at 27.) At other times, multiple cameras recorded aspects of the same event. (Tr. 104, 1485-86, 1558, 2764, 2766; Exs. C-106, C-107; Sec’y Mot. 5.) Most of the time, a hundred or more cameras are available. (Tr. 1486, 1898, 1913; Exs. C-3 at 21, 42, C-19, C-20; Resp. Opp’n to Sanctions Mot. 2-3.) Even accepting Ms. Kolln’s uncorroborated testimony on the frequency of camera unavailability, there remains a significant mismatch between the number of workplace violence incidents and the number of incidents for which video was produced.³⁴

Respondents’ failure to produce even the simplest list of videos searched for or saved is telling. (Tr. 1903-4, 1906-7.) They chose not to suspend their routine destruction of ESI after being subpoenaed and after litigation formally commenced. Respondents knew OSHA was interested in the hazard of workplace violence. Without ongoing litigation or investigations, Respondent reviewed video of all incidents resulting in employee injuries, patient restraints, patient seclusions, and as part of UHS-DE visits. (Tr. 1686, 1899, 1903, 1906, 2762, 2766-67; Exs. C-22 at 3, RCS-41.) Respondents knew the video system captured and automatically saved video of workplace violence incidents for thirty days. (Tr. 1898-99.) Workers would access such videos, review them, and then could have saved them to permit the Secretary to do the same. (Tr. 1898-99, 2766-67.) Instead, on multiple occasions after the subpoena was served and litigation was reasonably anticipated, no one did so.

Incidents occurred in places the video camera system covered but for which no video was produced. (Tr. 104, 771, 973, 979-80, 983, 991-92, 1437, 1507, 1541.) The Secretary showed a significant discrepancy between the number of incidents during the inspection period for which Respondents claimed to review video and the number of incidents for which video was provided to the Secretary. (Exs. C-19, C-20, C-49, C-64, C-66.) Respondents’ records repeatedly refer to videos being reviewed without a corresponding production of the same evidence for the Secretary’s review. (Tr. 1908-9; Ex. C-20.)

Respondents were on notice of the litigation and the relevance of video related to workplace violence at least by November 20, 2019. (Order 3.) The failure to preserve videos after litigation was reasonably anticipated of incidents that Respondents would review in the ordinary course of business was not reasonable. They failed to intervene and allowed ESI to be destroyed. The deletion of videos was not accidental or beyond Respondents’ control. See In re Krause, 367 B.R. 740, 766 (Bankr. D. Kan. 2007) (sanctioning debtor who continued routine deletion of emails and failed to deactivate “wiping” software which routinely removed information), aff’d, 637 F.3d 1160 (10th Cir. 2011); Philips Elecs. N. Am. Corp. v. BC Tech., 773 F. Supp.2d 1149, 1197 (D. Utah 2011) (sanctioning defendant who had a duty to preserve ESI by preventing it from getting lost, inadvertently overwritten, or wiped out). They routinely reviewed video surveillance outside of active litigation and had procedures to save video. (Tr. 913, 915, 1476-77, 1686, 1834, 1899, 1903, 1906, 2762, 2766-67; Exs. C-14 at 3; C-19, C-20, C-22, C-75, RCS-41 at 4, 6.)

Routine destruction under a company policy can constitute spoliation. Youngevity, Int’l, et al. v. Smith, No. 3:16-cv-704-BRM-JLB, 2020 WL 7048687 at *2 (S.D. Ca. July 28, 2020) (“Defendants’ failure to prevent destruction by backing up their phones’ contents or disabling automatic deletion functions was not reasonable because they had control over their text messages and should have taken affirmative steps to prevent their destruction when they became aware of their potential relevance.”). Once litigation is reasonably anticipated, parties must suspend policies related to deleting or destroying files and preserve relevant ESI. See e.g., Browder v. City of Albuquerque, 187 F.Supp.3d 1288, 1294-95 (D.N.M. 2016) (noting “a continuing responsibility to ensure” the preservation of “relevant information” and that this requires more than “mere notification;” counsel must take “affirmative steps to monitor compliance, to talk to key employees in an effort to understand how evidence will be stored, to continually ensure that the party is preserving relevant evidence.”); Cache La Poudre Feeds, LLC v. Land O’Lakes, Inc., 244 F.R.D. 614, 635-36 (D. Colo. 2007) (in a case decided before Rule 37(e)’s amendment, finding that the failure to preserve potentially relevant and responsive ESI by wiping clean computer hard drives and counsel’s failure to monitor the ESI discovery process properly warranted monetary sanction).

Respondents did audits and camera reviews throughout the inspection period. (Tr. 883-84; Exs. C-48, C-75, RCS-41; CS Br. 40, 56, 71.) They have the sophistication to understand the obligation to retain evidence before litigation formally commences. UHS-DE has been involved in several other proceedings before the Commission, during which the Secretary sought and used video of workplace violence incidents to establish general duty clause violations for the hazard of workplace violence.³⁵

The Advisory Committee notes to Rule 37 explain that ESI can be restored or replaced if the ESI is available from a third party or if there is more than one electronic copy. See Borum v. Brentwood Vill., LLC, 332 F.R.D. 38, 46 (D.D.C. 2019). That is not available to the Secretary for this matter.

Respondents note the availability of other evidence of workplace violence events was provided during discovery and introduced at the hearing.³⁶ Written records and availability of certain witnesses are not equivalent to what was destroyed. For example, the log of HPR reports indicates that video of multiple incidents was reviewed. (Ex. C-20.) However, there is little documentation about what the reviews uncovered. (Tr. 1906-9; Ex. C-20.) The videos may have captured details not found in the written records or the witnesses’ memories. See Woodward v. Wal-Mart Stores E., LP, 801 F. Supp. 2d 1363, 1373 (M.D. Ga. 2011) (finding that employee testimony about an event “hardly works” to address the loss of video); Storey, 2017 WL 2623775, at *5 (discussing the “unique and irreplaceable nature” of video evidence); BankDirect Cap. Fin., LLC v. Cap. Premium Fin., Inc., No. 15 C 10340, 2018 WL 1616725, at *10 (N.D. Ill. Apr. 4, 2018) (“BankDirect”) (proffered substitute for destroyed ESI was less compelling and not the “evidentiary equivalent” of what was destroyed); Moody v. CSX Transp., Inc., 271 F. Supp. 3d 410, 429-30 (W.D.N.Y. 2017); Abdulahi v. Wal-Mart Stores E., L.P., 76 F. Supp. 3d 1393, 1396-97 (N.D. Ga. 2014) (other evidence such as emails and testimony did not remove prejudice caused by the destruction of video).

Rule 37(e)(1) allows for curative measures when lost ESI causes prejudice to another party. “An evaluation of prejudice from the loss of information necessarily includes an evaluation of the information's importance in the litigation.” Bush v. Bowling, No. 19-CV-000098-GFK-FHM, 2020 WL 5423986, at *6 (N.D. Okla. Sept. 10, 2020) (finding prejudice from the loss of video despite its somewhat limited probative value and the availability of other evidence). In UHS Centennial, the employer’s destruction of videos warranted sanctions. 2022 WL 4075583, at *3. The judge imposed adverse inferences that the destroyed video of workplace violence incidents at a behavioral health facility would have shown: (1) the hazard of workplace violence was present in the workplace; (2) the employer recognized the presence of the workplace violence hazard; (3) the employer had actual knowledge of the workplace violence hazard and was aware its employees were exposed to it; and (4) such exposure could result in seriously bodily harm or death. Id. at *3, 23.

Respondents stipulated to some of the same findings imposed as adverse inferences in UHS Centennial. Those stipulations mitigated some of the prejudice the ESI’s destruction caused. Still, the lack of certain direct evidence complicated the Secretary’s presentation of this matter. The destroyed videos relate to disputed elements of the Secretary’s burden, including: (1) the harm from the hazard, (2) knowledge of the hazard’s presence, (3) the effectiveness of the Worksite’s abatement, and (4) the proposed abatement. To illustrate, Respondents challenge the contention that the calling of codes did not adequately provide enough staffing for workplace violence incidents. Citing a manager’s testimony, Respondents claim that typical response times are measured in seconds. (CS Br. 38.) In contrast, an employee described the response time to an incident that led to her injury as “several minutes.” (Tr. 753, 757.) Other witnesses contradict the claim that a response time was typically 15-20 seconds. (Tr. 807, 2309.) The preserved videos show when more staff arrived but represent only a fraction of the hazard occurrences. (Exs. C-106, C-108; Tr. 1553, 1714, 2309.) Ms. Nolan indicated she responded to every Code Green on her shifts and would perform debriefings. (Tr. 2343-44.) Video of her responding to workplace violence incidents would have supported or refuted her claims about the routine availability of managers to assist with codes.³⁷ Respondents also questioned the CO’s testimony that employees had to call multiple Code Greens before assistance arrived. (Tr. 303-4.) Again, video from more incidents could have corroborated or refuted this testimony. (Ex. C-106; Tr. 1549.)

Other questions on cross-examination highlighted some gaps in memories since the events occurred. (Tr. 717-18, 720, 1531-34, 1537-38.) A former nurse contacted the reporting line after a patient hit her with a laundry bin on February 12, 2020, months after OSHA’s investigation commenced. (Tr. 1531-34, 1537-38, 1540; Ex. C-15.) The witness believed the incident should have been captured on video. (Tr. 1541.) But, at that time, an opportunity to review camera footage had never been made available to injured workers. (Tr. 1541-42.) The witness had difficulty recalling the exact number of responders and whether the incident was caused by patient aggression or the patient’s refusal to take medication to reduce agitation, aggression, and anxiety. (Tr. 1537-40.) Although Respondents note the witness’ failure to remember this detail, they do not explain the difference to a worker, if any, between being hit for no reason versus a patient hitting a worker because the patient did not want medication. (CS Br. 62-63.) Respondents now seize on such gaps in memory to attack the former workers’ credibility. Id. at 47-65. Videos could have addressed such gaps and helped with credibility assessments.

Respondents recognized the utility of video and made extensive use of video they selectively preserved. (Tr. 1542-60, 1551-52, 2347-62; CS Br. 64-65, 91, 107-8.) They argue about what could have occurred off-camera without confirming that they preserved all available angles of incidents of workplace violence. (Tr. 1552, 1557-58, 2764.) Respondents had the opportunity to review evidence that may have corroborated or refuted testimony but did not give the Secretary the same opportunity.

Respondents also argue the availability of documents precludes a finding of prejudice. (Resp. to Sec’y Sanctions Mot. at 2, 13-15; CS Br. 13 n.7.) In fact, the documents highlight what was lost. For example, Respondents produced records related to various restraints of Patient 23 between January 6 and 7, 2020. (Ex. C-64; Stip. 55.) The incidents occurred well after November 2019, met the criteria for when Respondents ordinarily reviewed video, and happened in areas with cameras. Yet, Respondents preserved video of only one incident on January 7, 2020. They did not provide video of the other incidents that occurred on January 6th and at other times on January 7th. (Ex. C-108; Sec’y Br. 131.) Respondents do not explain why video of one incident was available yet others occurring so close in time and in similar areas were not. The written patient debriefing does not answer whether workers were adequately protected or the proposed abatement’s effectiveness. (Sec’y Br. 130-32.) The destroyed ESI could have shown how quickly staff responded, whether the latch on the window was broken, whether the patient was threatening before he entered the nurses’ station, and if the MHS could have called for help sooner. (Tr. 717-18, 720.)

Other examples of the disconnect between documents and recollections are seen in the Sedgwick Reports compiled by a contractor after injured workers call a hotline. (Stip. 39; Tr. 116-17, 2728-29; Exs. RCS-23, C-94 at 15.) Often, those reports give conflicting information about the injury’s cause. Rather than stating that an injury was caused only by a patient or patients, often, the Sedgwick Reports incorrectly state that a “fellow worker” caused the injury when a patient caused it. (Exs. C-3 at 53, C-12, C-14.) Other times, the reports simply state that a “human” caused the injury, without further clarification. Videos might have answered questions unaddressed in the written records and assisted with weighing credibility.

Most of the videos produced related to events in December 2019 and some did not show an aggressive patient.³⁸ Video was provided for only four incidents in all of 2020. And none of that related to the incidents on January 3, January 6, or February 12, 2020. (Tr. 672-73, 682, 1474, 1476; Exs. C-20, C-66.) Respondents attack the Secretary for not introducing every clip preserved. (CS Br. 13 n. 7.) While only three clips were introduced into evidence, the Secretary’s experts reviewed and relied on all videos produced in their reports. (Exs. C-94 at 22, 30; C-92 at 10.)

The point of precluding spoliation of evidence is that one side does not unilaterally get to decide what to preserve. When one side (a) has possession, custody, or control of material videos, (b) has the opportunity to review the video, (c) can preserve video, and (d) only preserves some of the videos, it deprives the other side of viewing all relevant evidence. Nation-Wide Check Corp., Inc. v. Forest Hills Distrib., 692 F.2d 214, 218 (1st Cir. 1982) (Breyer, J., drawing inferences from the destruction of documents under common law precedents). Discovery permits opposing sides to see the evidence and determine its worth. One side cannot destroy relevant evidence and then claim it would not have benefitted its opponent.

Certainly, the lost ESI might not have been helpful to the Secretary. However, it is extremely difficult to support speculation as to its lack of utility when Respondents failed to preserve what would be the best evidence of this conclusion. Paisley Park Enters. v. Boxhill, et al., 330 F.R.D. 226, 235-36 (D. Minn. 2019) (“Even when the information lost is cumulative to some extent, the loss of the information still has an impact because [a party] cannot present the overwhelming quantity of evidence they otherwise would have to support their case”); Ala. Aircraft Indus., Inc. v. Boeing, Co., No. 20-11141, 2022 WL 433457, at *16 (11th Cir. Feb. 14, 2022) (party could not substantiate the argument that the lost ESI was relatively unimportant when they deleted the information that would support such a conclusion). See also Food Lion, Inc. v. United Food and Com. Workers Int’l Union, 103 F.3d 1007, 1012 (D.C. Cir. 1997) (consolidated) (duty to preserve is broader than the duty to produce something in discovery).

The ESI destruction prejudiced the Secretary by forcing the Secretary to rely on less compelling evidence and piece events together from multiple sources. (Sec’y Br. 130-32.) See BankDirect, 2016 WL 65869448, *10 (availability of some ESI did not remove the prejudice caused by the destruction of “unguarded emails” the party “wrongfully place[d] out of reach”); In re: Gold King Mine Release in San Juan Cnty., Colo., No. 1:18-md-02824-WJ, 2022 WL 2230759, at *6-7 (D.N.M. June 21, 2022) (consolidated) (finding prejudice when spoliated ESI was relevant to a critical issue about which evidence was conflicting); In re: Ethicon, Inc., No. 2:12-cv-00497, 2016 WL 5869448, at *4 (S.D.W.Va. Oct. 6, 2016) (prejudice under Rule 37(e) may be found when the destruction causes a party to “piece together information from other sources”).

Sanctions available under Rule 37(e)(1) cannot be greater than necessary to cure the prejudice. See Kindergartners Count, Inc. v. Demoulin, et al., 209 F.R.D. 466, 468-69 (D. Kan. 2002) (appropriate sanction was to deem defendant’s defamatory conduct established). Respondents’ disregard for preserving evidence is troubling. Even so, the destroyed ESI was not so essential as to present an insurmountable obstacle to Complainant’s ability to prove its case. Further, the prejudice from the loss was mitigated by Respondents agreeing to several stipulations. Respondents stipulated that they and their industry recognize the hazard of workplace violence. (Stips. 9-10.) They stipulated that the hazard is present at the Worksite, employees were exposed to it, and it could result in serious staff injuries. (Stips. 7-8, 11.)

Nonetheless, these stipulations are insufficient to cure the prejudice. The Secretary is entitled to additional curative sanctions. The destroyed ESI will be presumed to have supported finding knowledge of the hazard of workplace violence at the Worksite and that the hazard caused actual serious harm.

The Secretary’s Motion and subsequent post-hearing brief focus on sanctions that can only be imposed after a finding of an intent to deprive. In support, the Secretary cites the notice of the need to preserve video, Respondents’ capability of doing so, and their lack of any reasonable explanation for why they failed to do so. (Sec’y Br. 131-34.) Rule 37(e) “leaves judges with discretion to determine how best to assess prejudice in particular cases.” Fed. R. Civ. Pro. 37 advisory committee note to the 2015 amendment, Subdivision (e)(1).

Respondents knew the utility of reviewing videos to assess safety hazards. Indeed, they claimed it was part of their WVPP. (Ex. C-22.) Video was routinely reviewed for a variety of reasons, such as patient restraints. (Tr. 698, 883-84, 910-16, 1476-77, 1832, 1834-35, 1913-14, 2762-66, 2771; Exs. C-14 at 3; C-19, C-20, C-22, C-39, C-48, C-75; RCS-41 at 4, 6.)

Intent is rarely proved by direct evidence. See, e.g., BankDirect, 2018 WL 1616725, at *2 (“[A] combination of events, each of which seems mundane when viewed in isolation, may present a very different picture when considered together.”); Moody, 271 F. Supp. 3d at 431-32 (finding intent based on defendants’ actions that allowed evidence to be overwritten and destroyed); Ottoson v. SMBC Leasing & Fin., Inc., 268 F. Supp. 3d 570, 581-82 (S.D.N.Y. 2017) (considering plaintiff’s conduct throughout the litigation and during discovery disputes); Ala. Aircraft Indus., Inc., 319 F.R.D. 730, 746-47 (N.D. Ala. 2017) (imposing sanctions under Rule 37(e)(2) where “unexplained, blatantly irresponsible behavior” led to ESI destruction), aff’d, No. 20-11141, 2022 WL 433457, at *13-16 (11th Cir. Feb. 14, 2022).

There is sufficient circumstantial evidence here to infer an intent to deprive the Secretary of the best evidence of the hazard and how Respondents’ abatement program addressed the hazard. Respondents knew OSHA was investigating workplace violence, they knew they had video evidence of the hazard, they knew the Secretary sought this ESI, and they knew they were contesting the allegations in the Citation. They offer no sound explanation for why companies with access to counsel and risk managers failed to preserve relevant information after OSHA’s investigation commenced, after a subpoena, and after litigation commenced. See Fed. Trade Comm’n v. Noland, No. CV-20-00047-PHX-DWL, 2021 WL 3857413, at *14 (D. Ariz. Aug. 30, 2021) (imposing sanctions under Rule 37(e) when a party destroyed electronic evidence after learning it was under investigation by a federal agency); BankDirect, 2018 WL 1616725, at *5-7, 9 (rejecting party’s array of excuses for destroying ESI when litigation was reasonably anticipated).

To impose sanctions under Rule 37(e)(2), a finding of prejudice is not strictly necessary. In re Google, 2023 WL 2673109 at *9. The finding of intent under subdivision (e)(2) supports “not only an inference that the lost information was unfavorable to the party that intentionally destroyed it, but also an inference that the opposing party was prejudiced by the loss of information that would have favored its position.” Comm. Notes, Subdivision (e)(2). Although such a finding is not required, as addressed, the Secretary was prejudiced by Respondents’ failure to preserve relevant, discoverable evidence of instances of the hazard and their response.

Respondents had the requisite state of mind for Rule 37(e)(2)’s remedies. Still, the harshest sanctions are not appropriate here. The “remedy should fit the wrong.” Advisory committee’s notes to 2015 amendment. The request for a definitive adverse inference establishing that the Secretary met the abatement prong of the general duty clause test is rejected. (Sec’y Br. 134.) Instead, the undersigned finds that the destroyed videos would have been unfavorable to Respondents on the issues of knowledge, abatement, and gravity. Further, the destroyed videos would have supported the Secretary’s claims regarding knowledge, abatement, and gravity. See Envision Waste Servs., LLC, No. 12-1600, 2018 WL 1735661, at *10 (OSHRC Apr. 4, 2018) (concluding that the failure to produce authentic documents corroborated evidence of employer's non-compliance).

III.OSHA Has Authority Over the Worksite and Cited Hazard

Before turning to the merits, Respondents’ challenge to OSHA’s scope of authority will be considered. Respondents argue that workplace violence is not within the scope of the general duty clause or OSHA’s jurisdiction.³⁹ The Secretary maintains that although Respondents undoubtedly have an obligation to patient care, they also have an obligation to protect employees from hazards reasonably likely to cause serious injury or death. See Safeway, Inc. v. OSHRC, 382 F.3d 1189, 1193-94 (10th Cir. 2004) (worker’s presence at the facility following his superior’s instructions was sufficient to support the OSH Act’s applicability).

Workplace violence is a hazard within the meaning of the OSH Act. It has long been recognized in the healthcare industry, and “according to U.S. Bureau of Labor Statistics data, the incidence of violence-related health care worker injuries has steadily increased for at least a decade.” (Stips. 10, 54; Ex. C-94 at 6-10.) Despite their rhetoric, Respondents provide no support for finding that OSHA lacks jurisdiction over, or responsibility for, workplace safety at inpatient psychiatric hospitals or other similar places of employment. Each Respondent stipulated that the Commission “has jurisdiction” over this matter, and each acknowledged they are an employer as defined in the OSH Act.⁴⁰ At best, Respondents are arguing against permitting OSHA to cite medical facilities for hazards that cannot be eliminated.

The OSH Act does not exempt businesses providing medical care. Applying the general duty clause to such employers, including for the hazard of workplace violence, has been repeatedly upheld. BHC Nw. Psychiatric Hosp. v. Sec’y of Labor, 951 F.3d 558, 566 (D.C. Cir. 2020) (“BHC”) (upholding citation for violation of the OSH Act based on the hazard of workplace violence at an inpatient psychiatric facility); UHS of Westwood Pembroke, Inc., UHS of De., No. 17-0737, 2022 WL 774272, at *1, 7-12 (OSHRC Mar. 3, 2022) (“UHS-DE/Pembroke”) (upholding a violation of the general duty clause for exposing employees to the hazard of patient on staff aggression at a behavioral health facility), aff’d, No. 22-1845, 2023 WL 243988 (3d Cir. May 4, 2023); UHS Centennial, 2022 WL 4075583, at *2-3 (same); Integra Health Mgmt., Inc., No. 13-1124, 2019 WL 1142920 at *1, 6 n.5 (OSHRC Mar. 4, 2019) (upholding the applicability of the general duty clause to workplace violence experienced by employees working with individuals with mental illness or criminal backgrounds). In Integra, the Commissioners unanimously upheld the applicability of the general duty clause to workplace violence experienced by employees working with individuals with mental illness or criminal backgrounds. 2019 WL 1142920, at *1, 6 n.5. The employer violated the general duty clause by failing to “adequately address a workplace violence hazard-specifically, the risk of Integra’s employees being physically assaulted by a client with a history of violent behavior during a face-to-face meeting.”⁴¹

Respondents knew the hazard was present at the Worksite and that they could take steps to reduce it.⁴² They acknowledge the Worksite had policies and procedures which, they argue, aimed to mitigate the serious and “longstanding” hazard of workplace violence. (CS Br. 7, 21.) Respondents’ challenge to the scope of OSHA’s jurisdiction is rejected.

Respondents next contend that the oversight of other regulators and an industry credential reduces the Secretary’s authority over the Worksite. (CS Br. 20-21, 66.) Like other types of employers, Respondents must comply with laws other than the OSH Act. “Cedar Springs Hospital is certified as a Medicare and Medicaid hospital by the Centers for Medicare and Medicaid Services (“CMS”).” (Stip. 12.) To maintain this certification, the Worksite must comply with federal regulations concerning “quality standards in hospitals.”⁴³ The Worksite is “regulated by Colorado state agencies,” including the CDPHE.⁴⁴ And it is “accredited by The Joint Commission (TJC) per its standards.” (Stip. 14.)

Even assuming that the Worksite was in complete compliance with all other regulatory requirements and TJC standards, it does not mean Respondents did enough to protect workers from workplace violence. While the TJC or other regulators may not require the proposed abatement measures, there is no evidence that Respondents could not adopt them because of other regulatory requirements. See UHS Centennial, 2022 WL 4075583, at *30-32 (noting that the TJC’s Joint Sentinel Alert is not mandatory but does identify industry-recognized methods for addressing the complex hazard of workplace violence); Int’l Union, United Auto., Aerospace & Agr. Implement Workers of Am., Gen. Dynamics Land Sys., Div. v. Sec’y of Labor, 815 F.2d 1570, 1576 (D.C. Cir. 1987) (adherence to specific standards did not absolve employers from general duty to provide a safe place of employment).

Other workplace violence cases have involved situations where the entity must comply with regulations besides the OSH Act and for which an industry credential is important. In each instance, the existence of other regulators and/or the need to maintain a credential did not undermine the Secretary’s authority to cite the employer for the hazard of workplace violence. See UHS-DE/Pembroke, 2022 WL 774272, at *4 (noting the CEO’s responsibility to meet TJC and state standards); HRI, 2019 WL 989735, at *21 (discussing state requirements); UHS Centennial, 2022 WL 4075583, at *23 n.27 (rejecting argument that TJC or Colorado law precluded proposed abatement measures). See also Integra, 2019 WL 1142920, at *7 (rejecting the argument that public policy concerns related to serving people with histories of violent behavior precluded a general duty clause citation); Waldon, 16 BNA OSHC at 1058 (finding abatement feasible even though the nursing home was in a “highly regulated business”).

Rather than distinguish those cases, which involve similar businesses and hazards, Respondents point to the dissent in SeaWorld of Fla. LLC v. Perez, 748 F.3d 1202 (D.C. Cir. 2014). In SeaWorld, the amusement park violated the general duty clause by exposing trainers to recognized hazards when working closely with animals during performances. 748 F.3d at 1205. The majority rejected the dissent’s arguments about policy decisions, stating, “Congress has vested in the Secretary and the Commission general authority to protect employees from unhealthy and unsafe work places.” Id. at 1211. Although the SeaWorld dissent takes issue with OSHA overseeing employee safety in the entertainment and sports industries, it does not state or imply that OSHA lacks jurisdiction over workplace violence in general or psychiatric care providers. Id. at 1216-22.

Respondents identify no other regulator or credentialing authority charged with protecting workers as opposed to those focused primarily on the health and safety of the consumers of the Worksite’s services. See Shamokin Filler Co., Inc. v. Fed. Mine Safety & Health Review Comm’n, 772 F.3d 330, 332-33 (3d Cir. 2014) (OSHA is the default agency for worker safety and health). It does not appear that CDPHE, TJC, or any regulatory authority besides OSHA assessed the Worksite during the pendency of OSHA’s investigation. Moreover, there is no evidence that TJC, or any other regulatory authority besides OSHA, concluded that the Respondents’ handling of workplace violence was appropriate or that there were no actions Respondents could take to materially reduce employee exposure to workplace violence. Respondents cite no portion of the OSH Act, the implementing regulations, or precedent to support their contentions that an exemption for their industry is directly or implicitly in the OSH Act. Likewise, complying with other applicable regulations and requirements neither conflicts nor precludes complying with the OSH Act.

Respondents assert that the Citation’s issuance for the hazard of workplace violence at a behavioral health facility violated the Due Process Clause of the Fourteenth Amendment. (CS Br. 66, 129.) Similar challenges to applying the general duty clause to workplace violence hazards at behavioral health facilities were rejected. BHC, 951 F.3d at 566; UHS-DE/Pembroke, 2022 WL 774272, at *18-21, 41-42 (ALJ collecting cases and rejecting employer’s due process claims related to the Citation and proposed abatement); UHS-DE/ Suncoast, 2023 WL 2388069, at *107-10 (ALJ concluding that general duty clause was not unconstitutionally vague as applied to workplace violence hazard at a behavioral health facility). See also Integra, 2019 WL 1142920 at * 14, n.15. Rejection of the notice claims is also appropriate here. See Brennan v. OSHRC (Santa Fe Trail Transport Co.), 505 F.2d 869, 872-73 (10th Cir. 1974) (distinguishing the Due Process requirements for criminal matters and rejecting claim that a standard was unconstitutionally vague).

Respondents knew workplace violence was a hazard to which their employees were exposed and from which they suffered injuries at this Worksite.⁴⁵ They, and their industry, also knew of OSHA’s concern about workplace violence overall and the hazard’s presence in behavioral health facilities. See UHS-DE/Pembroke, 2022 WL 774272, at *19 (ALJ found both actual knowledge and that the behavioral health industry was on notice of the hazard); UHS-DE/Suncoast, 2023 WL 2388069, at *2. Further, there is no dispute that experts familiar with the industry would consider the hazard when prescribing a safety program. See Nat’l Realty, 489 F.2d at 1266.

Knowledge of the hazard provides adequate notice to satisfy the requirement of Due Process. See, e.g., Cape & Vineyard Div. of New Bedford Gas & Edison Light v. OSHRC, 512 F.2d 1148, 1152(1st Cir. 1975) (finding that actual knowledge of the hazard provides fair notice); Bethlehem Steel Corp. v. OSHRC, 607 F.2d 871, 875 (3d Cir. 1979) (finding that fair notice is addressed by the requirement that the hazard is recognized). Like the present matter, in BHC, the proposed abatement accorded with “well-known industry best practices” and, in many instances, aligned with the safety measures the respondents identified but failed to apply consistently. 951 F.3d at 566. See also SeaWorld, 748 F.3d at 1216 (finding that employer could have anticipated that abatement measures it applied after incidents would be required); Babcock & Wilcox, Co. v. Sec’y of Labor, 622 F.2d 1160, 1164-65 (3d Cir. 1980) (affirming finding of liability when the company failed to take feasible precautions to reduce the injury risk); A.C. Castle Constr. Co. v. Sec’y of Labor, 882 F.3d 34, 38, 43-44 (1st Cir. 2018) (rejecting fair notice claims and limiting the doctrine’s scope). Despite claims to the contrary, the abatement practices used at the Worksite did not align with OSHA’s “Guidelines for Preventing Workplace Violence for Healthcare and Social Service Workers” (“Guidelines”) or the methods identified in Respondents’ unfollowed policies.

Management was aware of the obligation to protect workers from workplace violence and informed on how to do so. See Integra, 2019 WL 1142920 at *14, n.15 (rejecting constitutional vagueness challenge because the proposed abatement measures were “available to, and readily knowable by, the industry”); UHS-DE/Suncoast, 2023 WL 2388069, at *2 (UHS-DE and the other party acknowledged that they recognized the hazard of workplace violence, knew it was present at the behavioral health facility, and knew that measures taken were not effective in materially reducing the hazard); UHS-DE/Pembroke, 2022 WL 774272, at *7-12 (employer knew the hazard of workplace violence was present at their behavioral health facility and failed to abate it sufficiently). As in BHC, Respondents “can hardly object” that they were “blindsided by the utility of measures” they “already embraced, at least on paper.” 951 F.3d at 566. See also CF&T Available Concrete Pumping, Inc., 15 BNA OSHC 2195, 2198, n.9 (No. 90-329, 1993) (noting that the “mere existence of a safety program on paper does not establish that the program was effectively implemented on the worksite”); Pepperidge Farm, 17 BNA OSHC 1993, 2007-8 (No. 89-265, 1997) (employer failed to implement abatement it identified). Respondents’ Due Process challenges to citations for the hazard of workplace violence are rejected.

Respondents then attempt to turn their notice argument on its head. They acknowledge that the Guidelines and OSHA’s “Preventing Workplace Violence: A Road Map for Healthcare Facilities” (“OSHA Road Map”) inform employers about how to assess their workplaces and develop appropriate ways to minimize the hazard of workplace violence. (CS Br. 120-23; Exs. C-95, C-96.) The Guidelines explain the hazard’s risk, discuss appropriate practices to minimize it, and explain how to get further assistance. (Ex. C-95.) Rather than acknowledging that these documents support OSHA’s authority over the hazard, they ask for them to be viewed as a basis for vacating the Citation. (CS Br. 125.)

The Guidelines and OSHA Road Map specify employers’ responsibilities under the OSH Act, including the right for all workers to have “working conditions that do not pose a risk of serious harm.” (Exs. C-95 at 52, C-96 at 5.) This is not a new right. It is directly from the OSH Act. With or without the documents, employers must ensure their workplaces are “free from recognized hazards that are causing or are likely to cause death or serious physical harm.” 29 U.S.C. § 654(a)(1). (Exs. C-95 at 3-46, 52-56, C-96 at 5.) The Guidelines and OSHA Road Map are not equivalent to OSHA standards for specific hazards. The documents acknowledge this. (Exs. C-95 at 2, 4; C-96 at 2, 3-4.) The hazard of workplace violence is complex and varies by workplace. (Exs. C-95 at 7-8, C-96.) The documents guide employers in how to develop effective abatement for their particular workplace. They are tools to assist employers in meeting their OSH Act obligations. Employers are responsible for assessing the risk and developing appropriate responses. (Ex. C-95 at 3-46.)

The Secretary is not impermissibly enforcing the Guidelines or OSHA Road Map. She is attempting to ensure compliance with section 5(a)(1) of the OSH Act. Such compliance is required regardless of whether OSHA created guidance materials or offered to provide other assistance. See UHS Centennial, 2022 WL 4075583, at *30 (although the Guidelines and OSHA Road Map do not create new legal obligations, employers are “bound by the general duty clause and by its reasonable and prudent employer standard”).

IV.Expert Testimony

Expert witnesses testified for both sides. Jane Lipscomb, Ph.D., and Dr. Forman, testified on the Secretary’s behalf, and Vincent Staggs, Ph.D., testified for Respondents. Each expert’s written report was admitted to the record. Each person satisfied the threshold requirements to be qualified to offer expert testimony, but their respective areas of expertise differed. See F&H Coatings, LLC v. Acosta, 900 F.3d 1214, 1222-23 (10th Cir. 2018) (upholding the ALJ's finding that the expert was appropriately qualified to offer expert testimony). For the reasons discussed below, the opinions of Doctors Forman and Lipscomb were more relevant and helpful in resolving this matter.

Dr. Forman was accepted as an expert in the evaluation and clinical management of psychiatric patients. (Tr. 1183.) He has over fourteen years of experience working as a medical doctor in psychiatric settings. (Ex. C-91.) He is board-certified in general psychiatry, forensic psychiatry, and addiction medicine. (Tr. 1152-53, 1168-69; Ex. C-91.) He has extensive experience treating violent and aggressive patients while they are in psychiatric care. (Tr. 1162-64; Ex. C-91.) Dr. Forman testified before the Commission in UHS-DE/Suncoast. In that matter, he was qualified as an expert to: (1) testify regarding the causes and prevention of patient-on-staff violence in a psychiatric hospital setting, (2) render opinions regarding feasible means of abatement on the issue of workplace violence, and (3) testify whether the employers’ workplace violence policies and procedures were adequate to address patient-to-staff aggression, including what clinical treatment is appropriate to address aggression and whether such clinical treatment was provided. 2023 WL 2388069, at *47.

In this matter, Dr. Forman assessed: (1) whether the Worksite was adequately addressing the hazard of workplace violence during the time of OSHA’s inspection, (2) whether the abatement measures proposed are feasible, and (3) whether the measures can be expected to reduce workplace violence materially. (Ex. C-92.) He reviewed and offered opinions on Respondents’ abatement and the additional measures proposed. Prior to reaching his opinions, Dr. Forman reviewed a large volume of material, including deposition transcripts, patient records, site visit reports, staff surveys, staffing reports, photographs, employee interview transcripts, policies and procedures, and available videos. (Tr. 1175-76; Ex. C-92.)

Dr. Lipscomb has worked in workplace violence prevention in healthcare for over 30 years. (Tr. 1587; Ex. C-93.) She has researched the topic extensively, published related papers, and consulted with numerous governmental and private entities about the cited hazard. (Ex. C-93.) Like Dr. Forman, Dr. Lipscomb previously testified as an expert in Commission proceedings. She was accepted as an expert in UHS-DE/Suncoast, 2023 WL 2388069, at *45-48, BHC, 2019 WL 989734, at *4-6, and HRI, 2019 WL 989735, at * 5. See also BHC, 951 F.3d at 561, 564.

Dr. Lipscomb worked for OSHA and the National Institute for Occupational Safety and Health (NIOSH) to identify practices to prevent workplace violence in healthcare facilities. (Tr. 1589-94, 1597; Ex. C-94 at 2-3.) Her work included fifteen years of field research on workplace violence in four states. (Tr. 1587; Ex. C-94 at 3.) She designed training programs for healthcare workers on how to mitigate workplace violence. (Tr. 1584-85, 1593-94; Exs. C-93, C-94 at 3.) She has many publications, including over twenty peer-reviewed publications about workplace violence. (Tr. 1587-89; Ex. C-94 at 3.) She frequently lectures about workplace violence in healthcare. (Tr. 1592; Exs. C-93, C-94 at 3.) She has been trained in research methodologies and has taught epidemiology for several years. (Tr. 1579, 1582.)

Dr. Lipscomb has long studied the cited hazard and how to implement effective abatement methods. (Tr. 1587; Exs. C-93, C-94.) Most of her work and related publications center on the same type of workplace violence at issue here: patients acting aggressively against healthcare workers. (Exs. C-93, C-94.) See also BHC, 2019 WL 989734, at *5, 7-8 (discussing Dr. Lipscomb’s experience and crediting her findings regarding an employer’s failure to abate the hazard of workplace violence and the existence of feasible methods of abatement); UHS Centennial, 2022 WL 4075583, at *28 (noting the depth of Dr. Lipscomb’s experience with the hazard of workplace violence in healthcare and relying heavily on her testimony and report); UHS-DE/Suncoast, 2023 WL 2388069, at *45, 47-48 (discussing Dr. Lipscomb’s experience and research on workplace violence).

At the hearing, Dr. Lipscomb was accepted as an expert in workplace violence in healthcare and behavioral health settings, including issues regarding the adequacy of workplace violence prevention programs and other methods for preventing workplace violence. (Tr. 1598, 1602-3.) Her report and testimony focused on the adequacy of Respondents’ approach to the hazard and whether they could take additional feasible steps to materially reduce the hazard of patient-on-staff violence at their Worksite. (Tr. 1603-88; Ex. C-94.)

Dr. Staggs’ background is in quantitative psychology. (Tr. 2094; Ex. RCS-94.) He is not a licensed psychologist and has not completed any specialized coursework on workplace violence. (Tr. 2122, 2124-25, 2139-40; Ex. RCS-94.) He has never worked in or directed an inpatient psychiatric care unit. (Tr. 2124-25; 2130; Ex. RCS-94.) He has never treated a patient, been responsible for admission decisions, responded to a psychiatric emergency, or restrained a patient. (Tr. 2124-26; 2130; Ex. RCS-94.)

Dr. Staggs reviewed 100-500 pages of documents and looked at the transcripts from other Commission proceedings in which UHS-DE affiliated entitles were parties. (Tr. 2129, 2293.) He also discussed the matter with four unaffiliated people with the Worksite (his sister, his mother-in-law, a friend’s husband, and a nurse in Florida). (Tr. 2127-28.) He did not speak with any of Respondents’ employees. (Tr. 2126-29.) He reviewed the narrative portion of the CO’s report but did not read the worker interview statements in full. (Tr. 2128-29, 2166.) He did not visit the Worksite, watch any of the available videos, or review any layouts or floorplans for the Worksite. (Tr. 2126-29.) Dr. Staggs was unfamiliar with how patients were observed and the Worksite’s assault rate.⁴⁶

Dr. Staggs offered no opinion on the efficacy of Respondents’ existing abatement. (Tr. 2091.) Nor did he offer an opinion on each aspect of the Secretary’s proposed abatement. (Tr. 2091.) Instead, assessed four aspects of the Secretary’s proposal: (1) adequate staffing, (2) the use of security, (3) reconfiguring the nurses’ stations, and (4) the use of reliable communication devices. Id.

Of these areas, Dr. Staggs was accepted as an expert in staffing issues.⁴⁷ He was not in a position to assess the feasibility and effectiveness of the Secretary’s other proposed abatement methods. (Tr. 2141, 2147-48, 2154, 2267.) He lacked specialized knowledge, training, or direct experience working with a dedicated response staff, designing or working in nurses’ stations, or using a radio or panic alarm in a healthcare setting. (Tr. 2130-34, 2137-38, 2153.) And he had not published any articles or studies in those areas. (Tr. 2019-20.) Accordingly, he was not qualified as an expert on security response teams, nurses’ stations, or communication devices. (Tr. 2148-55, 2217-18.)

Although each person satisfied the threshold requirements to be qualified to offer expert testimony, their respective opinions are not entitled to equal weight. See i4i Ltd. P’ship v. Microsoft Corp., 598 F.3d 831, 852 (Fed. Cir. 2010) (“When the methodology is sound, and the evidence relied upon sufficiently related to the case at hand, disputes about the degree of relevancy or accuracy (above this minimum threshold) may go to the testimony’s weight, but not its admissibility”), aff’d, 564 U.S. 91 (2011). The testimony of Doctors Forman and Lipscomb focused more directly on the Worksite and the feasibility and effectiveness of the proposed abatement. See BHC, 951 F.3d at 561, 564 (upholding ALJ’s crediting Dr. Lipscomb’s testimony over another expert’s). They assessed both the existing abatement and the Secretary’s proposed abatement. Dr. Staggs’ review was more limited. He reviewed less material, particularly as related to conditions at this Worksite. The scope of his opinions was more circumscribed because of his experience and the extent of his review.

V.Analysis

The general duty clause requires every employer to provide employees with a workplace “free from recognized hazards that are causing or are likely to cause death or serious physical harm.” 29 U.S.C. § 654(a)(1). F&H, 900 F.3d at 1224 (affirming citation for violating the general duty clause). As interpreted by the Commission, to establish a violation of this clause, the Secretary must show: (1) there was an activity or condition in the employer’s workplace that constituted a hazard to employees; (2) either the cited employer or its industry recognized that the condition or activity was hazardous; (3) the hazard was causing or was likely to cause death or serious physical harm; and (4) there were feasible means to eliminate the hazard or materially reduce it. 900 F.2d at 1224; Safeway, Inc. v. OSHRC, 382 F.3d 1189, 1194-95 (10th Cir. 2004) (affirming a violation of the general duty clause when the employer was aware of the hazard and the employer could have eliminated it); Waldon, 16 BNA OSHC at 1058. The evidence must also show that the employer knew or, with reasonable diligence, could have known of the hazardous condition. Otis Elevator Co., 21 BNA OSHC 2204, 2207 (No. 03-1344, 2007).

“There is no existing Section 5(a)(2) OSHA standard for preventing workplace violence in healthcare.” (Stip. 53.) However, the general duty clause has repeatedly been found applicable to workplace violence hazards. BHC, 951 F.3d at 563-67 (upholding a citation for a general duty clause violation for a workplace violence hazard at a psychiatric hospital); Integra, 2019 WL 1142920, at *4 (finding “allegation of workplace violence … is a cognizable ‘hazard’ under the Act”); UHS-DE/Pembroke, 2023 WL 3243988, at *1 (upholding a citation for a general duty clause violation for workplace violence hazard at a psychiatric hospital). Non-precedential decisions also concluded that the general duty clause applied to workplace violence hazards. See UHS Centennial, 2022 WL 4075583, at *23 n.27 (finding the hazard of workplace violence at a behavioral health facility to be within the scope of the general duty clause and affirming the citation); HRI, 2019 WL 989735, at *2-8 (applying the Waldon test to assess a citation issued for workplace violence at a behavioral health facility); Megawest Fin., Inc., No. 93-2879, 1995 WL 383233, at *6 (OSHRCALJ May 8, 1995) (finding that the Secretary was “not precluded from asserting that workplace violence constitutes a general duty clause violation). The inability to eliminate a hazard does not preclude the Secretary from asserting a general duty clause violation. See, e.g., Sci. Applications Int’l Corp., No. 14-1668, 2020 WL 1941193, at *4 (OSHRC Apr. 16, 2020) (“SAIC”) (general duty clause applicable to drowning hazard that employer could not eliminate); Arcadian Corp., 20 BNA OSHC 2001, 2011 (No. 93-0628, 2004) (“Arcadian”) (Secretary does not have to show that abatement would eliminate the hazard).

Respondents stipulated that employees of Cedar Springs and UHS-DE were exposed to the hazard of workplace violence at the Worksite. (Stips. 7-8.) It was not unusual for workers to be struck by patients. (Tr. 691.) Workers were punched, bitten, kicked, scratched, and had hair pulled. (Tr. 388-89; Ex. C-2b at 7.) Such assaults frequently occurred, with many injuries resulting from the violence.⁴⁸ This element is established.

Respondents knew the hazard was present and was injuring Worksite employees. (Stips. 7-11; Exs. C-2b, C-3 at 12-16, C-12, C-13 at 4-5, C-14 at 3, C-15 at 3, C-16, C-17, C-43 at 2, C-55, C-57, C-60, C-64, C-65, C-66, C-73.) Managers witnessed, experienced, and had violent incidents reported directly to them. Id. Management also received reports of injuries from workplace violence or reviewed video of such incidents after they occurred. (Tr. 389, 1835, 1903; Exs. C-2a at 14, C-12, C-20, C-48, RCS-41.) Respondents recognized the hazard of patient-on-staff violence, as did their industry, and they had knowledge of its continued presence at the Worksite.⁴⁹

A hazard is likely to cause death or serious physical harm if the likely consequence of employee exposure is serious physical harm. Morrison-Knudsen Co./Yonkers Contracting Co., 16 BNA OSHC 1105, 1122 (No. 88-572, 1993). “Staff injuries from patient violence and/or assaults by patients against staff could result in serious injuries.” (Stip. 11.) See Vanco Constr., 11 BNA OSHC 1058, 1060 (No. 79-4945, 1982). The Secretary established many instances of serious physical harm from the cited hazard, such as workers being choked, headbutted, hit, punched, slapped, kicked, and bit.⁵⁰ Some incidents resulted in head and eye injuries. (Exs. C-3 at 12-16, C-10, C-11, C-12, C-17, C-20, C-59, C-60.) See Vanco, 11 BNA OSHC at 1060.

The general duty clause requires employers to “take all feasible steps” to protect workers from recognized hazards. Gen. Dynamics Corp., Quincy Shipbuilding Div. v. OSHRC, 599 F.2d 453, 464 (1st Cir. 1979). But, before discussing what additional feasible actions could have been taken to reduce the known hazard, we must first consider what the employer already did. See Mo. Basin Well Serv. Inc., 26 BNA OSHC 2314, 2319 (No. 13-1817, 2018); SeaWorld, 748 F.3d at 1210-15 (distinguishing Pelron Corp, 12 BNA OSHC 1833 (No. 82-388, 1986) and U.S. Postal Serv., No. 04-0316, 2006 WL 6463045, at *8 (OSHRC Nov. 20, 2006)).

The Secretary does not dispute that some mitigation was in place at the time of the inspection. Her position is that it was inadequate and failed to reduce the hazard to the extent feasible. Doctors Forman and Lipscomb offered opinions supporting the Secretary’s position. Both opined that the Worksite’s abatement was ineffective at addressing the hazard to the extent feasible. (Exs. C-92, C-94.) Dr. Staggs did not render an opinion on the existing abatement’s effectiveness. (Tr. 2091.)

This section addresses the Worksite’s high injury rate. It then turns to Respondents’ existing abatement, starting with patient management before going on to discuss other aspects of the existing abatement, including, (1) adequate staffing to manage the environment (milieu), (2) protective intake processes, (3) training, (4) incident reporting and debriefing to review and track violence, (5) Respondents’ restraint reduction initiative, and (6) the WVPP. (CS Br. 21-47.)

Violence resulting in worker injuries occurred multiple times before and during the investigation.⁵¹ The CO gathered information about injuries from patient aggression from multiple sources. These included the OSHA logs, injury reports (called Sedgwick reports), and information from workers. (Tr. 110-14, 116-19; Exs. C-10, C-11, C-12, C-13, C-14, C-15.) Injuries from the hazard resulted in lengthy days away and workers needing restrictive duty. (Tr. 110-15, 119-20; Exs. C-3, C-10, C-11.) In 2019, the year the inspection commenced, injuries from workplace violence accounted for 73% of injuries Respondents recorded on the Worksite’s OSHA logs. (Ex. C-3 at 8.) Nearly all the injuries that required days away from work related to workplace violence. Id. These frequent injuries, while not conclusive, support finding that the Worksite’s abatement methods were inadequate as implemented. See SeaWorld, 748 F.3d at 1210-12, 1215 (existing safety procedures found inadequate where evidence showed training and protocols did not prevent continued incidents of injuries).

Respondents argue that the incidents were “unforeseeable.” (CS Br. 75, 77.) The only case Respondents cite when discussing the role of injuries in assessing the adequacy of abatement is an Eighth Circuit decision from the early days of the OSH Act. Id. at 74-76 (discussing Brennan v. Vy Lactos Labs., Inc., 494 F.2d 460 (8th Cir. 1974). In that matter, the Eighth Circuit remanded the matter for the Commission to determine if the employer had actual knowledge of the hazard. 494 F.3d at 464. After remand, the employer withdrew its notice of contest, and the citation was affirmed. Vy Lactos Labs., Inc., 1 BNA OSHC 1774 (No. 31, 1974). Here, the Secretary showed both actual knowledge and recognition of the hazard.

Respondents’ position is unpersuasive. Not all workplace violence is unpreventable. While predicting which patient would act violently was difficult, the unpredictability necessitated more, not less, effort to implement engineering controls, assess risk, and have adequate staffing. (Ex. C-94 at 8, 21, 33.) Respondents know that modifying patient behavior is an uncertain process. Id. Yet, they skipped implementing known controls to minimize risk from unpredictable events. Id. Their abatement resulted in frequent injuries from the hazard and placed workers at risk of more serious injuries.

Many measures Respondents cite as being part of their abatement focus on providing medical care. (CS Br. 25-26, 32-33; Sec’y Br. 15-18, 78-82; Exs. C-92 at 11, C-94 at 44.) They cite clinical processes such as patient observations, communication among staff, treatment team meetings and medical treatment plan updates. Id. In terms of observations, workers observed each patient at least every fifteen minutes to ensure patient safety. (Tr. 2328-29, 2433.) The extent to which information learned through the observation rounds was used to improve worker safety is unclear. MHSs are not required to review prior observation sheets when they start their shifts. (Tr. 2554.) When conducting observation rounds, staff did not typically have radios or other reliable communication devices to call for assistance if they observed aggression or signs a patient might become aggressive. (Tr. 102-3, 690-91, 793-94, 1648; Exs. C-2b, C-92 at 11-12, C-94 at 25-26, RCS-32.) While they could yell for help, doing so could exacerbate the hazard. (Tr. 138, 143, 401, 691, 779, 1516, 1649-50, 2293, 2442.) If observations supported having more staff, there was no automatic process for adding more workers to a unit.⁵² Further, Respondents did not task people solely with security that could provide support. (Tr. 1457, 1510, 1735.) Similarly, placing one or more patients on assault precautions does not automatically increase staff for the unit. (Tr. 842.) For fall, seizure, and infectious control precautions, in addition to noting the precautions in the patient’s file, the patients were also given a colored armband and a magnet for the frame of the door to their bedroom.⁵³ No similar actions outside the paper medical file were used for aggression or assault. (Tr. 2553-54; Exs. RCS-8, C-94 at 16.)

MHSs participated in shift handoff discussions but typically did not receive information regarding specific incidents of patient aggression. (Tr. 642-43, 2554.) The written information completed after a restraint or seclusion was not immediately shared. It first went to the Risk Manager. (Tr. 559-60, 2523.) The purpose of that review was not staff safety. (Tr. 2523-24.) The information collected after workplace violence did not always indicate that an assault on staff was part of the incident. The information collected focused on the patients and their medical care. (Tr. 418-19, 559-60; Exs. C-2b, C-94 at 40-42.) If a patient had two or more restraints in a short period, they would look differently at how to care for the patient. (Tr. 2523-24.) Debriefings focused on patients, not assessing how to prevent or minimize worker injuries. (Tr. 419, 560 630-31; Exs. C-2b, C-94.)

At some point later, the information gathered about the restraint or seclusion would be placed in a patient’s medical file. (Tr. 559-60.) MHSs were not required to review these. (Tr. 560, 630-1, 1431, 2554; Ex. C-92 at 22.) A therapist indicated she did not always have enough information about the patients she interacted with. (Tr. 1420-21.)

The effectiveness of staff communication as an abatement method was also hampered by inconsistent incident reporting. (Tr. 157, 322, 419.) Workers described not reporting assaults. (Tr. 418-19, 746-47, 1448-49; Exs. C-2b at 13, C-3 at 37, 39-40, 44, C-94 at 14.) An MHS explained that he did not report assaults resulting in cuts, scratches, bruising, or what he considered minor injuries. (Tr. 746-47.) A therapist similarly explained that she did not report all injuries because of time constraints. (Tr. 1448-49.)

Patients acting violently or aggressively may be discussed when determining their medical treatment. (Tr. 726, 2516, 2647, 2769-70, 2772-75.) Worker safety was not the focus of treatment meetings. Id. Likewise, patient treatment plans were supposed to be completed after a patient’s behavior necessitated a restraint or seclusion. These treatment updates were not designed or intended to be about worker safety. (Tr. 1417-18; Ex. C-3 at 41, 45-46.)

Workers described incidents for which aggression towards staff did not result in changes to patient management in ways that improved staff safety. (Ex. C-92 at 16.) A nursing supervisor reported to the Director of Nursing about a patient who assaulted at least ten workers in multiple incidents within a week. (Tr. 515, 517; Ex. C-59.) Her report, covering just two days, describes many bites, assaults, a punch to the eye area, and other aggressive behavior. (Ex. C-59.) The nurse did not consider this report out of the ordinary for the Worksite. (Tr. 517; Exs. C-43, C-59, C-60.) A single shift could have multiple assaults, including punches, kicks, and bites. (Tr. 511-12; Ex. C-43.)

Respondents stressed that OSHA is not charged with regulating patient care. (CS Br. 21.) Perhaps to put the acknowledged hazard workers routinely faced and suffered injuries from beyond OSHA’s jurisdiction, many of Respondents’ purported abatement measures are patient care actions. Too often, Respondents attempt to have it both ways. They contend the Secretary may not present any evidence about patient care but then argue their patient care practices constitute adequate abatement.⁵⁴

The Citation is not an attempt to regulate or find fault with clinical care. (Tr. 28, 653-61, 935.) None of the proposed abatement measures relate to clinical care. The Secretary faults not patient care but the handling of the workplace violence workers experienced. (Stips. 7-8.) The clinical processes Respondents cite as part of their abatement may impact worker safety but were not focused on workers or mitigating the cited hazard’s impact on workers. (Tr. 1121, 1194; Exs. C-92 at 11, 19-24, C-94 at 14-17, 21, 44.)

Respondents note their approaches to staffing, which included unit staff that could be supplemented during an emergency, as a key aspect of their workplace violence prevention efforts. (CS Br. 24, 26-32, 36-39.) They also cite measures that occur before and after violent incidents, such as intake processes, training, incident reporting, debriefings, and tracking incidents. Id. at 33-22-23, 34-36, 39-44. And they cite their Restraint Reduction Initiative and other aspects of their WVPP. Id. at 45-47. As implemented, Respondents’ actions did not adequately protect workers.

Respondents recognized the role Worksite staffing plays in addressing the hazard. The WVPP discusses staff’s responsibility to prevent and mitigate workplace violence. (Ex. C-22.) Staffing levels impacted worker safety, and Respondents took steps to maintain “adequate” staffing. (CS Br. 28.) The issue is the efficacy of those steps.

Respondents used a staffing grid that set out staff ratios for the number of patients. (Tr. 834-35; Ex. C-68.) Respondents incorrectly assert that the grid sets out a minimum number of three staff for twelve patients. (CS Br. 57.) Other than the requirement of one nurse for each acute unit, the Worksite had no minimum staffing level. (Tr. 159, 838, 1013, 2593-94, 2617; Exs. C-68; RCS-47.) As stated on the grid, staffing could be further reduced beyond the lowest ratio “based on the unit acuity and staff skill mix.” (Ex. C-68.) Cf. HRI, 2019 WL 989735, at *23 (Secretary failed to show that approach to staffing at a behavioral health facility where the grid was the minimum level of staff and staffing levels were routinely adjusted upward based on aggression and other factors). Further, during the overnight shift, the grid called for only two workers per unit, regardless of the census. (Tr. 159, 838, 2584, 2586-94; Exs. C-68, RCS-47.) Overnight, the staffing ratio did not increase if more patients were awake or needed assault precautions. (Tr. 159, 792, 838, 840, 2330-31; Ex. C-3 at 43.)

Staffing was supposed to be monitored and adjusted based on the number of patients and their acuity. (Ex. RCS-47.) An additional MHS was supposed to be added to the day shift if the number of patients exceeded various thresholds. (Exs. C-68, RCS-47.) In practice, the grid’s ratio was not consistently met or increased for acuity. (Tr. 480, 688, 839, 1010, 2376; Exs. C-33, C-34, C-51, C-55, C-69, C-70; CS Br. 26, 30.) There were “more or less constant staff shortages.” (Tr. 688.)

Respondents set the level of staffing they felt was necessary and then took “a very lackadaisical approach” to meeting that level. (Tr. 1227.) Inadequate staffing was discussed in multiple meetings and written reports. (Tr. 392-93, 475-80, 484, 487-89, 492-95, 845-46, 988-90; Exs. C-33, C-34, C-52, C-54, C-55, C-61.) The Director of Nursing (Mattson) initially claimed staffing reports showing significant understaffing were inaccurate. (Tr. 2490-95, 2498-2503, 2506.) She then clarified that there may have been significantly fewer workers than what the grid called for. (Tr. 2565-78, 2582-94; Exs. C-33, C-55, C-68, RCS-90.) Ms. Degenhardt was present during the time discussed in the shift reports, and her reports are contemporaneous with the events described therein. Her testimony is credited over Ms. Mattson’s. (Tr. 477-81, 492-95, 499-500, 504-5; Exs. C-33, C-34, C-37, C-55, C-57, C-68.)

By limiting every unit to just two workers for the overnight shift, there was chronic understaffing. (Tr. 479-80; Exs. C-2b, C-9, C-34, C-37, C-58.) A single callout created a scramble. (Tr. 159, 688, 1320-21, 1728-29, 2329; Exs. C-3 at 42-44, C-51, C-52, C-55, C-57; CS Br. 29.) UHS Centennial confronted a similar situation. There, like here, employees calling out sick resulted in a struggle to fill the gaps. 2022 WL 4075583, at *10. While the employer viewed this as an “unforeseen” circumstance, First Judge Augustine pointed out that the possibility of insufficient staffing from callouts was a routine concern that Respondent could have accounted for in its staffing decisions. Id. at *49. As in UHS Centennial, predicting who would call out might have been difficult. Still, constantly scheduling only two people per unit left little room for the routine circumstance of someone being unable to work or for patient acuity or numbers to increase during a shift. Id.

The grid was a “paper representation” of who was at the Worksite. (Tr. 1215-17; Ex. C-70.) It did not set a minimum number of staff for worker safety or provide a minimum number of people who could promptly respond to violence. (Tr. 1215-17.) Sometimes, there was a drastic discrepancy between what the grid proposed and working conditions. (Tr. 392-93.) In practice, staffing was not maintained at a level adequate to protect workers. (Tr. 790, 1209-10, 1218-21, 1226-27, 1654-57, 1659, 1799, 2610; Exs. C-92 at 11, 13-17, C-94 at 28-33.)

Workers described situations where they felt there was an insufficient number of workers to maintain safety. (Tr. 392-93, 480, 985-90, 992-95, 1506-8, 1654-55; Exs. C-2b, C-51, C-55, C-56, C-57, C-69, C-70, C-92 at 12, 15, C-94 at 25, 27-28, 30, 34-35, RCS-41 at 7.) They had to intervene in violent situations alone or lacked adequate staff to implement appropriate abatement to prevent worker injuries.

At times, employees were alone in units or with multiple patients. (Tr. 172, 363-64, 473, 484-88, 528-29, 578, 791, 841-42, 958, 961-63, 1022, 1394, 1414, 1424, 1515-16, 1659, 1672-73, 2386-87, 2561-62, 2614, 2596; Exs. C-3, C-9, C-52, C-54, C-70, C-92 at 13, 15, C-94 at 27-32; CS Br. 29-30.) Working alone was not the result of idiosyncratic or discrete events. Id. It routinely occurred. Id. During the investigation, workers were alone when they needed to attempt a hold of a violent patient. (Tr. 961, 963, 1484, 1659-61; Exs. C-54, C-65, C-106, C-108.) On December 8, 2019, a nurse had to restrain an aggressive patient by herself. (Tr. 1482-84; Ex. C-106.) There should have been at least two to three workers doing the restraint instead of her alone. (Tr. 1484.) In another instance, on January 7, 2020, a staff member was alone when a patient grabbed her. (Tr. 1658-61; Ex. C-108.) A housekeeper arrived to assist, but the two workers could not contain the patient. (Tr. 1800-1; Ex. C-108.)

Verbally calling for assistance does not work if no one is there to hear it or assist with the violence. (Tr. 963, 1394.) If a unit lacked its one MHS, no one from that unit could assist with Code Greens or other emergencies elsewhere. (CS Br. 29-30.) The single nurse supervisor might be able to assist, but that left people working alone when the supervisor covered each nurse’s break throughout the overnight shift. (Tr. 2596; Exs. C-43, C-55, C-92 at 15.) It was difficult to arrange the staff to keep the units safe. (Tr. 472; Exs. C-51, C-52, C-92, C-94.) Often, each unit was at the “bare minimum,” preventing supervisors from adding staff when needed to address violence. (Tr. 470-73, 688; Exs. C-9, C-51, C-52, C-55, C-57, C-68, C-70, C-92, RCS-41 at 7, RCS-47.) Although the Worksite’s policies permitted adding or moving staff for acuity, the number of workers scheduled precluded that from occurring. Id.

Respondents insist that being left alone “was not considered a dangerous practice in the industry.” (CS Br. 37 citing Tr. 2458.) The cited testimony did not go so far. The Director of Nursing testified that being alone was “not necessarily” a “dangerous situation.” (Tr. 2458.) As support, she gave examples of when a worker was not actually alone because other workers would still be on the unit, or situations when a worker was left with one patient “ill and in their bed.” (Tr. 2458.) Similarly, Dr. Staggs noted he was aware of staff being briefly alone, such as when one worker took out the trash. (Tr. 2293-94.)

But workers described a wide range of situations when they were alone with multiple patients for more than the brief moments Ms. Mattson and Dr. Staggs cite. The Secretary’s concern is not with 1:1 patient-to-staff ratios or situations when several workers are in the unit. Her position is that staffing practices that frequently left workers alone with multiple patients were inadequate. On occasion, no one was able to respond to calls for assistance. (Tr. 399-400; Ex. C-2b at 8-10.) An MHS described needing to intervene alone against a group of patients. He believed that at least three staff members were needed to manage the situation appropriately. (Tr. 685-87.) The same MHS believed that staffing did not change for acuity and that there were “more or less constant staff shortages.” (Tr. 688.)

Dr. Staggs did not assess whether Respondents’ approach to staffing left employees working alone too frequently. Dr. Lipscomb confirmed the workers’ views about increased susceptibility to the hazard when working alone. In her view, there was consensus that short-staffing or leaving staff alone with an aggressive patient was inadequate to protect workers from workplace violence. (Tr. 1664-67; Ex. C-94 at 31-32.) Dr. Forman agreed that staff were vulnerable when alone. (Ex. C-92 at 11, 13, 15.) The testimony of Doctors Forman and Lipscomb is credited.

Respondents note that no “laws, regulations or [Joint Commission] standards” prohibit staff from being left alone in a unit with patients and argue it is not considered dangerous. (CS Br. 30.) The absence of a specific standard does not mean there was no hazard. (Sec’y Br. 107; Tr. 106-7; Ex. C-1.) For this Worksite, the Secretary showed that working alone made preventing and mitigating workplace violence more difficult. See BHC, 2019 WL 989734, at *32 (discussing working alone with patients at a behavioral health facility as a risk factor for workplace violence). Cf. Brennan v. OSHRC (Hanovia Lamp Div. of Canred Precision Indus.), 502 F.2d 946, 952 (3d Cir. 1974) (remanding decision vacating a citation alleging a violation of the general duty clause to address whether employer exposed employee to serious hazard by permitting him to work alone).

Witnesses described situations where they felt there were insufficient workers to maintain safety or prevent staff injuries. (Tr. 392-93, 480, 985-90, 992-95, 1506-8, 1654-55; Exs. C-2b, C-51, C-55, C-56, C-57, C-69, C-70, C-92 at 12, 15, C-94 at 25, 27-28, 30, 34-35, RCS-41 at 7.)

Restraining patients during aggressive behavior has led to worker injuries.⁵⁵ Respondents’ WVPP and other policies aim to reduce the frequency of restraints and improve the safety of those that must occur. (CS Br. 25, 43.) In practice, insufficient staffing made it more difficult to prevent restraints and perform necessary ones without worker injury. (Tr. 470-73, 480, 623-25, 985-88, 1424, 1428-29, 1451, 1484, 1506-8; Exs. C-51, C-65.) Restraining a patient with more people is safer than restraints attempted by one worker. (Tr. 1196-97, 1482-84; Ex. C-9.) There were times when multiple requests for assistance were necessary because too few staff responded to initial requests. (Tr. 1518; Ex. C-3 at 44.)

Ms. Degenhardt observed a patient throwing his body against and kicking a back door. (Tr. 470; Ex. C-51.) She and another staff member could not restrain him. (Tr. 470-71.) She called for more assistance, but it took a while for sufficient help to respond. (Tr. 470-72.) Five people were needed for the restraint, which had to continue for some time. (Tr. 471.) Workers got tired and their arms began to shake. Id. Additional workers from other units had to “swap” places to relieve those performing the restraint. Id. With every unit at the “bare minimum,” ensuring enough staff were available for the rest of the Worksite was difficult. (Tr. 472-73.) Restraints requiring multiple workers were not unusual. (Ex. C-64.)

Respondents train workers to have three or four people to perform a restraint. (Tr. 1484.) There should be one person on the body, one on the legs, and one talking to the patient. Id. At times, an additional person is needed for the feet. Id. Other staff should move patients not involved in the incident to their rooms or off the unit. (Tr. 1491-92, 1494-95; Ex. C-107.) At the inspection’s start, it was permissible for one worker to attempt a restraint. (Tr. 719, 1484.) But such attempts were supposed to be brief and cease as soon as possible. Id.

Staffing levels precluded consistently implementing the training. On December 8, 2019, patients were assaulting each other. (Tr. 1480-83, 1487; Exs. C-106, C-107.) A nurse had to restrain one of the aggressive patients by herself. (Tr. 1482-84, 1486; Ex. C-106.) The video shows her struggling to hold the patient. (Ex. C-106.) The patient elbows the nurse in the shoulder and head. (Tr. 1483-85; Ex. C-106.) The patient also attempts to head-butt the nurse. (Tr. 1483-84; Ex. C-106.) The practices seen in the videos, including the need to attempt a restraint alone, were not unusual. (Tr. 1494.)

Witnesses described other incidents in which workers believed staff levels were inadequate for staff safety. In January 2020, a nurse and an MHS were with patients who became increasingly aggressive throughout the shift. (Tr. 753-56.) Patients removed the door from an office and were destroying other property. (Tr. 756-57.) The nurse had access to a radio and called a code to try to obtain assistance. (Tr. 757.) It took several minutes for anyone to arrive because codes were occurring in two other units simultaneously. (Tr. 757-59.) Even when everyone available arrived, there was not enough staff to contain the situation. (Tr. 760, 771-73.) The police were then called to assist. (Tr. 760-61, 814.) One worker was punched in the face, one was shoved into a door frame, and others were hit. (Tr. 759-60, 808-10.) A worker was hospitalized and missed fifteen workdays. (Tr. 762-65, 767.) A nurse involved in the incident believed better staffing could have controlled the situation before it escalated beyond their control. (Tr. 760, 772.)

The following month, an MHS was attacked while trying to stop an assault. (Tr. 684.) The MHS believed the situation required at least three workers, but he had to intervene alone. (Tr. 685-87.) In January 2021, another incident requiring more staff than was present occurred. Patients were assaulting staff, making weapons, and destroying property. (Tr. 992-93.) A patient jumped on an MHS’s back and attempted to put him in a chokehold. (Tr. 994, 998.) The police had to be called to assist the staff. (Tr. 995, 1040; Ex. C-73.) An MHS involved in the riot said that staff levels were insufficient to handle the situation safely. (Tr. 994-95, 999.) He did not believe that staffing levels improved after the incident. (Tr. 995-96, 1009-10; Ex. C-73.)

At times, staff could not timely perform tasks related to worker security. Patient care responsibilities conflicted with workers being able to respond promptly, or in some cases, at all, to requests to assist with workplace violence. (Tr. 160-61, 179-80, 747-48; Ex. RCS-29 at 22.) At times, the police had to be called to provide the necessary assistance. (Tr. 85, 414-15, 518-19, 760, 814, 993-95; Ex. C-2b at 12.) Everyone tasked with security was also tasked with many other responsibilities. (Tr. 417; Exs. RCS-29 at 4, 19-22, RCS-30, RCS-72, C-92 at 1, 17.)

Respondents recognized that having more patients per worker can impact the number of aggressive episodes. (Ex. RCS-41 at 7.) They identified the number and type of staff needed for the patient census and then failed to ensure that staffing level was in place. About two months before the inspection commenced, the CEO was notified of “significant staff shortage.” (Ex. C-80 at 1.) UHS-DE’s milieu specialist noted: “short staffing has had a very negative impact on accountability and oversight. It is imperative we maintain [] established staffing levels.” Id. at 3. So, the staffing level OSHA observed was not a brief period of insufficient staffing. Id. at 1, 3. In fact, at the inspection’s start, the staffing level was “the best” it had been “in a long time.” (Ex. C-36 at 1.)

A Code Green was a way to request available workers from other units to assist with an aggressive patient or patients. (Tr. 121, 142; Exs. C-9, RCS-5.) Former workers described responding to one or more Code Greens per shift. (Tr. 689, 792, 1494; Exs. C-57, C-59.) Frequently, more than one code occurred at the same time. (Tr. 689, 792-93, 1494, 1518.)

To commence the process, a worker could use a radio, the unit phone, or by yelling “Code Green.” (Tr. 103, 138, 142, 364, 400-1, 461-62, 690-91, 963, 1516, 1649-50, 2888; Exs. C-31, RCS-5, RCS-18 at 5.) Not everyone could use a radio to commence a Code Green. There were not enough functioning radios for all staff members to carry one. (Tr. 102, 138-39, 400, 506, 508, 691, 793-94, 1414, 1419, 1516, 1648; Exs. C-2b, C-42, C-57.) During the inspection, a former charge nurse twice reported issues with radios to senior management. (Tr. 508-59; Exs. C-42, C-57.) She believed there were not enough radios for the units to be safe. Id. Typically, staff only had a radio when escorting one or more patients from the patient care unit to another location. (Tr. 102, 138-39, 506, 690-91, 793-94; Ex. C-57.) While the patient care units generally had more than one working radio, only one was in the A&R department. (Tr. 1648-49.) So, A&R staff did not carry a radio when transporting a new patient to a patient care unit for the first time. (Tr. 138-39, 1648-49.)

When they could not access a radio, staff could try the phone at the nurses’ station or yell for help. (Tr. 103, 138, 142, 364, 401, 461-62, 963, 1516, 1649, 2888; Ex. C-31.) Staff sometimes had difficulty getting to the phone to commence a code. (Tr. 103, 364, 401, 461-62, 564-65, 1433, 1517; Ex. C-31.) There were no phones in the offices where the therapist worked. (Tr. 103, 1419, 1650, 1723.) Even when working near the phone, the nature of the workplace violence incident sometimes precluded using it. (Tr. 103, 461-62, 2392; Ex. C-31.) Workers would have to “yell ‘Help’ as long as possible” and “hope” they would be heard. (Tr. 691, 963.) When working alone, yelling for help has an obvious limitation, as no one would be there to hear it. (Tr. 1394.)

Besides issues with commencing a Code Green, the process was also ineffective when there was a lack of staff available to respond. (Tr. 180; Ex. C-2b at 9.) Staff could not respond when engaged in the 1:1 care of a patient, conducting body scans (i.e., examining a patient for injuries and searching for contraband), conducting observation rounds, providing medication, or transporting patients. (Tr. 121-21, 160-61, 747-48, 1797, 2373-74, 2377; Exs. C-3 at 42, C-92 at 14, RCS-7 at 1, RCS-30, RCS-31.) Workers also could not respond if they were already engaged in the de-escalation of a patient or if other situations prevented them from leaving their assigned unit. (Tr. 747-48, 2373-74, 1797-99; Exs. C-3 at 42, RCS-29 at 4, 19-22.) Certain staff, such as therapists, were not required to respond to Code Greens. (Tr. 1418.) It was common to have codes called multiple times because of an insufficient response to the first call. (Tr. 162-63, 399-400, 1482-83, 1494, 1517-18; Ex. C-3 at 44.) Worse, at times, no one was able to respond promptly. (Tr. 162-63, 399-400, 807; Ex. C-2b at 9-10.)

Calling a Code Green had adverse ripple effects. When patients heard the code or saw staff leaving their unit to respond, sometimes they would escalate aggressive behavior. (Tr. 142-44, 366, 1518; Exs. C-3 at 44, C-94 at 26.) Those units would then have less staff to de-escalate the patients or perform other tasks to prevent or reduce aggression. (Tr. 142, 366, 1495, 1518; Ex. RCS-29 at 4, 19-22.)

The Commission addressed the importance of reliable methods to summon assistance for those working at inpatient behavioral health facilities. In UHS-DE/Pembroke, like at the Worksite, there were not enough walkie-talkies for employees to always carry. 2022 WL 774272, at *9-10. Instead, employees had to use phones, which were in only limited locations, or yell for assistance. Id. at *9. The Commission upheld the ALJ’s finding that this approach was not sufficiently effective at abating the hazard of workplace violence at the behavioral health facility. Id. at *10.

A similar finding is appropriate here. Doctors Lipscomb and Forman concluded that the ability to obtain assistance was inadequate. (Tr. 1648, 1653; Exs. C-92 at 11-12, C-94 at 24-26.) Sometimes this was because of difficulties in calling the code. (Tr. 506, 508-9, 694-95, 794; Exs. C-42, C-57 at 4.) At other times it was because there was not enough staff who could respond promptly. (Tr. 162-63, 399-400, 1518.) Slower responses increased the likelihood of injury and could exacerbate injuries. (Tr. 1199-1200; Exs. C-92 at 11-12; C-94 at 26.)

Respondents argue that the employee witnesses were biased. The compelling witness testimony was supported by overwhelming evidence of routine violence and serious injuries from the hazard. Respondents do not rebut the evidence showing that the amount of available staffing contributed to or exacerbated injuries. Further, they destroyed evidence of whether adequate numbers of staff were available to respond to incidents.

Insufficient staffing was a recurring and significant problem that exacerbated the hazard.⁵⁶ To put it mildly, Respondents’ claims about their ability to augment staffing were overstated. Typically, non-direct care staff were only available on one of the three weekday shifts and not on weekends. With the facility operating 24/7, most of the time, few non-direct care workers could assist with abating workplace violence or its precursors.

Respondents recognized that having more patients per worker can impact the number of aggressive episodes. (Ex. RCS-41 at 7.) Dr. Forman opined that the Worksite’s staffing did not appropriately account for patient acuity and fell short of what is required to maintain a safe workplace.⁵⁷ Dr. Lipscomb agreed that the Worksite’s staffing was inadequate for staff safety and led to incidents of workplace violence. (Tr. 1626, 1654, 1656-58, 1664-65; Ex. C-94 at 27-33.) Respondents’ pre-inspection approach to staffing did not adequately abate the hazard.

Admissions and intake processes are part of the Worksite’s abatement. (CS Br. 22-25, 36, 68-69, 103-4.) Most patients are brought directly to a patient care unit after being referred by a hospital. (Tr. 258-59, 2310-11, 2419.) When a patient arrived after a hospital referral, they were physically searched in their rooms by two staff members for contraband and weapons. (Tr. 2319-20.) Their belongings were not scanned with a metal detector or immediately searched. (Tr. 1501.) At the start of OSHA’s inspection, such equipment was not in the buildings with the patient care units. (Tr. 1501; Ex. C-38.) Respondents claim that patient belongings were placed “in a locked location.” (CS Br. 22.) However, at the investigation’s start, staff placed patient belongings under the desk in the nurses’ station. The nurses’ stations could be locked but as noted, some patients were able to gain access.⁵⁸

The belongings remained at the nurses’ station until staff had time to search and inventory their contents. (Tr. 2892; Ex. C-3 at 29.) The searches were to be conducted either during the shift the patient arrived on or on the next shift. (Tr. 2321.) It could be many hours before anyone knew if weapons or other dangerous items were in the belongings. (Exs. C-3 at 29, C-94 at 35.) For example, a nurse discovered bags that had been at the nurses’ station for two days before they were searched. (Tr. 497-99, 582-83; Ex. C-56.) The belongings contained hazardous materials, including butane fuel and numerous lighters. (Tr. 497-99; Ex. C-56.)

The assessment for contraband also had concerning gaps for patients who first went to the A&R department. Potential patients kept their belongings until they went into an assessment room. (Tr. 91.) The belongings were not checked with metal detectors or searched before the assessment. Id. If admitted, the patient and their belongings were supposed to be scanned with a metal detecting wand. (Tr. 91, 176; Exs. C-3 at 29, C-39.) The wands were not particularly sensitive and would not detect metal if the patient was wearing baggy clothing or if the wand did not come extremely close to the person. (Tr. 359, 902-3, 1204, 1672-73, 1963, 2551; Exs. C-3 at 29, C-39.) A&R staff did not physically search patients or their belongings. (Tr. 91, 176, 260, 359; Exs. C-3 at 29, C-39.) There was no inventory, and no one knew whether the patients had any contraband not picked up by the metal detector. (Tr. 899, 2548-49; Exs. C-38, C-39.)

A staff member would transport the patient with their belongings from A&R to one of the buildings with patient care units. (Tr. 91, 260, 896.) Initially, there was no policy or procedure about whether staff or the patient held the belongings when the patient went outside of the A&R building to a patient care unit. (Tr. 901-2, 2548, 2551; Exs. C-39, RCS-15.) Sometimes staff carried the belongings, and sometimes patients did. (Tr. 91, 175, 261, 1204, 1314-15.) Like patients arriving after a referral, the belongings of patients who entered through A&R were placed at the nurses’ station to be inventoried later. (Tr. 260, 360, 1502, 2892; Exs. C-38, C-39, C-92 at 13, RCS-72.)

In December 2019, a patient was able to bring a large butcher knife into the care unit. (Tr. 174, 898, 1496-98; Exs. C-3 at 28-30, C-39.) The patient’s belongings were not searched in the A&R department. (Tr. 91, 176, 899, 2548-49; Ex. C-39.) After admission, the patient carried his belongings to a patient care unit (Hickory). (Tr. 899; Ex. C-39.) The staff member escorting him did not have a radio while accompanying the patient across the Worksite. (Tr. 175, 899; Ex. C-94 at 24.) The patient was able to keep the knife in his possession until the following day. (Tr. 1496-97.) Additional contraband in the patient’s belongings was also found after the knife was discovered. (Tr. 1498.) Management reviewed video of the admissions process and the patient’s entry into the unit. (Tr. 899; Ex. C-39.) An MHS from the unit staff was terminated, but no one from the A&R department was. (Tr. 264-65, 360, 1570; Ex. C-3 at 29.) Dr. Lipscomb considered it a matter of luck that something did not go seriously wrong. (Tr. 175-76, 361, 1672.)

Respondents had policies addressing screening patients and their belongings for items that presented a safety concern. (Tr. 260, 2321, 2423, 2425, 2547-51; Exs. C-22, RCS-15, RCS-72.) At the start of OSHA’s investigation, the policies did not address several key aspects of the screening process. For instance, the “Initial Safety Screen of Patients and Securing Belongings” policy did not address if patients could retain possession of their belongings after being admitted in the A&R department. (Tr. 2548, 2551; Ex. RCS-15.) The policy refers to using a metal detector wand on the patient but does not address searching or using metal detecting wands over patient belongings. Id. Similarly, the “Control of Contraband” policy did not state that any screening of belongings was to occur in the A&R department. (Ex. RCS-72.) It also did not specify when to screen belongings after patients arrived on the units. Id.

Respondents recognized the risks associated with patients bringing dangerous materials into the units. Their training materials stated (in bold) that inventories would occur upon arrival. (Tr. 2891-92; Ex. RCS-30 at 2.) In practice, this was not followed. (Ex. C-3 at 29 “Employees stated in interviews that it could be several hours before a patient’s belongings were inventoried.”.) Dangerous items were brought into the care units and left unsecured. (Tr. 2892; Exs. C-92 at 12-13; C-94 at 24-25, 35-46.) Possessions remained at the nurses’ stations for hours when staff had too many patient care tasks to complete the searches. (Tr. 360, 582-83, 1570, 1629, 2321, 2548; Ex. C-56.)

Doctors Lipscomb and Forman believed the policies and practices in place at the start of OSHA’s investigation were inadequate for the hazard. (Tr. 1202-5, 1208, 1313, 1397-98, 1670, 1672-74; Exs. C-22, C-92 at 12-13, C-94 at 34-36.) Their expert opinions, coupled with the other evidence, established that Respondents’ approach for identifying and securing dangerous items was inadequate to protect staff from the hazard when OSHA’s inspection commenced.

All employees received training in de-escalating aggressive behavior and responding to behavioral health emergencies. (CS Br. 33-36.) The Secretary agrees that attempting verbal de-escalation can be a good approach to some behavior. (Sec’y Br. 81.) The proposed additional abatement does not include a stand-alone provision on training. Nonetheless, the Secretary maintains that training, to be effective mitigation, required adequate resources so workers could implement the actions called for by the training. Id.

At the time of the inspection, workers were trained in a program called SAMA. (Ex. C-3 at 40, 46.) The training instructed workers that restraining a patient alone is generally prohibited. (Tr. 2889; Ex. RCS-22 at 39.) In practice, workers were often alone when patients were aggressive toward them. Sometimes, they were the only staff member around when patients were assaulting other patients or self-harming. They did not have time to start the Code Green process or wait for a second staff member before attempting to restrain a patient. (Tr. 640-41, 963.) Although permissible, it was typically challenging for one person to maintain a hold alone successfully. (Tr. 2390; Ex. C-79 at 2.) Even five people would not be able to restrain certain patients safely. (Tr. 516-19; Ex. C-59.) A former lead MHS discussed an unsuccessful solo hold attempt. (Tr. 623-25.) He tried to stop a patient from entering the nurses' station’s window. (Tr. 623.) As he tried to restrain the patient, the patient broke free, and the worker’s finger was lacerated. (Tr. 623-24, 626.) The patient was still not in control and the worker had to initiate a second hold alone. (Tr. 625.) This time, the worker’s knee was injured. Id.

Difficulties with initiating calls for assistance and receiving adequate and timely responses complicated workers’ ability to perform the techniques taught in a manner that minimized injury to the worker. Additionally, as Dr. Forman explained, “nothing is more confusing to staff than training them to do things one way then not giving them enough teammates to use the training in real-world situations.” (Ex. C-92 at 14.) He also discussed the value of having staff with specialized security training. Id. at 17. Dr. Lipscomb also raised concerns with not having staff with specialized security training. (Ex. C-94 at 16, 36-40.) Considering the conditions workers confronted, the techniques taught could not be implemented consistently and were not sufficient.

Promptly after a restraint or seclusion, a nurse was supposed to complete a debriefing form. (Tr. 182-83, 2344; Exs. C-57, C-64, C-65, C-66, C-79.) Later, the Risk Management Director (Kolln) and others were supposed to review video of restraints and seclusions. (Tr. 1835, 1903; Exs. C-20, C-22, C-79.) There were notable inconsistencies and gaps in how debriefings occurred. (Ex. C-3 at 41.) The debriefing process focused on patient feedback and patient safety. (Tr. 182-83, 1634; Ex. C-94 at 41.) It was not particularly formal. (Tr. 888; Ex. C-94 at 40-42.) Although the forms had a section to document staff/patient debriefings, the information included was often incomplete. (See Exs. C-64 at 9, 23, 36, 59, C-94 at 41.) It is unclear how the forms kept in the paper patient medical file were used for staff safety.

In June 2019, Kendra Stea, a UHS-DE Program Manager/Clinical Trainer, visited the Worksite. (Stip. 42; Tr. 917; Ex. C-79 at 4.) She advised the Director of Nursing and other senior leaders that “all incidents should have a staff debriefing as well as a patient debriefing. By all accounts this was intermittent, or non-existent.” (Tr. 919, 1242, 1335; Exs. C-79 at 5; RCS-41 at 5.) The on-site trainers appeared to know how to conduct debriefings, but there was an issue of “accountability” in ensuring the practice consistently occurred. (Ex. C-79 at 1, 5.) Ms. Stea advised the Worksite’s senior management to move from their current approach “toward 100% camera review and 100% staff debriefing.” Id. at 8. This June 2019 recommendation for camera reviews and including staff in such reviews was not new and had previously been raised as something for the Worksite to implement. (Tr. 2611, Exs. C-79, RCS-41 at 5-6.) A month later, debriefings were still a “struggle.” (Tr. 2611-12; Ex. RCS-41 at 7.) And, by fall, still not much had changed. Debriefings were still “not adequate.” (Ex. C-80 at 3.)

Frequently, direct care workers were not involved in the debriefing process, even for workplace violence incidents during which they were injured. (Tr. 183, 643, 651, 672-73, 682, 966, 980, 983, 992, 1420, 1433, 1438, 1471-73, 1495, 1507, 1510, 2394-97; Exs. C-3 at 45, C-64, C-79 at 5, 8, C-92 at 19, C-94 at 40-42.) Debriefings could occur when the key staff involved were unavailable. (Tr. 2344; Exs. C-57, C-64, C-92 at 19.) Ms. Nolan acknowledged that although she would try to gather as many people as possible, the debriefing might include only one person. (Tr. 2344.) Routinely, staff were needed back in their assigned units to complete other tasks after a code. (Tr. 1473, 2344; Ex. RCS-41 at 7.) They would not participate in any debriefing because they needed to quickly return to their respective units where, at night, they typically had left one worker alone while they assisted with the code. Id.

Despite being involved in numerous restraints of Patient 23, an MHS did not recall participating in any debriefings after those events. (Tr. 643, 651, 672.) Similarly, a nurse did not participate in a debriefing on February 12, 2020, after a patient injured her. (Tr. 1472.) The debriefing form is consistent with her testimony, as it does not identify the workers involved with the restraint. (Tr. 1471-72; Ex. C-66 at 5.) A worker explained that sometimes there would be discussions about what happened but did not recall ever discussing measures that could be taken to handle things differently in the future. (Tr. 735; Ex C-3 at 45-46.)

The input of involved workers was not consistently captured after incidents and subsequent reviews also left them out. (Exs. C-79 at 8, C-80 at 3, RCS-41 at 5.) The Worksite had several committees, but none focused on workplace violence or staff safety. (Tr. 154-55, 229, 321-22, 1634-35, 1701, 1825-26, 1920, 1983, 2520, 2661; Exs. RCS-3, RCS-84 at 8.) Direct care workers did not participate in camera reviews. (Tr. 1634; Ex. RCS-41 at 5-6.) They also did not regularly attend PIC or PSC meetings, where aggregate data related to restraints and seclusions was reviewed.⁵⁹ The data reported at those meetings came from the MIDAS database, which typically did not note worker injuries. (Tr. 560, 1825; Ex. C-94 at 19.) Respondents did not effectively analyze systemic risk factors, such as the number and type of staffing present during incidents or where they occurred. (Ex. C-94 at 19.) Management appeared confused about who was responsible for tracking and trending worker injuries related to patient aggression. Id. (the Director of Risk Management and Human Resources each indicated that the other was responsible). The Risk Management Director reviewed reports to see if patients had to visit an emergency room, but those reports would not indicate if workers needed medical care from the incident or because of other workplace violence. (Tr. 2603.)

The incident investigation and analysis involved limited data collection and analysis. (Ex. C-94 at 19.) The information gathered focused on assessing the individual patient’s treatment, not improving worker safety. (Ex. C-3 at 41, 46.) It amounted to a “paper exercise.” (Tr. 2611; Exs. RCS-41 at 5-6, C-79.) Little attention was paid to getting feedback in the form of post-incident debriefings. (Exs. RCS-41 at 5-7, C-79 at 5, C-94 at 17-19.) Worse, the limited data obtained was not effectively used to reduce the hazard’s incidence or severity. (Exs. C-92 at 18-19; C-94 at 19.) As Dr. Forman explained, the post-incident investigation and debriefing at the Worksite was deficient, sometimes to the point of being “non-existent.” (Exs. C-79, C-92 at 11, RCS-41 at 5-6.) This denied management and staff the opportunity to improve future responses. (Ex. C-92 at 11, 18-19.) His assessment, backed by other record evidence, contradicts Respondents’ claims that they use debriefing and incident review to abate the hazard effectively. In short, Respondents’ claims about tracking, trending, and analyzing data related to workplace violence as a method to reduce the hazard are significantly over-stated. Their approach, as implemented, did not adequately address, or contribute to staff safety. (Tr. 1634, 1684-85; Exs. C-3 at 45, C-64, C-92 at 19, C-94 at 40-41.)

The impact of Respondents’ restraint reduction efforts on the hazard is unclear. (Exs. C-3 at 51-52, C-79 at 4; RCS-41.) Ms. Stea noted issues with how restraint reduction was implemented at the Worksite: “the current team structure has not been very effective lately.” (Stip. 42; Ex. C-79 at 4.) Respondents did not produce any records from meetings during OSHA’s six-month investigation. (Ex. RCS-41.) The minutes from the restraint reduction team meetings in 2018 and the first half of 2019 do not refer to staff injuries. (Tr. 2610-12; Ex. RCS-41.)

Dr. Gray, the Medical Director, indicated that patient restraint and seclusions were discussed in Medical Executive Committee (“MEC”) meetings. (Tr. 2662; CS Br. 45.) He did not provide details about the discussions, and no minutes or other records from MEC meetings related to this issue were introduced into the record to corroborate his assertion. Restraints and seclusion were agenda items for the Performance Improvement Committee (PIC). (Ex. RCS-84.) However, the extent to which the PIC meetings reduced the hazard’s incidence was not established. The PIC’s meeting minutes for January 2019 and January 2020 had identical discussions regarding seclusions and restraints. (Tr. 2834-38; Exs. RCS-84 at 1, 8, 30, 37, C-3 at 35.) The minutes discuss a reduction in restraints from 2017 to 2018, without discussing the number of restraints in 2019. (Tr. 1827-28, Exs. C-74 at 1, RCS-84 at 1, 30.) Similarly, the January 2020 meeting minutes for the Governing Board reflect restraint information from 2018 rather than 2019. (Tr. 1875-76; Ex. C-76.)

The record suggests difficulties with analyzing restraints and implementing actions to minimize worker injuries from the cited hazard. (Tr. 1951-52, 2612; Exs. C-3 at 51-52, C-80 at 2, RCS-41, RCS-84 at 1, 8, 30, 37.)

Respondents have a WVPP. (Tr. 145, 1817; Ex. C-22.) The document is dated April 2019, but the Risk Management Director (Kolln) could not recall when it was implemented. (Tr. 1817; Ex. C-22.) The Secretary established a rift between what the WVPP calls for and Worksite practices. This undermined the abatement’s effectiveness. (Tr. 144-51, 1621-26, 1629-30; Exs. C-3 at 32-54, C-92 at 12, 17, 19.)

First, the WVPP called for “zero tolerance” for all types of workplace violence, including violence committed by patients upon staff members. (Ex. C-22 at 1.) Dr. Lipscomb testified that she did not agree with Respondents’ claim of zero tolerance and characterized it as “an empty sort of policy statement.” (Tr. 1755. See also Tr. 146, 696, and 1456.) Dr. Forman’s report characterized the WVPP “zero tolerance” position as “ill-defined” and noted Respondents’ “high tolerance” for workplace violence. (Ex. C-92 at 17.) Workers agreed. (Tr. 146, 696, 1453-56.) Respondents appear to acknowledge they did not adhere to “zero tolerance” as the WVPP states. (CS Br. 94.) Instead, they took a “practical approach” to patient aggression against staff. Id.

Second, while Respondents claim workers were trained on the WVPP and could access it, multiple workers did not know there was a WVPP, what it recommended, or what it required. (Tr. 148-49, 695-96, 797, 1452-54, 1623-24, 1822; Ex. C-3 at 33; CS Br. 42, UHS-DE Br. 23.) Workers indicated they did not receive training specific to the WVPP and did not receive a paper copy of it. (Tr. 148-49, 695-96, 797, 1452-53.) The binder of information provided to new employees did not include the WVPP or specify how to access it. (Tr. 148-49, 695-96, 1056-57, 1452-53, 1822.) The Human Resources Director (Robert Troudt) did not cover the WVPP in new employee orientation and was unsure if there was annual training on it.⁶⁰ Similarly, Ms. Kolln did not know if there was training on the WVPP. (Tr. 1822-23.)

Management did not appear to be well-informed about the plan. (Ex. C-3 at 32-33.) The WVPP set a goal to reduce staff injuries from patient aggression by 15% from their 2018 level. (Ex. C-22 at 1.) Ms. Kolln did not know where the goal came from, how many injuries a 15% reduction would correspond to, or if the goal was ever met. (Tr. 1823-24; Ex. C-3 at 33.) Mr. Troudt was similarly unaware if the goal was met in 2019 or the following year. (Tr. 1058.)

Responsibility and accountability for the plan was muddled. (Ex. C-3 at 34.) Ms. Kolln could not recall or explain multiple aspects of the short document. (Tr. 1823-24, 1854; Exs. C-22, C-94.) She could not recall participating in any review or evaluation of the WVPP, even though the document calls for annual reviews. (Tr. 1854.) Besides Ms. Kolln, the WVPP also specified that the Director of Human Services was a responsible person for the plan. (Ex. C-22 at 7.) There was no such position at the Worksite. (Tr. 1055-56.) There was a similar sounding position (Human Resources Director), but that person did not have responsibilities to prevent workplace violence involving patient aggression toward staff. (Tr. 1054.)

Third, the WVPP’s review and evaluation elements were not implemented. The WVPP called for “Employee Accident Reports” to be reviewed to identify injury and illness trends. (Ex. C-22 at 7.) This was to be done to see whether corrective action was needed. Id. However, Ms. Kolln was unaware of such reports being used at the Worksite. (Tr. 1854-55.) Similarly, the WVPP called for reviewing workers’ compensation claim reports, OSHA 300 logs, and “Root Cause Analyses conducted for employee injuries.”⁶¹ But Ms. Kolln either did not receive or did not review such documents. (Tr. 1855; Ex. C-3 at 45.) She could not confirm the next two items the WVPP lists (post-incident debriefings and existing corrective action plans) were part of the review procedure. (Tr. 1855-56.)

The only item on the list she did acknowledge reviewing for illness and injury trends was the HPR (Healthcare Peer Review) reports. (Ex. C-22 at 8.) Those reports were generated from information in the Worksite’s MIDAS database. (Tr. 1824-25.) The MIDAS system allows workers to input data about incidents and categorize them. (Tr. 1853.) It is a patient-based system. (Tr. 560, 1825, 1836.) It is not used to record or report worker injuries and there is no field to enter them. (Tr. 560, 1824-25.) This undercuts their utility for identifying trends and implementing corrective action for worker safety.

Fourth, the WVPP calls for more robust monitoring and assessment of workplace violence than what occurred. The WVPP tasked two committees (PSC and PIC) with monitoring patient aggression towards staff. (Ex. C-22 at 1-2.) Specific incidents of patient aggression to staff were not typically discussed at those committees' meetings. (Tr. 1825-26.) As the name implies, the Patient Safety Council (PSC) focused on high-risk incidents involving patient care. (Tr. 1829-30; Stip. 43.) Its meeting minutes support this characterization. They do not reflect discussion of specific incidents of patient-on-staff violence or worker injuries. (Tr. 1831-33; Ex. C-75.) For the PIC, worker injuries were not a priority or performance improvement area. (Stip. 44; Tr. 1062-64, 1826-28; Exs. C-3 at 49, C-74 at 1-2.) The committee focused on issues related to patient care, such as restraints, medication errors, and falls. (Tr. 1826-27; Ex. C-74.)

There are more gaps between what the WVPP called for and what occurred at the Worksite concerning the analysis of aggression. (Tr. 146-51.) The WVPP specifies that “senior management will ensure that patient aggression events are analyzed and that trends are identified for the corrective action process.” (Ex. C-22 at 3-4.) To conduct the trending the WVPP required, Ms. Kolln reviewed incidents categorized as “patient aggression towards staff” in the MIDAS database. (Tr. 1835, 1841.) The information she obtained this way does not indicate whether there was an employee injury from the aggression. (Tr. 1836, 1846, 1849.) When entering information into MIDAS, staff can categorize incidents in several different ways. (Tr. 1841-46; Exs. C-19, C-20.) If an incident involved an assault by a patient against a worker, it might be categorized as “patient out of control,” or some other category rather than “aggression/assault.” (Tr. 1841-46, 1848-49; Exs. C-19, C-20.) Ms. Kolln did not know why some incidents involving staff assaults were categorized as “patient out of control” rather than “aggression/assault.” (Tr. 1845-46.) She did not know why there were different categories and did not explain why she only looked at reports specifically categorized as “aggression/assault patient toward staff” when conducting the assessment. (Tr. 1835-36, 1841, 1852-53.) The Secretary identified multiple incidents that involved assaults against staff but were categorized as “patient out of control.” (Tr. 1838, 1841-45, 1848-49; Exs. C-19, C-20.) Looking at only one of the relevant categories limited the exercise’s utility as mitigation.⁶²

Fifth, the WVPP calls for employee participation and lists actions that purportedly demonstrated such participation. (Ex. C-22 at 2-3.) Some workers were unaware of the two committees (PSC & PIC) the WVPP cites as evidence of employee participation in addressing the hazard. (Tr. 1010.) Direct care staff did not attend the PIC meetings and were not typically at the PSC meetings either. (Stips. 43-44; Tr. 154, 301-2.) Moreover, there was not much participation from direct care in the restraint and seclusion reduction team’s meetings. (Tr. 2829, 2831-32; Ex. RCS-41.) Employee participation in preventing and mitigating workplace violence in behavioral health is a “foundational” and “critical” element of prevention. (Tr. 1607.) Respondents’ WVPP failed to encourage employee participation and this aspect of the WVPP was deficient. (Tr. 150, 1622-24, 1633-35; Ex C-3 at 37-39, 52.)

Sixth, the “Hazard Identification and Control” section was not fully implemented, and the measures taken focused on patient rather than staff safety. (Exs. C-22 at 4-6; C-94 at 19-20.) Management was not conducting adequate hazard analysis and surveillance related to workplace violence at the time of the inspection. (Tr. 153, 1629-30.) Respondents had a process to conduct “environmental rounds.” (CS Br. 36, 68-69.) However, as executed, these primarily looked for maintenance issues and objects patients might use for self-harm. (Tr. 1629-30; Ex. C-94 at 19.)

Respondents complain that the Secretary is trying to enforce their policies. That is a mischaracterization. Respondents claimed their WVPP was implemented and adhered to OSHA guidance on how to mitigate workplace violence in healthcare. The Secretary showed notable confusion about the plan and how its implementation differed from what is considered appropriate for mitigating the hazard at facilities like the Worksite. (Exs. C-22, C-92 at 20, C-94 at 14-20.) Overall, the WVPP was focused more on patient safety than worker safety. (Tr. 148-49; Exs. C-22, C-94 at 14-20.) “Policies are very important, but the effectiveness of a policy is going to be largely dependent upon [] the fidelity to it and the implementation of it.” (Tr. 1250-51; Ex. C-22.) As implemented, the WVPP was not effective at addressing the hazard at this Worksite. (Tr. 146-51, 1621-26, 1629-30; Ex. C-92 at 12, 17, 19.) See Nelson Tree Servs., Inc. v. Sec’y of Labor, 60 F.3d 1207, 1211 (6th Cir. 1995) (fact that an employer “incorporated the relevant” safety measures “into its own safety manual does not satisfy its obligation” under the General Duty Clause unless those measures are actually “followed and enforced.”)

Respondents took some actions to address the known hazard of workplace violence. However, the hazard persisted and remained a frequent cause of serious injuries. Respondents lacked a comprehensive and implemented approach to workplace violence abatement. Many things they claim to be part of their abatement efforts appear to be “re-purposed,” i.e., they re-characterized a policy or procedure related to patient care and argue it is meant to protect staff. Undoubtedly, many things done at the Worksite are dual-purpose in that they help protect patients and staff. Still, upon examination, it is apparent that fully implementing abatement measures to adequately protect staff from the hazard was not prioritized. The actions taken were “paper exercises” rather than committed efforts to take feasible measures to reduce the incidence of the hazard. (Tr. 2611; Exs. C-3, C-92 at 14; C-94 at 14, RCS-41.) See BHC, 951 F.3d at 565 (finding inadequate abatement when employer had policies “on paper” but did not fully implement them); Nelson Tree, 60 F.3d at 1211.

Respondents claim they have implemented effective abatement. But when the Secretary shows that the claimed abatement is not in place, they contend the Secretary is enforcing policies. Stripping away the hyperbole, it is plain that Respondents knew the hazard was present and had identified means to reduce its incidence at the Worksite. They laid out steps in written policies and provided certain training. Nonetheless, they failed to ensure adequate staff and resources to implement the abatement. This failure undermined the abatement’s effectiveness at protecting workers from the cited hazard and rendered it inadequate.

As the next section addresses, the Secretary showed that there were several additional feasible actions Respondents could have taken to mitigate the hazard further.

When the existing abatement is inadequate, the Secretary must propose some actions the employer could have taken. Nat’l Realty, 489 F.2d at 1268. Any measure will not suffice. The Secretary must establish that her proposed action or actions will be effective at addressing the hazard. Arcadian, 20 BNA OSHC at 2011. Further, the Secretary cannot propose measures fundamentally incompatible with the business such that implementing them will result in bankruptcy or preclude operations altogether. Pelron, 12 BNA OSHC at 1834-35 (proposed abatement would have required the plant to close, so it was not feasible). Simply put, if there’s a known hazard capable of causing serious injury or death, but nothing can be done to eliminate or materially reduce it, there is no OSH Act violation.

The Citation sets out both engineering controls and administrative controls as feasible and effective means to abate the hazard of workplace violence. The engineering controls include: (1) reconfiguring nurse stations to prevent patients from entering them, and (2) providing communication devices and silent alert signals to employees who work near patients. Broadly, the proposed administrative controls to further reduce the hazard are:

(1) develop and implement additional procedures related to identifying and securing contraband;

(2) implement a comprehensive WVPP that includes conducting investigations and debriefing after workplace violence;

(3) maintain staffing that is adequate to safely address changes in patient acuity and patient census; and

(4) designate specific staff with specialized training in security to monitor patients for potential aggression and respond to calls for assistance.

(Sec’y Br. 94-122.) In disputing this aspect of the Secretary’s burden, Respondents argue that: (1) the Citation proposes various alternative methods of abatement, and they already implemented some of the measures, (2) the Secretary failed to show the proposed abatement methods will reduce the hazard, and (3) the Secretary did not establish the proposed abatement methods are feasible. (CS Br. 77-120.) In some respects, Respondents’ arguments are muddled and inconsistent. They contend they took actions the Citation proposed as abatement and then argue that the Secretary did not prove that those actions were operationally or economically possible. Id. at 67-74, 111-20. Cutting through the discrepancies, it is apparent that: (1) the Citation proposed multiple actions, not various alternatives; (2) the proposed methods of abatement will materially reduce the hazard, and (3) Respondents can implement the proposed abatement.

Respondents, relying heavily on A.H. Sturgill, No. 13-0224, 2019 WL 1099857 (OSHRC Feb. 28, 2019), argue that the Secretary proposed these actions as alternatives rather than separate actions to be part of a comprehensive abatement. (CS Br. 98, 117-20, 127, 128.) In its most recent decision addressing the hazard of workplace violence, the Commission does not rely on A.H. Sturgill. UHS-DE/Suncoast, 2023 WL 2388069, at *2-3. Instead, it identifies UHS-DE/Pembroke as the way to assess the feasibility and effectiveness of a multi-prong proposal to address the hazard of workplace violence. Id. at *3. When, as is the case here, the proposed abatement is a multi-component process rather than a series of options, the Secretary must show that at least one of the proposed measures was not already implemented and that the same measure is feasible and effective in addressing the hazard. UHS-DE/Pembroke, 2022 WL 774272, at *8; UHS-DE/Suncoast, 2023 WL 2388069, at *3 (upholding citation when parties agreed on the feasibility and effectiveness of six of the eight proposed abatement measures for the hazard of workplace violence). See also BHC, 2019 WL 989734, at *26, 40 (finding the abatement prong met when the Secretary identified six main measures to address deficiencies in the employer’s existing program to address the hazard of workplace violence at a behavioral health hospital); Pepperidge Farm, 17 BNA OSHC at 2033-34 (holding that the Secretary may require an employer to engage in an abatement process, the goal of which is to determine what action or combination of actions will eliminate or materially reduce the hazard).

The measures proposed here include different types of controls. Some are more preventative, and some are more focused on mitigation. (Exs. C-92 at 11; C-94.) Like in UHS-DE/Pembroke and UHS-DE/Suncoast, they are directed at different root causes of the hazard. The issue “is not whether each measure can have a material effect on specific instances of hazards, but instead whether all measures, working together, would prevent a broader universe of hazards.” UHS-DE/Pembroke, 2023 WL 3243988, at *2. See also Nat’l Realty, 489 F.2d at 1266-67 (“all preventable forms and instances of hazardous conduct must … be entirely excluded from the workplace”). The Secretary’s proposed multi-action approach to abating workplace violence aligns with the hazard’s “nature.” 2022 WL 774272, at *8. The hazard “arises in different contexts and conditions,” necessitating multiple abatement measures. Id.; 2023 WL 3243988, at *2 (concluding that the proposed abatement measures for workplace violence were cumulative with each addressing “distinct aspects” of the hazard); United States Postal Serv., No. 16-1813, 2023 WL 2263314, at *2 n.2 (OSHRC Feb. 17, 2023) (concluding that the proposed measures were a comprehensive heat stress safety program rather than alternatives).

In any event, unlike Sturgill, when the Citation was issued, Respondents had not yet implemented the proposed additional feasible measures in the manner the Citation describes. See 2019 WL 1099857, at *9. Respondents sometimes took partial measures related to the proposed abatement but had not yet taken measures that offered the same level of protection as those in the Citation. See Chevron Oil Co., 11 BNA OSHC 1329, 1334, n. 16 (No. 10799, 1983) (employer can develop different solutions if the alternative methods achieve at least as much reduction).

Citing Mid-South Waffles, Inc., No. 12-1022, 2019 WL 990226 (OSHRC Feb. 15, 2019) and A.H. Sturgill, Respondents also contend that the “proposed abatements are not sufficiently specific.” (CS Br. 126-27.) In Mid-South, two Commissioners concluded that the only abatement method was a description of the result the employer had to achieve rather than the specific additional steps it should have taken. 2019 WL 990226, at *6. The issue in Mid-South and A.H. Sturgill was the Secretary’s failure to identify the measures beyond those already implemented. Mid-South, 2019 WL 990226, at *6; Sturgill, 2019 WL 1099857, at *10; UHS-DE/Pembroke, 2022 WL 774272, at *7-8 (distinguishing Sturgill).

Here, Respondents’ abatement was not fully implemented, and the Secretary proposed the specific additional steps they could take to move their program from one “on paper” into one that effectively addressed the hazard to the extent feasible. She sets out additional effective actions Respondents could have taken to reduce the hazard to the extent feasible. UHS-DE/Pembroke and UHS-DE/Suncoast are more apt comparisons than Mid-South or the other cases Respondents cite.

Before getting into the specifics of each abatement, an overview of the effectiveness and feasibility test prong may be helpful. The parties agree that there are no feasible measures to eliminate the hazard. Still, the abatement prong is met when there is an abatement method that would materially reduce the hazard. Arcadian, 20 BNA OSHC at 2011. Reliable expert testimony is sufficient to establish that an abatement method meets that requirement. See Integra, 2019 WL 1142920, at *13-14 (finding that reliable expert testimony is sufficient to establish that an abatement method would materially reduce a hazard, even if the expert cannot quantify the reduction). Respondents note the Secretary’s acknowledgment in discovery responses (Ex. RCS-118) that she did not quantify the reduction of the hazard Respondents should expect to see in implementing the abatement measures. (CS Br. 78). However, the Secretary “need not quantify” the extent to which her abatement and its components would have materially reduced the likelihood of patient-on-staff violence. BHC, 951 F.3d at 565. Evidence of the successful use of a similar approach elsewhere can also establish effectiveness. See Pepperidge Farm, 17 BNA OSHC at 2034; Wheeling-Pittsburgh Steel Corp., 10 BNA OSHC 1242, 1246 n.5 (No. 76-4807, 1981) (finding abatement method feasible when it had previously been used at the cited facility), aff’d, 688 F.2d 828 (3d Cir. 1982)(table).

After showing the proposed abatement would effectively mitigate the hazard, the Secretary must then show that his proposal is feasible, i.e., capable of being done. Baroid Div. of NL Indus., Inc. v. Sec’y of Labor, 660 F.2d 439, 447 (10th Cir. 1981). This requirement comes from caselaw interpreting the OSH Act’s general duty clause. The OSH Act requires each employer to provide a workplace “free from recognized hazards that are causing or are likely to cause death or serious physical harm to his employees.” 29 U.S.C. § 654(a). If multiple abatement measures would cumulatively reduce a recognized hazard, the employer must implement every method (or an equally effective method) until no preventable hazard remains. Nat’l Realty, 489 F.2d at 1266 (Congress intended to require the elimination of recognized and preventable hazards). In other words, employers must continue to act until the workplace is “free” of the hazard. Id., Gen. Dynamics Corp., Quincy Shipbuilding Div. v. OSHRC, 599 F.2d 453, 464 (1st Cir. 1979) (employer must “take all feasible steps”).

However, the requirement to “free” workplaces of recognized hazards cannot be stretched too far. The OSH Act does not permit the imposition of strict liability for any instance of the hazard. Nat’l Realty, 489 F.2d at 1266-67. The statute is limited to hazards that are “serious” and “recognized.” 29 U.S.C. § 654(a). In addition, as interpreted, the Secretary must also show that employers can do something about the hazard. Thus, the Secretary cannot propose an abatement measure that is impossible to implement and still prove a violation of the OSH Act. The proposed abatement must be something employers can put into effect. Baroid, 660 F.2d at 446-47; Acme Energy Servs. v. OSHRC, 542 F. App'x. 356, 367 (5th Cir. 2013) (unpublished) (Secretary only needs to show the abatement is capable of being done). See also Am. Textile Mfrs. Institute, Inc. v. Donovan, 452 U.S. 490, 506-9, 518-22, 540-1 (1981) (equating feasibility with achievability and finding that the Secretary was not required to undertake cost-benefit analysis); Brabham-Parker Lumber Co., 11 BNA OSHC 1201, 1201 (No. 78-6060, 1983) (consolidated) (finding, in the context of a noise standard, feasible means achievable).

Often, there’s significant overlap between the evidence of effectiveness and the feasibility evidence. For instance, the implementation of the abatement elsewhere or post inspection changes can establish both effectiveness and feasibility. Puffer’s Hardware, Inc. v. Donovan, 742 F.2d 12, 19 (1st Cir. 1984) (owners of similar equipment had implemented the proposed abatement); SeaWorld, 748 F.3d at 1215; Modern Drop Forge, Co. v. Sec’y of Labor, 683 F.3d 1105, 1114 (7th Cir. 1982). See also SAIC, 2020 WL 1941193, at *8; UHS-DE/Pembroke, 2022 WL 774272, at *8-12.

Like other workplace violence matters before the Commission, the Secretary’s proposals are grounded in the Guidelines, OSHA Road Map, and the TJC Alert. (Tr. 71, 128, 130, 134, 179, 1608, 1618-19; Exs. C-95, C-94, C-96, C-100.) The Guidelines were developed based on “industry best practices and feedback from stakeholders, and provide recommendations for developing policies and procedures to eliminate or reduce workplace violence in a range of healthcare ... settings.” (Ex. C-95 at 7). BHC, 951 F.3d at 564-66 (Secretary must show that the employer failed to implement measures that “a reasonably prudent employer familiar with the circumstances of the industry” would have taken). The Guidelines include a model WVPP, which is broken down into parts and addresses its application in various settings. (Ex. C-95 at 12-46.) The Guidelines do not provide universal one size fits all recommendations. Because of workplace variations, they direct employers to assess their worksites and implement the recommendations in the manner most effective for their worksites. Id. at 7, 12-46. See BHC, 951 F.3d at 565 (“the Secretary specified a thorough self-evaluation” for the employer “to determine what action or combination of actions will,” in the form of a comprehensive workplace violence prevention program, “eliminate or materially reduce the hazard”).

As First Judge Augustine noted, “the Guidelines’ recommendations are informed by an extensive bibliography of workplace violence studies, including studies by the Center for Disease Control (CDC), international standards associations, state departments of health, multiple publications from Dr. Lipscomb, and the [TJC Alert].” UHS Centennial, 2022 WL 4075583, at *31. OSHA supplemented the Guidelines with the OSHA Road Map, which provides concrete examples of how healthcare facilities have applied the practices the Guidelines called for to reduce the hazard. (Ex. C-96.) Similarly, the TJC Alert identifies factors associated with the hazard and provides recommendations. The identified factors include several with particular relevance to this matter: (1) lack of emergency communication devices, (2) understaffing, (3) staff working in isolation; and (4) inadequate security.⁶³ The TJC Alert’s recommended actions for reducing workplace violence track the OSHA Guidelines. Like the OSHA Guidelines, they call for: (1) management commitment; (2) employee participation; (3) tracking and trending of workplace violence data; (4) identifying source(s) of the hazard and developing a response; and (5) providing training on de-escalation, self-defense, and responding to emergency codes. (Exs. C-94 at 10-12, C-100.)

There is no evidence that Respondents cannot adopt any of the abatement measures the Secretary identified. See BHC, 2019 WL 989734 at *26-40 (finding the following to be feasible and effective methods of abatement at a behavioral health facility:(1) performing a comprehensive evaluation of workplace violence at the facility and developing appropriate policies based upon the evaluation; (2) ensuring that units have appropriate levels of staff given the acuity of the workplace violence hazard; (3) improving procedures for summoning assistance when patients become agitated or violent; (4) improving how incidents of workplace violence are documented and how employees are debriefed after such incidents; (5) having a safety committee obtain input about the hazard from front-line staff; and (6) training); BHC, 951 F.3d at 564-66 (affirming the ALJ’s finding that the Secretary established feasible means to materially reduce the hazard of workplace violence); UHS Centennial, 2022 WL 4075583, at *41-57 (finding employer could have materially reduced the hazard of workplace violence at a behavioral health facility by (1) implementing a comprehensive WVPP, (2) using personal communication devices and/or panic alarms; (3) reconfiguring the nurses’ station: (4) maintaining staffing that is adequate for census and patient acuity; and (5) conducting incident investigations and debriefings); UHS-DE/Pembroke, 2022 WL 774272, at *8-12 (finding that providing panic alarms and equipment for de-escalation were feasible and effective means of abatement for the hazard of workplace violence at a behavioral health facility); UHS-DE/Suncoast, 2023 WL 2388069, at *3 (noting that on review the respondents did not dispute that the following abatement measures constituted feasible and effective methods of abating the hazard of workplace violence at a behavioral health facility: (1) implementation of a comprehensive WVPP, (2) reconfiguration of the nurses' workstations, (3) revising intake procedures, (4) creating a law enforcement liaison position, (5) certain types of staff training, and (6) investigation and debriefing following each incident of workplace violence). Cf. HRI, 2019 WL 989735, at *8-23 (Secretary failed to show that robust and implemented workplace violence prevention program was inadequate for the particular facility).

Respondents appear to assert that the Secretary had to establish the exact cost of the abatement and then obtain Respondents’ financial records to show that they could afford the measures. (CS Br. 111-16.) Such a test is not directed by the OSH Act or in the precedent interpreting the general duty clause. As noted, courts developed the feasibility analysis to ensure strict liability was not imposed. Nat’l Realty, 489 F.2d at 1266-67. While there are cases with expert testimony on an employer’s finances and the abatement’s cost, the Secretary can meet the feasibility prong without the proposal’s exact cost. BHC, 951 F.3d at 558-566, SeaWorld, 748 F.3d at 1215-16; UHS-DE/Pembroke, 2022 WL 774272, at *9-12; SAIC, 2020 WL 1941193 at *8-10. Here, the proposed actions have been “recognized by safety experts as feasible,” and Respondents can implement them. See Beverly, 19 BNA OSHC at 1191; Nat’l Realty, 489 F.2d at 1266 n.37 (“The question is whether a precaution is recognized by safety experts as feasible, not whether the precaution’s use has become customary”); BHC, 951 F.3d at 565 (examining whether a reasonably prudent employer familiar with the industry’s circumstances has protected against the hazard in the manner the citation specified).

Respondents do not counter the Secretary’s evidence that the proposed abatement measures could be put into place without threatening their economic viability. Their citation to Smith Steel Casting Co. v. Brock, 800 F.2d 1329 (5th Cir. 1986) is of little relevance here. (CS Br. 113-15.) Those proceedings concerned an air contaminants standard, not the general duty clause. Smith Steel Casting Co., 15 BNA OSHC 1001, 1002 (No. 80-2069) (consolidated). The employer challenged the process used to promulgate the standard. Id. The matter did not concern assessing whether the abatement proposed is sufficiently feasible such that requiring it will amount to strict liability under the OSH Act. See Nat’l Realty, 489 F.2d at 1266-67 (precautions against hazards are not infeasible for purposes of the general duty clause merely because they are expensive). An employer’s provision of its net profit in a completely different context does not bear on whether such evidence is necessary to establish the feasibility of abatement for violating the general duty clause. Compare Smith Steel, 15 BNA OSHC at 1009 (finding Secretary met the required showings to uphold standard’s validity) with BHC, 951 F.3d at 558-566, SeaWorld, 748 F.3d at 1215-16; UHS-DE/Pembroke, 2022 WL 774272, at *9-12; SAIC, 2020 WL 1941193, at *8-10.

Respondents try to morph the feasibility requirement’s test from one focused on “economic viability” to one about short-term profitability. See Nat’l Realty, 489 F.2d at 1266 n. 37. The Secretary does not deny that there are costs associated with compliance with the OSH Act. There is no requirement that the abatement be cost-free. The issue is whether the costs are so exorbitant that the employer cannot remain in business. Id. Here, there is no evidence implementing the proposed abatement would put Respondents out of business or even cause them to become unprofitable. (Tr. 2003; Exs. C-89, C-90.) Dr. Staggs provided no testimony on the economic feasibility of any measure or Respondents’ ability to implement the proposals. (Tr. 2303; Ex. RCS-95.) Further, he did not contend that the existing abatement was reasonable. In contrast, in Beverly, the employer tried to rebut the Secretary’s evidence of economic feasibility by presenting evidence of the abatement’s cost and its ability to pay. 19 BNA OSHC at 1191. The ALJ failed to resolve the dispute between the Secretary’s expert and the employer’s rebuttal evidence. Id. at 1192. This led to a remand for further analysis of the competing evidence. Id. See also Pelron, 12 BNA OSHC at 1835 (Secretary did not show the existing safety program was inadequate and eliminating the hazard would require the business to cease). The undersigned is not confronted with competing evidence of Respondents’ ability to pay.

The lack of pushback on the evidence of feasibility is unsurprising. Many aspects of the proposed abatement were, or by the time of the hearing had become, part of Respondents’ written policies and procedures. There is no evidence Respondents cannot execute the policies they have adopted. When seeking TJC accreditation or during CDPHE reviews, it does not appear that Respondents claimed it was economically, technologically, or operationally infeasible to abide by their written policies and procedures. (CS Br. 12, 18-20; Stips. 12-16; Tr. 338, 341-42, 1616, 2074, 2676-77, 2905; Ex. C-9.) Other aspects of the proposal, such as reconfiguring the nurses’ station, increasing staff, and revising policies, occurred post-Citation without impinging on the Worksite’s economic viability. See SeaWorld, 748 F.3d at 1215; SAIC, 2020 WL 1941193, at *8-10. (Tr. 1989-90; Exs. C-89, C-90, C-94 at 23.) As Dr. Forman stated, there is “no greater proof that something is feasible” than showing it has already been done. (Ex. C-92 at 23.) The remaining measures were not novel as similar measures had been adopted elsewhere. See Modern Drop, 683 F.3d at 1114; Puffer’s, 742 F.2d at 19; UHS-DE/Suncoast, 2023 WL 2388069, at *3 (noting the measures the employers conceded could be implemented); UHS-DE/Pembroke, 2022 WL 774272, at *1-2, 9-12 (discussing measures in place and additional abatement that was feasible and effective abatement for the workplace violence); UHS Centennial, 2022 WL 4075583, at *13-22, 32-57 (same); BHC, 2019 WL 989734, at *8-40 (same). Moreover, Respondents can sometimes negotiate higher payments from Medicare or Medicaid. (Tr. 2910-11.) Their reimbursement rate is not permanently fixed.⁶⁴ The Secretary showed that a reasonable employer would have been doing more than what Respondents were doing to protect workers. (Exs. C-92, C-94.)

Respondents seek to add new requirements to the Secretary’s burden under the OSH Act. In their view, to comply with past OSH Act interpretations the Secretary had to show that Respondents would remain highly profitable if they implemented the proposed abatement. (CS Br. 112-16.) Such a requirement is not in the OSH Act or relevant precedents. As the Supreme Court explained, “When Congress passed the [OSH Act] … it chose to place pre-eminent value on assuring employees a safe and healthful working environment, limited only by the feasibility of achieving such an environment.” 452 U.S. at 541. The judicially developed feasibility requirement needs to be understood in the context of why the concept was developed. It is a protection against strict liability. Rendering a business less profitable does not result in strict liability under the OSH Act. National Realty was concerned about requiring measures that would eliminate a business. 489 F.2d at 1266 n.37. The Secretary does not have to establish that her proposed abatement would allow an employer to remain highly profitable. It is sufficient to show, as the Secretary did here, that the proposed abatements will not preclude Respondents’ economic viability. See Entergy Corp. v. Riverkeeper, Inc., 556 U.S. 208, 223 (2009) (noting that the OSH Act’s failure to mention cost-benefit analysis led to the Court’s conclusion that the agency was not required to engage in cost-benefit analysis).

The Commission also recognizes an affirmative defense of economic infeasibility. That defense requires the employer to establish that compliance is extremely costly and that it cannot absorb or pass on the cost. See e.g., Faultless Div., Bliss & Laughlin Indus., Inc., 674 F.2d 1177, 1190 (7th Cir. 1992) (fact that control would be “burdensome” was not enough as test is whether “the employer’s existence as an entity is financially imperiled by compliance”). Respondents made no such showing and do not raise the affirmative defense of economic infeasibility in their Answer or briefs.

Respondents also assert certain of the proposed measures create a greater hazard than the measures in place. (CS Br. 88, 101, 117.) The affirmative defense of greater hazard was not raised in the Answer and is not at issue here. When citing a violation of the general duty clause, the Secretary has the initial burden of proving “that an abatement method exists that would provide protection against the cited hazard.” CSA Equip. Co., LLC, No. 12-1287, 2019 WL 1375918, at *9 (OSHRC Mar. 19, 2019). After that, the burden then shifts to the employer to produce evidence showing or tending to show” that the Secretary’s methods “will cause consequences so adverse as to render their use infeasible.” Id.

Respondents misrepresent the evidence in their brief. The Secretary showed that compared to the abatement in place, her proposals would reduce, not increase, the hazard. Dr. Staggs did not assess Respondents’ existing abatement. He also did not conclude the Secretary’s proposals would create a greater hazard. He simply discussed “some” studies suggesting certain types of security personnel were more likely to use unnecessary physical intervention. (Tr. 2220-21.) He referred to a study involving psychiatric units in Germany using people “untrained” in mental health. (Tr. 2220-21.) He also cited a study where the security personnel did not have any positive interactions with patients. (Tr. 2277-78.) He did not state that security increased harm to workers. He did not conclude that a position like the post-inspection Milieu Specialists who responded to patient emergencies but did not conduct rounding was problematic. Id. He did not call into question the discussion of security in the OSHA Road Map and in Dr. Lipscomb’s testimony. He agreed that Respondents could structure the security positions so the individuals interact with patients and still have no care duties. (Tr. 2278.) Dr. Staggs was not an expert in nurses’ stations, and he did not assess the protection offered by the design in place at the time of the inspection. He pointed to no studies that suggested the Secretary’s proposal would create a greater hazard. (Tr. 2282-83, 2885-89, 2292.) Respondents failed to rebut the Secretary’s evidence.

When a citation lists a series of steps to abate a hazard, the Secretary needs to prove that at least one of the measures was not implemented and that this measure is effective at reducing the hazard and feasible. UHS-DE/Pembroke, 2022 WL 774272, at * 8. In other words, if any individual abatement proposal satisfies the requirements, the abatement prong of the test is met. 2022 WL 774272, at * 8; UHS-DE/Suncoast, 2023 WL 2388069, at *3.

In contrast to these recent on-point decisions, to argue the Secretary must establish every proposed abatement would reduce the hazard, Respondents point to Litton Sys., Inc., 10 BNA OSHC 1179 (No. 76-900, 1981) and the Second Circuit decision’s Carlyle Compressor Co. v. Sec’y of Labor, 683 F.2d 673 (2d Cir. 1982).⁶⁵ Litton is consistent with UHS-DE/Pembroke and UHS-DE/Suncoast in stating that the Secretary “must only show that there was a means of abatement which would materially reduce the hazard.” 10 BNA OSHC at 1182 (emphasis added). “The means of abatement, unlike the hazard itself, does not have to be recognized by an employer or the employer’s industry. … The means of abatement must only be ‘feasible.’” Id. (emphasis added, internal citation omitted). Carlyle is also consistent. There, the Secretary proposed a single abatement measure, the installation of a barrier guard. 683 F.2d at 676. The Second Circuit notes that the Secretary must specify the “particular steps” (plural) an employer must take to correct the problem. Id. at 677. But it concludes that showing one action was feasible was sufficient to meet the burden of proof necessary to establish a general duty clause violation. Id. at 678.

Some of the Secretary’s proposals were subsequently implemented at the Worksite and others had been successfully implemented at similar worksites. Expert testimony supported each of the actions and the overall package of measures. See Nat’l Realty, 489 F.2d at 1266 n.37 (recognition by safety experts as feasible meets the required showing for a violation of the general duty clause). Respondents did not rebut the Secretary’s evidence of the feasibility and effectiveness of the abatement prong. Dr. Staggs acknowledged things could be done to prevent and mitigate the effects of inpatient violence. (Ex. RCS-95 at 4.) Although he concluded that certain of the Secretary’s proposals would not be effective at mitigating the hazard, he offered no opinion on others. (Tr. 2302-3; Ex. RCS-95.) He reached no conclusions as to any proposal’s feasibility. (Tr. 2303; Ex. RCS-95.)

Though establishing one feasible method of abatement that will materially reduce the hazard is sufficient, the undersigned will discuss each proposal.

Engineering Controls

Re-configure the nurses’ stations in the units to include design features that prevent patients from jumping over, reaching into, over, or otherwise entering into the workstations. Ensure items in the workstations, such as but not limited to hole punchers, staplers, telephones, cords, pens, computers, computer peripherals, and other items are not accessible by the patients, so they cannot be used as weapons.

The design of the nurses’ stations varied by unit. (See Section II.B.1.b.) Contrary to Respondents’ position, the Secretary did not have to establish that the nurses ‘station was the “most common” location for injuries to show this abatement method was feasible. (CS Br. 87.) They cite no precedent for this argument.

What the Secretary did show is that there were many worker injuries and patients could weaponize many items needed at the nurses’ station. (Tr. 993, 1643-45; Exs. C-92 at 23, C-94 at 22.) See UHS-DE/Suncoast, 2023 WL 2388069, at *70-71 (patient used scissors from nursing station to stab worker). On numerous occasions aggressive patients entered or tried to enter the nurses’ stations. (Tr. 94-95, 628, 859-60, 960, 992-93, 1024-25, 1028-29, 1254, 1489, 1850-52; Exs. C-2b at 10, C-19, C-20, C-92 at 22-23.) Several such incidents resulted in workplace violence.⁶⁶ Other evidence, such as incident reports and worker testimony, indicates that injuries at the nurses’ station were more common than Respondents claim. (Tr. 1643-45; Exs. C-92 at 21-22; C-94 at 22.) The record and the expert opinions show that the nurses’ stations, as they were previously configured, provided inadequate protection.

Dr. Gray, the Medical Director, acknowledged he was not an expert on the topic but noted a “general trend” of removing nurses’ stations to the extent possible. (Tr. 2635, 2668-69.) In contrast, Dr. Forman concluded that the design of the nurses’ stations was inadequate in light of the typical patient acuity. (Tr. 1254; Ex. C-92 at 22-23.) He opined that reconfiguring the nurses’ station would be feasible and effectively reduce the hazard. (Tr. 1255-56, 1258; Ex. C-92 at 22-23.) As support, he discussed other facilities with such abatement and Respondents’ successful post-Citation changes. (Tr. 1255-56; Ex. C-92 at 23.) See SeaWorld, 748 F.3d at 1215 (extending actions taken with one animal to all similar work was feasible); Con Agra, Inc., 11 BNA OSHC 1141, 1144-45 (No. 79-1146, 1983) (finding extension of existing abatement feasible). Dr. Forman had experience with fully enclosed nurses’ stations at inpatient psychiatric facilities. (Tr. 1256.) He opined that making it more difficult for patients to access the interior of the nurses’ stations did not negatively impact the therapeutic environment. (Tr. 1256-57; Ex. C-92 at 22-23.)

Dr. Lipscomb concurred that the nurses’ stations were inadequate, and that the proposal was feasible and effective. (Tr. 1643-47; Ex. C-94 at 22-24.) Preventing patients from entering the nurses’ station increased staff safety and other hospitals had implemented similar abatement. (Tr. 412-13, 1050, 1489-90, 1643, 1646-47, 2702-4; Ex. C-94 at 23.) See UHS-DE/Suncoast, 2023 WL 2388069, at *95; Pepperidge Farm, 17 BNA OSHC at 2034 (viewing “successful use of a similar approach elsewhere” and expert testimony as elements of an effective abatement method). The CO also asserted that re-configuring the nurses’ stations would reduce the likelihood of injuries from direct attacks or when patients gain access to items kept at the nurses’ station. (Ex. C-3 at 55.)

Dr. Staggs offered no opinion on the proposal’s economic feasibility, but his report touched on other aspects of feasibility. (Tr. 2303; Ex. RCS-95 at 9.) Dr. Staggs was not accepted as an expert who could opine on the feasibility of this measure. He is not a licensed psychiatrist and has never worked in or around a nurses’ station. He has not published articles or studies on nurses’ stations. (Tr. 2119-20.) His expertise lies in assessing the quantitative data’s quality in various studies. In that area, he noted three studies on nurses’ stations, whose data he considered to be of “very low quality.” (Tr. 2227-33, 2288-89.)

More persuasive than the studies Dr. Staggs discussed, is the expert testimony of Doctors Forman and Lipscomb and actual experience at this Worksite. (Tr. 1050, 1256-57, 1489-91, 1643-47, 2670; Exs. C-2b at 10, C-92, C-94 at 23.) Workers discussed their experience with how nurses’ stations that patients could not enter improved safety. (Tr. 412-13, 1489-90.) A former nurse explained how enclosed workstations at her current position prevent patients from getting inside. (Tr. 412-13.) Dr. Staggs agreed he would take seriously the workers’ assessments on how design changes for nursing stations affected the Worksite. (Tr. 2289-90.)

The post-Citation changes adopted at the Worksite were well received by staff. (Tr. 1256-57, 2703-4.) Barriers extending up to the ceiling were added, and there are full doors workers can lock. (Tr. 879-81, 1647, 2669-70; Ex. C-94 at 23.) The new enclosures have been helpful and enclosed nurses’ stations would have prevented Patient 23 from entering the area as occurred in January 2020. (Tr. 880-81, 946-47.) The changes provided worker protection without resulting in patient care problems. (Tr. 880-82, 1050, 1489-90, 1647, 2670, 2702-4; Exs. C-92 at 23, C-94 at 23.)

The post-Citation changes to have more protective nursing stations could have been implemented before OSHA’s inspection. Around the fall of 2019, locks were added to the doors of the area where the doctors worked on the Cottonwood unit. (Tr. 1508-9.) This prevented patients from accessing the area. (Tr. 1510.) The change was made after a patient flipped a worker over when the doctor attempted to place him in a hold. (Tr. 1509.) See SeaWorld, 748 F.3d at 1215 (possible to extend abatement measure); Con Agra, 11 BNA OSHC at 1144-45.

Other behavioral health facilities had more enclosed nursing stations.⁶⁷ Like at this Worksite, UHS-DE/Suncoast involved behavioral health patients entering the nurses’ stations. 2023 WL 2388069, at *70. Sometimes, this resulted in staff injuries. Id. In one instance, a psychiatric patient was able to reach into the nurses’ station and get a pair of scissors. Id. at *71. The patient then stabbed a nurse with the scissors. Id. The Secretary proposed a similar abatement method.⁶⁸ The ALJ rejected the argument that patient care precluded protective nurses’ stations. Id. at *94. Relying on expert testimony and the partial implementation of the proposal, the ALJ found reconfiguring the nurses’ stations to limit entry to be a feasible and effective means of abatement for the hazard of workplace violence.⁶⁹

Another ALJ reached a similar conclusion for a behavioral health facility in Colorado. There, “the nurses’ station was a locus for patient aggression,” and the layout did not offer effective protection. 2022 WL 4075583, at *5, 42. The Secretary proposed reconfiguring the nurses’ stations as abatement. Id. at *1, 5, 42. Employees who had been attacked at the nurses’ station believed a more substantial barrier would have prevented certain incidents, and the experts agreed. Id. at *43-44. The ALJ found that the proposed abatement was a feasible and effective means to mitigate workplace violence. Id. at *44.

The record and the expert opinions of Doctors Forman and Lipscomb showed that re-designing the nursing stations was feasible and provided material additional protection from the cited hazard. See 2022 WL 4075583, at *44; 2023 WL 2388069, at *94-95.

The next aspect of the proposed abatement relates to providing reliable communication devices so staff can convey a need for assistance:

Provide all staff members who work in close proximity with patients reliable communication devices to rapidly communicate need for assistance, including but not limited to times when staff members leave the unit to escort patients. Provide a means to silently communicate need for assistance without alerting the patient population. Install the alert signal in a way that is not visible to patients. In addition, provide training and procedures on the use and limitations of equipment.

The primary ways of requesting assistance were accessing the phone in the nurses’ station to use the overhead paging system or yelling. Workers typically had a radio when going between the buildings or in areas such as the gym or cafeteria, but at times there were not enough working devices to permit this. (Tr. 137-44; Ex. C-42.) The proposal is a feasible and effective method to address the deficiencies in the existing system. (Tr. 140-44; Ex. C-3 at 55.)

Dr. Staggs reviewed the scientific literature on workplace violence and claimed no studies showed that a particular device materially reduces workplace violence. (Tr. 2236.) He then clarified that one of the studies discussed in his report found that communication devices such as cell phones and personal duress alarms were a protective factor against violence. (Tr. 2299-2300.)

Dr. Staggs had no experience with radios or panic alarms and had not conducted any research on communication devices or alarms. (Tr. 2137-41.) He offered no opinion on the method’s feasibility or the suitability of Respondents’ approach at the time of the inspection. (Tr. 2293, 2302-3.) Like with nurses’ stations, Dr. Staggs’ report goes beyond the areas in which he was qualified as an expert. (Ex. RCS-95.) For instance, his report notes that he did not see evidence of insufficiencies in the quantity or quality of the radios. (Ex. RCS-95; Tr. 2302.) In addition to being beyond an evaluation of the research of others, he based this assessment on the post-inspection status of communication devices, not the situation before Respondents purchased over 100 new devices. (Tr. 141, 882, 2302, 2555-56; Ex. RCS-43.)

Dr. Lipscomb had a background in workplace violence prevention in the healthcare setting, and her opinion on the feasibility and effectiveness of this proposal is credited. (Tr. 1650-51, 1653-54.) She asserted that having a reliable and readily available communication device will reduce the likelihood of injuries from attacks. (Tr. 1650-53; Ex. C-94 at 26.) In terms of the need to silently communicate assistance, witnesses explained that patients would often escalate in response to audible Code Greens. Sometimes this was because they realized there would be less staff in their unit as workers moved to the unit from which the call came. (Tr. 142-43, 1518, 1651.) Respondents also realized there were areas without ready access to communication devices. A site survey conducted months before OSHA’s investigation identified a lack of panic buttons at the Worksite and directed senior management to “consider panic buttons in areas that do not have phones.” (Tr. 1870-71; Ex. C-78 at 2.) As for feasibility, according to Dr. Lipscomb, having reliable communication devices with the capability to ask for help silently was “commonplace” in Respondents’ industry and identified in the OSHA Guidelines and other sources as an engineering control to address the hazard. (Tr. 1650-54; Exs. C-94 at 19, 26, 33, C-95 at 10, 25, 32.)

Dr. Forman agreed that this proposal was feasible and effective. At an adequately staffed Worksite, providing staff with reliable communication devices would permit faster and more efficient responses to requests for assistance when dealing with aggressive patients and violence. (Tr. 1200-1; Ex. C-92 at 11-12.) The ability to summon help more quickly could lessen the duration and possibly the intensity of an assault. (Tr. 1200-1; Ex. C-92 at 12.) Communication devices permit the effective communication of relevant information needed to respond to an event appropriately. (Tr. 1196-97.) Dr. Forman explained the need for multiple effective methods of communication. (Tr. 1301-10, 1394.) Other similar workplaces utilize multiple methods to summon assistance depending on the situation. (Ex. C-92 at 12.)

Other evidence supports the conclusions of the Secretary’s experts. Witnesses addressed how having one communication device for each staff member would improve staff safety during incidents of workplace violence. (Tr. 137-144, 691, 794, 1516-17.) A former nurse explained how having a button to contact security immediately was more efficient than the methods of obtaining assistance at the Worksite. (Tr. 1516-17.) The abatement would improve reliability and provide more ways to summon assistance, including silently. (Tr. 142-43.) For example, therapists did not have radios and did not work from the nurses’ station. (Tr. 1414; Ex. C-94 at 25.) When confronted with workplace violence, a therapist described needing to rely on someone else hearing her and then having that person obtain further assistance. (Tr. 1432-33.) Several witnesses discussed other healthcare facilities with systems similar to the proposal for summoning assistance. (Tr. 401-2, 506, 508-9, 691, 693, 1200-1, 1516-17; Exs. C-43, C-92 at 12, C-94 at 26.)

The D.C. Circuit and the Commission found that providing reliable communication devices was a feasible and effective abatement method for workplace violence at behavioral health facilities. In BHC, the D.C. Circuit upheld the ALJ’s conclusion that the telephones and walkie-talkies used to summon assistance during incidents of aggression at a behavioral health hospital were inadequate. 951 F.3d at 564-65. Providing reliable and readily available means of communication was feasible and effective abatement for the hazard.⁷⁰ Similarly, in UHS-DE/Pembroke, the Commission concluded that providing personal panic alarms for all employees would be a feasible and effective means to materially reduce the hazard of workplace violence. 2022 WL 774272 at *9.

In UHS Centennial, the ALJ found that “implementing the use of walkie-talkies and/or personal panic alarms would be feasible and effective” at abating the hazard of workplace violence at a behavioral health facility. 2022 WL 4075583, at *42. Employees already had some walkie-talkies at the time of the inspection and, later, the employer increased the number of devices such that each person could have one. Id. The OSHA Guidelines and expert testimony indicated that “the use of walkie-talkies and/or panic alarms would be effective in reducing the likelihood or severity of the injury resulting from an act of patient aggression.”⁷¹

Akin to UHS Centennial, many aspects of the proposed abatement were implemented after OSHA’s investigation commenced. (Exs. C-92 at 12; C-94 at 36.) Respondents purchased more radios for the Worksite. (Tr. 141, 882-83, 2447, 2555-56; Ex. RCS-43.) The additional radios ensured every MHS working could be assigned their own radio.⁷² Further, there is no evidence that panic buttons could not be implemented in additional locations beyond the A&R department.⁷³

The Secretary’s experts opined that this method was feasible and effective. (Tr. 1200-1, 1301-2, 1650-53; Exs. C-92 at 11-12; C-94 at 25-26.) This type of abatement is widely used throughout healthcare, including in the context of behavioral health. (Exs. C-92 at 12; C-94 at 36.) Hence, as was the case in BHC, UHS-DE/Pembroke, UHS-DE/Suncoast, and UHS Centennial, the Secretary showed that ensuring all staff members working closely with patients have reliable communication devices is a feasible and effective means of abatement.

Administrative Controls

During OSHA’s investigation, incidents laid plain the Secretary’s concerns with Respondents’ approach to contraband. The proposed abatement includes several specific actions to mitigate the hazard by improving methods for identifying and securing contraband:

Develop and implement a policy in the [A&R] unit to secure potential patient belongings upon arrival for assessment, such as providing patients with a locker where they can secure personal belongings. Minimize the length of time patients spend in common areas with potential contraband, including weapons. Inspect patient belongings in a secure area prior to escorting patient to the unit. Develop and implement procedures for confiscating items in a manner which reduces or eliminates risk to A&R employees. Consider implementing a policy where local law enforcement is contacted to secure contraband such as illegal substances or weapons. Develop a means for A&R employees to keep their hands free and unobstructed while escorting patients to the unit.

Dr. Lipscomb opined that this abatement method was feasible and effective. (Tr. 1670, 1673-74; Ex. C-94 at 36.) Dr. Forman agreed that modifying how workers screened and secured patients’ belongings would effectively improve staff safety. (Tr. 1314-15; Ex. C-92 at 13-14.) As he explained, the “front door,” i.e., upon arrival, is when contraband should be secured. (Ex. C-92 at 12.) He indicated that this is what occurs at the facility where he works. Id. at 13. The CO noted that securing patient belongings earlier in the admissions process reduces patient access to weapons and reduces the risk of injury. (Ex. C-3 at 57-58.) Dr. Staggs offered no opinion on the proposal’s feasibility or effectiveness.

In addition to the expert opinions, more robust procedures at other similar facilities and the changes implemented after the inspection support finding this method to be feasible and effective. Witnesses described screening procedures used at other psychiatric facilities that provided more protection, such as security personnel screening all belongings and screenings conducted off-unit. (Tr. 701, 780-81; Ex. C-92 at 12-13, C-94 at 23-26.)

Respondents do not counter the expert opinions and other evidence of feasibility and effectiveness. Respondents note regulations impose requirements on handling patient belongings. (CS Br. 104.) Still, they did not rebut the evidence that implementing either the proposed abatement or something equally effective could be done and would improve worker protection from the cited hazard. (Tr. 1400-3.)

There is no requirement compelling patient access to contraband. Respondents do not contend that their training, which directed that inventories would occur “upon arrival” was inappropriate, let alone illegal. (Tr. 2892; Ex. RCS-30 at 2.) Respondents also do not contend that their post inspection changes to contraband screening and control violated any other applicable regulation. After the knife incident, staff no longer returned belongings to the patients after admission. (Tr. 902-3, 2430.) The location of the belongings searches for the Hickory unit changed from the patient’s bedroom to a separate room outside of the unit. (Tr. 2320.) Under the new approach, staff searched patients and their belongings with the more effective equipment before the patient is brought to a care unit. (Tr. 902-3, 1314, 1964, 2428, 2430; Exs. C-3 at 29, C-39.) Respondents also equipped the units with metal detecting wands to use on patient belongings that had not been processed in the Evergreen building. (Tr. 2428, 2430; Ex. C-39.)

The Secretary proposed that Respondents “implement the Workplace Violence Prevention Program.” Essentially, the Secretary calls on Respondents to properly implement the procedures identified in their WVPP and related policies. Briefly, the proposal calls for addressing several key deficiencies related to: (i) worker participation and the WVPP’s accessibility, (ii) reviewing and updating the WVPP, (iii) identifying hazards, and (iv) incident reporting, investigations, and debriefings. (Tr. 144-57; Ex. C-3 at 55-56, 58-59.)

The Secretary further elaborated on how, as part of the workplace violence prevention program, incidents of workplace violence should be investigated and debriefed:

Conduct an investigation and debriefing after each act of workplace violence, including near misses, with the attacked and/or injured employee and other involved employees, including root cause or similar analysis, lessons learned, and corrective actions to prevent re-occurrence. Maintain accurate records of patient assault on staff. Provide the attacked and/or injured employee and other involved employees an opportunity to provide feedback about specific measures that could prevent such future incidents. Review and evaluate each workplace violence related incident, both on a case-by-case basis and to monitor for trends in areas with high rates of incidents such as the acute units. Thoughtfully review and evaluate any formal or informal requests for increased staffing based on unit acuity. Ensure that formal and informal requests are valued, acknowledged, reviewed and evaluated, and the resultant outcome and actions are communicated in the most effectual forum.

The Citation numbers these proposals as 3 and 7. (Ex. C-1.) Item 7 is a more fulsome discussion of Item 3’s directive to address the deficiencies in the existing policies concerning incident debriefings.⁷⁴

On paper, the Worksite had elements of a workplace violence prevention program. But Respondents neglected to implement a “comprehensive and coherent approach to preventing patient-on-staff violence.” (Tr. 144-45, Ex. C-94 at 14.) See BHC, 951 F.3d at 565 (finding that a “comprehensively considered and applied program would materially reduce the hazard” of workplace violence). Neither individually nor collectively did the Worksite’s policies and procedures set out a “comprehensive, population-based approach to patient on staff violence.” (Tr. 144-57; Exs. C-3 at 32-54, C-92, C-94 at 16.)

Dr. Staggs offered no opinion on the WVPP’s content, communication, or implementation. (Tr. 2302-3; Ex. C-95.) He did not opine on the feasibility and effectiveness of the proposed changes. Id. In contrast, Dr. Lipscomb asserted that implementing this abatement was an effective component of reducing the hazard. (Tr. 1606-8; Ex. C-94 at 21.) Based on her experience and as identified in certain studies, implementing a comprehensive WVPP with a strong management commitment and employee involvement is effective at materially reducing workplace violence. (Tr. 1609-1615, 1621; Exs. C-94 at 20-21, C-95 at 12-14, C-97, C-98, C-100.) In terms of feasibility, she noted the ability of other similar facilities to adopt and implement such plans. (Exs. C-94 at 20-21, C-96.) The proposal’s feasibility is also reflected by the fact that it largely calls for Respondents to implement their policies and procedures.

The Risk Management Director discussed the value and need for direct care workers to be involved in abating the hazard. (Tr. 2786). Respondents recognized this as a critical component of mitigating the hazard. (CS Br. 42; Exs. C-95 at 12-13; C-96 at 9-11.) But reality diverged from what Respondents claimed they were doing to lessen the hazard. (Stips. 43-44; Tr. 149-52, 154-57, 1010, 1456-57, 1504-5; Exs. C-3, C-38, C-69, C-79.) The WVPP did not incorporate genuine employee participation. (Tr. 1606-7, 1621-24, 1633-34; Exs. C-3, C-94 at 14, 18-19, 21, 43.) Workers were not consistently aware of the Worksite’s WVPP or how to access it. (Tr. 155-56; Exs. C-3, C-94 at 20.)

Dr. Lipscomb and the CO addressed how appropriately incorporating these measures was feasible and would effectively reduce the hazard. (Tr. 149-50, 1613; Exs. C-3 a 56-57, C-94, C-98.) Respondents do not claim it would be infeasible or ineffective to follow their WVPP’s requirement to have employee participation or improve access to the WVPP. (Tr. 155-56, 1623-24; CS Br. 93-94.) See HRI, 2019 WL 989735, at *16-20 (behavioral health services provider recognized the need for employee input in workplace violence prevention and provided numerous ways to provide feedback that was then acted upon).

Respondents contend it was their policy to review the WVPP annually, with the reviews focused on the causes of employee injuries from workplace violence. (Exs. C-22 at 7-8, C-78 at 2.) They identified this as a means to reduce the hazard at their Worksite. Id. Yet, this data was neither adequately collected nor reviewed. In terms of the frequency of review, Respondents scheduled it for review every two years, as opposed to the annual review set out in the WVPP itself. (Tr. 1793; Exs. C-3, C-22 at 1, 7; C-78 at 2; C-94 at 20.) Even so, bi-annual reviews had not occurred. Id.

Routine review and monitoring assist with implementing the plan and facilitate incorporating the lessons learned through assessing past incidents.⁷⁵ The issue is how Respondents used the findings to mitigate the violence workers experienced. Unfortunately, the root cause analyses for employee injuries and corrective action plans contemplated by the WVPP were not employed effectively to mitigate the hazard. (Exs. C-22 at 7-8; C-94 at 20.)

Incorporating program assessment and updating would reduce the hazard and was feasible. (Ex. C-94 at 20-21.) Other facilities have done it, and the abatement correlated with significant reductions in workplace violence related injuries. (Exs. C-94 at 20-21, C-96.)

Turning to feasibility, besides the evidence of other facilities, there’s also Respondents’ claims about the Worksite’s WVPP. Respondents claimed the WVPP was updated in 2020. (CS Br. n. 103.) The testimony they cite does not clearly support the contention. (Tr. 1824.) Nevertheless, the parties agree that the WVPP was reviewed and revised the following year, in 2021. (Ex. C-94 at 20.) This shows that more frequent reviews and updates were feasible.

As part of implementing a workplace violence prevention program, the Secretary proposes Respondents “conduct a building-specific hazard analysis that analyzes the building as a whole.” Expert testimony supports the method’s feasibility and effectiveness. (Ex. C-94 at 19-20.) Assessments will prevent worker injury and reduce re-occurrence of similar events. (Ex. C-3 at 56-57.)

Post-inspection actions support these conclusions. In April 2020, after the inspection commenced, a UHS-DE employee, Eric Wang, sent Ms. Kolln detailed analyses of patient aggression and other criteria to use during future meetings of the PIC and PSC. (Ex. C-49.) Mr. Wang directed Ms. Kolln to update and use the data on a going-forward basis. Id. He explained that analyzing detailed data should be part of the Worksite’s overall strategy of “reducing patient aggression and restrictive intervention.” Id.

According to Ms. Kolln, the data provided was the same information she previously used. (Tr. 2808.) However, there is no evidence that the type of tracking and trending data Mr. Wang provided in April 2020 had been discussed during any previous committee meetings. (Ex. C-3 at 36-37, 51.) Ms. Kolln may have had access to such information, but there is no evidence it was presented or discussed in the detail and scope shown in the April 2020 charts put together by Mr. Wang. (Exs. RCS-84, C-75, C-76.) That tracking and trending information was far more robust than the information shared in past meetings.

There is no debate on feasibility. Respondents were able to implement more robust hazard assessment and analysis procedures. (Ex. C-49.) Likewise, Mr. Wang’s email supports Dr. Lipscomb’s testimony on its effectiveness. (Exs. C-49, C-94.) See BHC, 951 F.3d at 565 (noting employer’s flawed process for tracking and addressing patient-on-staff violence and upholding the Secretary’s proposed abatement in the form of a comprehensive workplace violence prevention program).

The proposed abatement requires several changes to reporting, debriefing, and investigating workplace violence. It calls for implementing “an effective reporting process” and for participants in workplace violence incidents to be debriefed. (Exs. C-1 at 8, C-3 at 56, 58-59.) Debriefings are to include all involved staff and the review of available camera footage. (Ex. C-3 at 56.) The findings should be analyzed with appropriate changes implemented. Id. at 56, 68-59. The results of the debriefings and investigations are to be used to limit the number and severity of future workplace violence incidents. Id. at 56-59.

Notably, the proposal aligns with the existing WVPP and what management previously identified as important. (Tr. 144-45; Exs. C-3 at 56, C-22; C-79 at 8, C-92 at 19, 24, C-94 at 41.) But Respondents’ protocols were less robust and inconsistently followed. (Exs. C-22, C-94.) See BHC, 951 F.3d at 565 (Secretary’s proposed program “would more effectively and consistently apply measures designed to reduce patient-on-staff violence”). The WVPP in place during the inspection was sparse, with many topics addressed with brief phrases and bullet points. (See Ex. C-22 at 3-4.) Insufficient attention was paid to investigating and analyzing employee injuries from the hazard. (Exs. C-92 at 18-20, C-94.) Restraint and seclusion data was reviewed. However, analysis of employee injuries from aggression did not occur to the same degree. (Ex. C-92 at 18-19.)

To address the shortcomings of how Respondents investigated incidents and used the findings, the Secretary proposes revisions to their procedures to “conduct an investigation and debriefing after each act of workplace violence, including near misses.” (Ex. C-1.) These investigations should include analysis and result in “corrective actions to prevent re-occurrence.” Id. The debriefings should: “provide the attacked and/or injured employee and other involved employees an opportunity to provide feedback about specific measures that could prevent such future incidents.” Id. Each workplace violence related incident should be reviewed and evaluated “to monitor for trends in areas with high rates of incidents such as the acute units.” Id. Respondents should: “maintain accurate records of patient assault on staff.” Id.

Dr. Forman explained how adequate post-incident debriefing would enhance worker safety. (Ex. C-92 at 18-19.) He characterized investigating and debriefing all workplace violence as “essential.” Id. Dr. Lipscomb agreed, describing an appropriate debriefing process as “critically important.” Id. at 42. The proposed process allows informed decisions about the corrective actions to prevent future staff injuries from workplace violence. Id. Dr. Staggs did not address the WVPP’s effectiveness or this method’s feasibility. (Ex. C-95.)

The WVPP states that analyzing aggression events permits the identification of trends and the implementation of appropriate corrective action. (Ex. C-22 at 3.) Respondents appear to acknowledge that post-incident debriefings were inconsistent. (CS Br. 110.) After the Citation’s issuance, they revised the debriefing and incident investigation process. Id. at 40. The changes included conducting video reviews with the staff involved in the incidents of workplace violence. (Tr. 883-84; Ex. RCS-41.) The forms also changed to allow for more input from staff and leadership. (Tr. 888.) The changes improved worker safety. (Tr. 884-88.) In 2021, the Restraint Reduction Initiative also included improved debriefing as part of its strategies to reduce the number of restraints or seclusions. (Tr. 2613.) Those changes reduced the number of restraint and seclusion events, which Respondents maintain correlates with improving staff safety. (Tr. 888; CS Br. 36.)

Respondents argue that the Secretary’s use of the phrase “near miss” in this proposal is unduly vague. (CS Br. 125.) However, Respondents used the same term. Workers were expected to report “near miss” incidents. (Tr. 320-23, 1238-39, 1243; CS Br. 109.) And, they had a program to recognize when a worker made a “good catch,” which was a “near miss” or something that “could have been a risk.” (Tr. 323, 1858, 2777-79; Ex. C-22 at 2.) As used in the proposal, the term is not unduly vague.

Respondents then pivot to assert that a near miss of an incident is not cognizable under the general duty clause. (CS Br. 110.) First, the Secretary is not citing Respondents for near misses. Actual incidents of workplace violence, including those resulting in serious injuries, occurred routinely. Respondents were aware of the hazard and knew workers suffered injuries from it at the Worksite. Second, the Secretary is arguing that to reduce incidents of the hazard, a feasible and effective component of abatement is to examine incidents in which serious injuries or death nearly occurred. Respondents already required documentation of at least four “good catches” or near misses per month. (Tr. 323, 1858, 2778-79, 2882.) What the Secretary is proposing is for Respondents to address the inconsistency in which incidents were reported and investigated. Reports of serious safety issues resulted in little follow-up or action. (Tr. 1009-10, 1036-37, 1438; Exs. C-2b, C-30, C-31, C-33, C-34, C-38, C-42, C-43, C-52, C-53, C-54, C-55, C-56, C-57, C-58, C-59, C-61, C-62, C-63, C-72.)

The Secretary showed that proposed abatement methods 3 and 7 would be feasible and effective at materially reducing the hazard.

Maintain staffing that is adequate to safely address changes in patient acuity and patient census. Staffing levels must allow for safety of staff during admission of new patients, behavioral health emergencies, one-on-one patient assignments, staff breaks, and the accompaniment of patients off-unit. Staffing levels must also allow for and ensure safety during educational instruction on and off the unit, therapeutic activity groups, and recreational periods. …

Designate specific staff with specialized training in security and/or hire trained security specialists to monitor patients for potential aggression on all shifts and to assist in preventing and responding to Code Green events occurring in the units. Staff must have skill necessary to re-frame issues, seeking how to keep the patient and persons in the surrounding area safe, while responding to aggressive behaviors. The staff designated to monitor and respond to patient aggression should not be given other assignments, such as patient rounds, which would prevent the designated person from immediately responding to an alarm or other notification of a Code Green. Conduct Code Green training sessions to allow all designated staff to practice and evaluate their skills in a variety of environments and scenarios.

This proposal is not that different from what Respondents claimed occurred. Consistent with the other proposed measures, the issue relates more to implementation. See BHC, 951 F.3d at 565 (“incomplete and inconsistently implemented safety protocols were inadequate to materially reduce the hazard posed by patient-on-staff violence”). The proposal provides an approach to address the gap between what Respondents’ policies and programs called for, and how the hazard was addressed in practice. See Gen. Dynamics, 599 F.2d at 464 (employers must “take all feasible steps” to protect against recognized hazards and implement every abatement measure “whenever it is recognized by safety experts as feasible, even though it is not of general usage in the industry”); Arcadian, 20 BNA OSHC at 2008 (violation established when employer failed to free workplace of a recognized hazard and there are feasible methods to abate or materially reduce the hazard).

The proposed abatement does not provide the number of workers necessary to protect patients or to address particular acuity scenarios. The Secretary makes no claims about how many workers Respondents need to provide medical care. Her argument is that frequently there was not enough staff to implement abatement and adequately protect workers from serious workplace violence injuries. (Sec’y Br. 104-5.)

Dr. Staggs agreed that there are demonstrated “links between nurse staffing and various adverse events in hospitals.” (Tr. 2111.) And he concluded that, “adequate staffing levels are crucial for patient and staff safety.” (Ex. RCS-109 at 6.) Still, more nursing staff does not necessarily result in a safer workplace.⁷⁷

Dr. Staggs’ research and the other studies he relied on have the classic chicken or egg, which came first, problem. He could not draw any conclusions about the direction of causality. (Tr. 2188-89, 2295; Exs. RCS-109 at 1-2, 6; RCS-113 at 10.) Facilities treating more violence prone patients tend to have more staff. (Tr. 2295; Exs. RCS-109 at 5, RCS-113 at 9-10.) And, as violence increases, facilities may add more staff. (Exs. RCS-109 at 2; RCS-113 at 10.) Besides the missing causal link, he acknowledged other limitations in the research. The studies did not consider patient diagnoses. The units studied were not representative of all U.S. psychiatric units. In addition, the nature of the field precludes the use of various controls.⁷⁸

At the hearing, Dr. Staggs suggested that the Worksite’s staff level was not an “outlier.” (Tr. 2196.) Neither this conclusion nor any supporting data are in his report, and this conclusion is rejected. (Ex. RCS-95.) He did not recall reviewing any information about the Worksite’s rate of assaults by patients against staff. (Tr. 2196, 2263.) Such information was available. (Tr. 1700; Exs. C-3 at 51, C-94 at 14.) Dr. Staggs also may have had access to benchmarking and research information on patient assaults and staffing. (Tr. 1700, 1763, 2101-5, 2108, 2190; Ex. C-94 at 14.) He developed a database that was part of the National Database of Nursing Quality Indicators (“NDNQI”). (Tr. 1761, 2106, 2114-15, 2123; Ex. RCS-109 at 1.) He did not review the NDNQI to assess or compare the Worksite to other similar facilities.⁷⁹ In any event, employers must “take all feasible steps” to protect against recognized hazards and implement every abatement measure “whenever it is recognized by safety experts as feasible, even though it is not of general usage in the industry.” Gen. Dynamics, 599 F.2d at 464.

Dr. Staggs did not assess the adequacy of the Worksite’s staffing. (Tr. 2259-60.) Respondents claimed, without citation, that there was “no consensus as to what constitutes adequate staffing in an inpatient psychiatric unit. (CS Br. 109.) However, Doctors Lipscomb and Forman opined on the adequacy of Respondents’ approach and this proposal’s effectiveness. They acknowledged that some research on staffing is equivocal, but persuasively addressed significant agreement on other aspects of staffing. There is consensus that if staffing levels necessitate a worker being left alone with aggressive patients, particularly with the frequency that occurred at the Worksite, the staffing is not adequately protective for workers. (Tr. 1667-68; Exs. C-92, C-94 at 28, 31-32.) Likewise, there is agreement that staffing levels need to change in response to acuity. In practice, often, there were not enough available on-site workers to prevent and mitigate workplace violence. (Exs. C-92, C-94.)

Worksite conditions bear out Doctors Lipscomb and Forman’s position. Understaffing made it more difficult to adequately protect workers during patient aggression events.⁸⁰ It can cause a “cascading” series of workplace violence hazards. (Ex. C-92 at 14.) Understaffing leads to injuries, which demoralize staff, which leads to attrition, further exacerbating the root problem of understaffing. (Tr. 1220-21, 2610; Exs. C-92 at 14, RCS-41 at 7.)

Workers asserted that maintaining sufficient staffing would reduce the hazard. A nurse described an event where a patient placed an MHS in a headlock. (Tr. 1506-7.) Another patient had to intervene before staff were able to arrive. Id. The nursing supervisor explained that having more staff would have helped in this situation. (Tr. 1508.) In another instance, multiple patients behaved aggressively. (Tr. 753-57, 759-60.) Staff could not contain the situation, and a nurse was injured. (Tr. 760-65, 767.) Adequate staffing could have prevented her injury. (Tr. 771-72.) In another matter, patient behavior escalated into what a former lead MHS characterized as a riot and staff injury. (Tr. 985-90, 992-5; Ex. C-2b.) There was not enough staff to handle the situation, and police had to be called. (Tr. 992-95.)

Post-inspection actions support finding that these proposals effectively mitigate the hazard. Respondents added several additional full-time positions. The Milieu Specialists work from around mid-morning to about 9:30 p.m. every day. (Tr. 850-51, 856.) They provide support, mentoring and additional help when patient behavior escalates. (Tr. 572, 851, 1668-69, 1683, 1961, 2515, 2806; Exs. C-76 at 3, 9, C-94 at 38.) They also fill in as MHSs when there are fewer staff present than what the staffing grid calls for.⁸¹

Respondents also added an additional float MHS to the night shift. (Tr. 857.) This person was in addition to the staff called for by the grid. Id. They added a Milieu Manager, Program Managers, an Assistant Director of Nursing, and a person to focus on recruiting and retaining employees. (Tr. 846-47, 850, 936, 1073, 1953-54, 1956, 1963, 2517-18, 2604; Ex. C-94 at 38, 40, 44; CS Br. 32.) The Milieu Managers can do a variety of things, including filling in or providing extra support to the unit staff. (Tr. 846-47, 850; Ex. C-94 at 40, 44.) The recruitment and retention position has helped with having an adequate number of workers. (Tr. 859, 1073-74; Ex. C-94 at 44; CS Br. 31-32.) Management confirmed that the new positions benefited worker safety. (Tr. 856-57, 1963, 2525.)

The OSHA Guidelines emphasize the importance of having adequate staffing to materially reduce the hazard of workplace violence. (Ex. C-95 at 10, 26.) The TJC Alert called on employers to analyze staffing levels and assign sufficient staff to units. (Ex. C-100 at 3, 5.) Doctors Lipscomb and Forman agreed. Dr. Lipscomb determined that this proposal would materially reduce the hazard of workplace violence and prevent incidents of staff injuries. (Ex. C-94 at 33.) Dr. Forman explained that appropriate staffing for patient acuity would yield “faster, better and safer outcomes for staff.” (Ex. C-92 at 17.)

Respondents contend that they do not know what sufficient or adequate staffing means. (CS Br. 109, 125.) At the same time, they claimed that the Worksite’s staffing “can” be adjusted for acuity and that “one of the primary duties of the nurse supervisor is to ensure adequate staffing.” (CS Br. 26-28.) The staffing policy indicated the facility was committed to providing the appropriate number of personnel during all hours of operation. (Exs. C-9 at 1; RCS-47.) Management was to ensure that the appropriate staff were available for the patients served. (Exs. C-9, RCS-47.) Thus, the issue is not that Respondents do not understand the Secretary’s proposal. Their policies identified the need to maintain adequate staffing, including by increasing staff for acuity and the admission of more patients. (CS Br. 26-30, 68; Exs. C-64, C-65, C-66.) The proposal is for Respondents to maintain sufficient staffing in practice. (Exs. C-92, C-94.) See HRI, 2019 WL 989735, at *13 (finding behavioral health hospital’s practice of “maintaining adequate staffing” was effective abatement for the hazard of workplace violence).

Respondents also cited having one-on-one processing with an agitated patient as a potential intervention to address aggression before it results in actual violence. (CS Br. 34, Exs. C-64, C-65, C-66.) One-on-one verbal interaction was frequently used as a measure to address behavior before a patient engaged in violence. (Tr. 2344; Exs. C-64 at 1, 6. 9, 12, 15, 18, 26, 37, 39, 42, 48, 53, 61, 63, 66, C-65 at 1, C-66 at 1, 5.) The form used for treatment plans also identified “talking to staff” as a method to help prevent patients from endangering others. (Exs. C-64 at 2, 11, 25, 35, 40, 52, 54, 75, 76, C-65 at 7.) After violent behavior, staff or the patients would specify that one-on-one talks with staff would help address future agitation. (Tr. 2344; Exs. C-64 at 4-5, 34, 70, 77, C-65 at 5.) Doctors also sometimes ordered one-to-one supervision. But this precaution could only be implemented on the shift during which it was ordered if there are enough MHSs in other units to re-assign. (CS Br. 29; Tr. 2370-71; Exs. C-3 at 43, C-58.)

The Secretary is not arguing that one-on-one discussions or assignments necessarily prevent any particular patient from engaging in or repeating violent acts. The patient statements indicating that one-on-one verbal conversations would benefit them if they were agitated were not offered for the truth of those assertions. (Tr. 635-36.) The point is that identifying triggers and addressing agitation through verbal de-escalation was a method Respondents found reduced workplace violence. (CS Br. 33-36, 39, 71.) Neither this nor many other actions the WVPP called for can be implemented without adequate staffing. (Ex. C-3 at 57; Sec’y Br. 96-98, 107, 113.)

The Secretary showed that this abatement method would materially reduce the hazard. See BHC, 2019 WL 989734, at *33 (concluding that a feasible method of abatement for the hazard of workplace violence at a behavioral health facility would be for the employer “to determine the appropriate number of staff each unit requires based on acuity of the workplace violence hazard”); UHS Centennial, 2022 WL 4075583, at *37, 44, 46-51 (discussing how addressing the restraints imposed by insufficient staffing would mitigate the hazard).

Respondents trained direct care workers in techniques to handle patient aggression. Too often, there simply were not enough people available when violence commenced to implement the actions called for and to minimize the risk of serious worker injury. Part of the Secretary’s proposal is to have designated staff who can immediately respond. (Tr. 177-78.) Doing so would reduce the hazard by preventing injuries and minimizing their severity. (Tr. 177-80, 253-54.) Trained security would be an effective deterrent to violent patient behavior. (Tr. 179; Ex. C-3 at 58.)

This proposed method was supported by expert testimony and its use at other similar worksites. See Pepperidge Farm, 17 BNA OSHC at 2034. Doctors Forman and Lipscomb detailed how having dedicated response staff, as described in the proposal, would decrease employee exposure to the cited hazard and prevent injuries. (Exs. C-92 at 12-13, 17, C-94 at 37, 39-40.) Dr. Forman has extensive experience working in medical facilities where specific individuals were dedicated to security tasks. (Ex. C-92 at 17.) Unlike staff with multiple roles, security can respond faster to requests for assistance. (Tr. 177; Exs. C-92 at 17, C-94 at 37.) When responses are delayed, the risk of injury and the potential for more severe injury increases. (Ex. C-94 at 37, 40.) This abatement has a preventative effect in addition to better handling incidents that arise. (Tr. 1226-28, 1231; Ex. C-92 at 17.) Dr. Forman noted that the presence of security staff during verbal de-escalation can make patients realize that violence will not achieve what they think it will. (Ex. C-92 at 17.) Dr. Lipscomb similarly opined that this abatement method would materially reduce the likelihood of injury. (Tr. 1675-77, 1682; Ex. C-94 at 37, 40.) Hiring and training security staff to work in behavioral health is a “best practice” in workplace violence prevention. (Ex. C-94 at 38.) See BHC, 951 F.3d at 566 (upholding abatement that aligned with “well-known industry best practices”). The OSHA Guidelines and TJC Alert cite using security to reduce the hazard. (Tr. 1679-82, 1732-33; Exs. C-95 at 10, 14, 26, C-96 at 21-22, C-100 at 3, 5-6.) And the OSHA Road Map provides examples of its successful implementation at behavioral health facilities. (Exs. C-94 at 38, C-96.) See UHS Centennial, 2022 WL 4075583, at *30-32 (discussing the OSHA Guidelines, OSHA Road Map, and TJC Sentinel Alert).

Former workers explained the effectiveness of this abatement in addressing the hazard in other facilities where they worked. (Tr. 417-18, 700-01, 780-81, 1457-58, 1510-12.) See Pepperidge Farm, 17 BNA OSHC at 2034 (“successful use of a similar approach elsewhere” and expert testimony supported finding abatement method was effective and feasible). All the psychiatric facilities where a former nurse supervisor worked after leaving the Worksite had security officers, and she found such staff very helpful. (Tr. 417-18.) Another witness discussed the University of New Mexico Hospital, which has employees who specialize in security services respond to codes. (Tr. 780-81.) Similarly, Ms. Wilkosz has worked at numerous other psychiatric facilities, and all of them had security. She explained that security comes in as a show of support, and their presence often causes patients to settle down. (Tr. 1510-12.) Workers also identified adding security as a method to increase safety in Respondents’ 2020 Safety Survey. (Tr. 1897; Ex. C-70.) Several workers indicated that based on their experience, security would increase employee safety. (Tr. 417-18, 781, 789, 1457-58, 1510-12.)

Against this evidence is the testimony of the CEO and Dr. Staggs. The CEO took a negative view of “security guards” who “do not do well with de-escalating patients” and those who respond with a “law enforcement attitude.” (Tr. 2901-2.) He was not asked about the Secretary’s specific proposal, which calls for either “designated specific staff with specialized training in security” or “trained security specialists.” (Tr. 177-78; Ex. C-1.) He did not claim that workers who had undergone Respondents’ training in de-escalation and other techniques and had to follow Respondents’ policies would inappropriately respond. Such an approach is in place at other facilities. (Tr. 817-18, 1164, 1230-32.) For example, a nurse explained that her current in-patient psychiatric facility has security who are trained in a technique similar to the SAMA training that the Worksite used. (Tr. 817-18.) The CEO did not address the widespread use of security in the relevant industry. (Tr. 417-18, 700, 780, 1139, 1149, 1510-12, 1680-81; Exs. C-92 at 12-13, 17; C-94 at 32-33, 38-39; C-96.)

The Secretary is not calling for personnel without training on responding appropriately to behavioral health emergencies. The CEO’s concern rings hollow considering the frequency with which employees needed to rely on actual law enforcement to respond to behavioral health emergencies at the Worksite. The Secretary proposes having trained staff on-site to contain violence better before actual law enforcement is necessary. See UHS-DE/Suncoast, 2023 WL 2388069, at *96-99 (having designated staff to respond to emergencies was preventative rather than the reactionary approach of calling law enforcement after situations became dangerous).

Similarly, Dr. Staggs’ testimony did not rebut or undermine the Secretary’s evidence. He lacked experience with, or expertise in, working with a dedicated response staff in a psychiatric setting. He was not accepted as an expert in how to abate the hazard of workplace violence. Ultimately, the staffing studies are inconclusive. Some studies drew no conclusions on the effect of security on assault rates and others noted benefits to security. (Tr. 2271-7, 2244-46, 2275-76.) Some were of weak design and were close to being a collection of anecdotes. (Tr. 2287-89.)

Neither Dr. Staggs nor the other record evidence sufficiently rebutted the opinions of Doctors Lipscomb and Forman. The record supports the conclusions of the Secretary’s experts and the CO. Feasibility and effectiveness are also shown by actions taken after the inspection commenced. There are now rapid response teams and additional staff. (Tr. 1539; Ex. C-94 at 38.) The response teams do not have specific patient care duties and are free to respond to psychiatric emergencies. (Tr. 2714-15; Exs. C-94 at 38; C-66 at 5.) These teams have been a success, according to Dr. Gray. (Tr. 2714-15.)

The Secretary established that having designated staff, i.e., those without other assignments that would prevent them from immediately responding to aggression, would be an effective method of reducing the cited hazard.

Among other evidence, the Secretary showed the feasibility of the proposed abatement methods through: (1) the testimony of conscientious experts, familiar with the industry; (2) Respondents’ ability to add multiple new staff positions after the inspection commenced; (3) Respondents’ size and ability to pay for additional abatement measures; and (4) the successful implementation of similar abatement methods elsewhere.

Looking first to the experts, Dr. Staggs offered no opinion on the feasibility of these or any other of the Secretary’s abatement proposals. Dr. Lipscomb opined that the proposals were feasible. (Tr. 1577, 1654, 1668-69, 1674-76, 1678-83; Ex. C-94.) She noted that the OSHA Road Map provides examples of how other employers in the industry used security. (Tr. 1678-83; Ex. C-96.) Dr. Forman also opined that the abatements were feasible. (Tr. 1372; Ex. C-92 at 24.)

By the time of the hearing, Respondents had revised and clarified their staffing policies. For instance, the form used to request additional staffing now included guidance on the conditions that call for adding more staff per patient. (Tr. 2559-60; Ex. RCS-91.) One of the specified factors is an episode of violence. (Tr. 2559-60.) There is no evidence the changes had any significant cost or impacted profitability. (Tr. 2003; Exs. C-89, C-90.)

After the inspection commenced, several new full-time positions were added. This includes the Milieu Specialists and overnight MHS float. (Tr. 572, 857, 1539, 1959-60, 2516, 2614-15, 2806-7.) To the extent that these new positions were not enough for Respondents to implement these proposed abatement methods, the Worksite’s former CFO discussed how the Worksite could have afforded to have higher staffing levels, such as having workers who could assist in preventing and responding to Code Green events.⁸² The former CEO, Brandon Askew, indicated that Cedar Springs was profitable and that it did not have to maintain any minimum profit level. (Tr. 1989-90, 2003, 2037-40, 2078-79; Exs. C-89, C-90.) The profitability statistic on Respondents’ “key indicator report” is the contribution margin. (Tr. 2076-78; Ex. C-90.) Mr. Askew believed the Worksite should have 19.2% “contribution margin” to be considered profitable.⁸³ From May 2019 through May 2020, the Worksite’s margin was 30.3%, much higher than the profitability threshold Mr. Askew identified. (Tr. 2046-47; Ex. C-90; Sec’y Br. n.38.) According to Mr. Askew, Cedar Springs could have added twelve additional around-the-clock MHS positions and would still have been profitable. (Tr. 2049-51.) The average cost per full-time employee in 2020 at the Worksite was $73,798, including benefits. (Tr. 1994-95, 2032-33, 2037-38, 2040-41, 2045, 2062, 2078; Exs. C-89, C-90.) An MHS’s salary would be lower than that average amount.⁸⁴

The post-inspection financial data is consistent with his testimony. Respondents added several positions after the inspection commenced and again after the Citation’s issuance. The Worksite’s contribution margin, a key indicator of profitability, remained significantly above the margin the CFO identified to remain profitable. (Tr. 2003, 2038, 2076-78; Exs. C-89, C-90.) Thus, not only could the Worksite remain profitable, there is also no evidence the proposed abatements would “clearly threaten” Respondents’ “economic viability.” Nat’l Realty, 489 F.2d at 1266 n. 37.

Other ALJs have also found similar proposals to be effective abatement for the hazard of workplace violence. In UHS-DE/Suncoast, no staff was tasked only with security-related responsibilities. 2023 WL 2388069, at *69, 96. The proposed abatement included two measures calling for designated staff to prevent and respond to violence.⁸⁵ The ALJ determined that the abatement would be effective mitigation for workplace violence. Id. at *89, 99. See also BHC, 2019 WL 989734, at *32-33 (employer added a security mental health technician without ongoing direct patient care responsibilities to assist with safety concerns). UHS-DE/Pembroke addressed another analogous situation. There, the Secretary proposed that the employer “provide security staff and/or crisis intervention specialists on all three shifts to assist in preventing and responding to violent events.” 2022 WL 774272, at *41, 50-52. It required the behavioral health facility to have “employees without specific direct care tasks focused on safety who could immediately respond to actual or threatened incidents of violence.” Id. at *51. The ALJ concluded that the Secretary established that the proposed abatement would materially reduce the hazard and was feasible.⁸⁶

In sum, expert testimony and the actions taken at similar worksites established this proposed method as feasible and effective mitigation. See Pepperidge Farm, 17 BNA OSHC at 2034. Feasibility and effectiveness are also shown by Respondents’ post-inspection abatement and by their claims that they can implement their policies.

The Secretary showed there were additional feasible measures Respondents could have taken to further reduce the number and severity of incidents of the hazard. Respondents knew protective workstations, reliable communication devices, and a fully implemented WVPP were critical means to address the hazard. It is readily apparent that reliable communication devices and the WVPP require adequate staffing to effectively reduce the hazard to the extent feasible. Respondents partially implemented several of the proposed abatement methods, including maintaining higher staffing levels, and remained profitable. (Tr. 2003; Exs. C-89, C-90.) As discussed by Doctors Lipscomb and Forman, and in the Guidelines, OSHA Road Map, and TJC alert, other similar employers were able to implement the abatement the Secretary proposes.

In some circumstances, the OSH Act’s remedial purposes are best effectuated by treating interrelated entities as a single-employer. C.T. Taylor Co., Inc., 20 BNA OSHC 1083, 1086-88 (No. 94-3241, 2003) (finding that in certain circumstances, the purposes of the OSH Act, including effective enforcement, “are well served” by holding two separate legal entities responsible for a violation); Altor, Inc., 23 BNA OSHC 1458, 1463-64 (No. 99-0958, 2011), aff’d, 498 F. App’x 145 (3d Cir. 2012) (unpublished). The Commission looks at three factors to determine whether the entities should be treated as a single-employer and be responsible for a single violation of the OSH Act. UHS-DE/Pembroke, 2022 WL 774272, at *2; UHS-DE/Suncoast, 2023 WL 2388069, at *3-10. First, do the entities share a common worksite? 2022 WL 774272, at *2; 2023 WL 2388069, at *3. Second, are the entities interrelated and integrated for safety and health matters? 2022 WL 774272, at *2; 2023 WL 2388069, at *3. Third, do the entities share a common president, management, supervision, or ownership?⁸⁷ The Secretary bears the burden of establishing a single-employer relationship. 2022 WL 774272, at *2; 2023 WL 2388069, at *3. See also Loretto, 23 BNA OSHC at 1358 n.4.

The Commission twice addressed whether UHS-DE operated as a single-employer with an affiliate. UHS-DE/Pembroke, 2022 WL 774272, at *2; UHS-DE/Suncoast, 2023 WL 2388069, at *3. Like the present matter, those cases concerned behavioral health hospitals for which employees of each entity had responsibilities for the same workplace. 2022 WL 774272, at *3; 2023 WL 2388069, at *4. UHS-DE and the hospitals were “interrelated and integrated” with respect to health and safety matters. 2022 WL 774272, at *4-5; 2023 WL 2388069, at *4-9. In terms of common management, like here, the CEOs of the hospitals were employed by UHS-DE. 2022 WL 774272, at *6; 2023 WL 2388069, at *4-5. In both cases, the Commission found that all three factors weighed in favor of finding a single-employer relationship. Id. The same finding is appropriate here.

“Cedar Springs’ address listed on its federal corporate tax return is 367 S. Gulph Road, King of Prussia, PA 19406.” (Stip. 20.) The parties stipulated that UHS-DE’s “corporate address” is at the same location.⁸⁸ However, the Worksite is in Colorado. (Stip. 3; Exs. R-UHS 3, R-UHS 6.) “UHS-DE does not own the property where Cedar Springs Hospital is located.” (Stip. 49.) The Management Agreement between UHS-DE and Cedar Springs indicates that Cedar Springs “owns” a behavioral health facility in Colorado. (Ex. C-84.)

Two business addresses do not preclude finding a common worksite. 2022 WL 774272, at *3; 2023 WL 2388069, at *4. See also A.C. Castle, 882 F.3d at 42 (noting that while a shared headquarters or business address “generally satisfies the common worksite factor,” it is not “necessary”). While two entities can “have different principal addresses and perform their primary work at different locations, the central inquiry is whether they share a common worksite.” 2022 WL 774272, at *3.

Cedar Springs employed most of the on-site personnel. (Stip. 51.) Still, UHS-DE approves and employs those who hold the key supervisory positions of CEO and CFO. (Stips. 21, 25, 26; Tr. 1917-18, 1979-80; Meloni Dep. 19:10-15, 21:13-24, 22:1-5.) The CEO was the direct supervisor of multiple people at the Worksite, including the CFO, the Director of Risk Management, the Director of Admissions and Referrals, the Human Resources Director, and the Director of Plant Operations. (Stips. 21, 25; Tr. 1919-20, 1981.) The CFO was often at the Worksite, including in the patient care units. (Tr. 1984.) He supervised several Cedar Springs’ employees and had hiring and firing authority. (Tr. 1981-82.) Additional UHS-DE employees were often at the Worksite and involved in addressing safety. (Stip. 42; Tr. 904, 916-17, 1240, 1859, 1863-64, 1867-68; Exs. C-29, C-45, C-77, C-78, C-79, C-80.)

These and other facts support finding a common worksite. Advance Specialty Co., 3 BNA OSHC 2072 (No. 2279, 1976) compels no different conclusion. (UHS Br. 31, 34, 41, 51.) That case held that mutual access to a shared hazard may be sufficient to establish a common worksite. 3 BNA OSHC at 2076. In this matter, the parties stipulated that employees of both entities were exposed to the cited hazard at the Worksite. (Stips. 7-8). See UHS-DE/Pembroke, 2023 WL 3243988, at *1 (concluding that mutual access is not required to find a common worksite but, in any event, workers from both entities had such access). In short, the common worksite factor supports finding a single-employer relationship between UHS-DE and Cedar Springs. See 2022 WL 774272, at *3 (finding the presence of one UHS-DE employee and involvement of additional UHS-DE employees sufficient to show a common worksite); 2023 WL 2388069, at *4 (finding a common worksite when three UHS-DE employees worked alongside other hospital employees at the same location); C.T. Taylor, 20 BNA OSHC at 1087 (finding single employer relationship when both entities shared an office and had a presence at the site where the hazard was located).

Very much akin to the facts of UHS-DE Suncoast and UHS-DE/Pembroke, Cedar Springs and UHS-DE were interrelated and integrated when it came to workplace safety, including in their approach to protecting their employees from the cited hazard. (Stips. 7-8.) Hence, this factor also favors finding a single-employer relationship. UHS-DE/Suncoast, 2023 WL 2388069, at *4; UHS-DE/Pembroke, 2022 WL 774272, at *4 (finding UHS-DE and a behavioral health facility had interrelated and integrated approaches to safety and health matters). See also C.T. Taylor, 20 BNA OSHC at 1087 (shared inquiries into safety line demonstrated that two entities “handled safety matters as one company”).

UHS-DE was involved in the Worksite’s tracking of worker injuries, including those occurring from workplace violence. “UHS-DE Loss Control Managers handle Cedar Springs worker’s compensation claims and track worker’s compensation expenses.” (Stip. 41.) UHS-DE retains a contractor (Sedgwick) to handle employee injury reporting for the Worksite and compile OSHA 300 logs. (Stip. 39-40; Tr. 909, 1059-60, 1064-65, 1067; Meloni Dep. 18:19-24, 19:1-3; Exs. C-3 at 47, C-12 thru C-17.) When workers call the reporting line, they were often asked for their “UHS location.” (Exs. C-12, C-13, C-14, C-17.) For reported injuries, someone from UHS-DE works with Cedar Springs employees to obtain more information. (Tr. 909-13, 1059-60; Exs. C-3 at 45, C-12, C-16, C-44, C-48.) UHS-DE employees sometimes instructed Cedar Springs employees to take further steps related to the reported injuries. Id. Sometimes a UHS-DE employee is the injured worker’s “loss control manager,” and may call the injured worker directly. (Stip. 42; Tr. 909, 1818; Exs. C-12, C-16, C-17, C-27.) The injured employees are directed to speak to the UHS-DE employee when they call. Id.

UHS-DE provides support to maintain the MIDAS database, which is used to record some incidents of patient-to-staff aggression. (Stip. 38.) UHS-DE can access data, such as patient aggression information from MIDAS. (Tr. 1859-61, 2807-8; Ex. C-49.) Cedar Springs employees also had access to UHS-DE’s data to compare the Worksite with similar facilities. (CS Br. 73 n. 84; Tr. 905.)

The interrelation and integration are also shown by: (a) the Management Agreement, (b) the involvement of UHS-DE’s personnel at the Worksite, and (c) UHS-DE involvement in training.

“UHS-DE is not a licensed healthcare provider.” (Stip. 50.) It “performs management services for Cedar Springs pursuant to a Management Agreement.” (Stip. 28; UHS-DE Br. 1, 12, 14, 35-38.) The agreement tracks the one the Commission addressed in UHS-DE/Pembroke. (Ex. C-84 at 2-3.) Under it, UHS-DE is responsible for “(a) the billing system; (b) the collection system; (c) the disbursement system; (d) the payroll system; (e) the insurance claim system; (f) the management information system; and (g) the patient safety improvement system.” (Stip. 31; Ex. C-84 at 2-3.) See UHS-DE/Pembroke, 2022 WL 774272, at *5 (quoting identical language from the management agreement between UHS-DE and Pembroke). The agreement also indicates that UHS-DE will provide Cedar Springs with personnel “necessary for the management” of the Worksite. (Tr. 1966-68; Exs. C-45; C-84 at 4.)

The management agreement reflects UHS-DE’s authority over the Worksite’s budget and finances. See 2022 WL 774272, at *5. UHS-DE assists Cedar Springs with preparing its annual capital and operating budgets. (Tr. 1991-2, 1994-95; Exs. C-84 at 2, C-89.) It requires Cedar Springs to pay a monthly fee and takes that fee directly from Cedar Springs’ accounts.⁸⁹ The CFO did not know how the fee amount was determined and had no role in negotiating it. (Stip. 29; Tr. 2002, 2052-53.) “UHS-DE provides and administers a number of systems for Cedar Springs,” including ones for payroll and benefits, the billing system, the collection system, the disbursement system, and the insurance claim system.⁹⁰ UHS-DE “files and certifies Cedar Springs’ federal income tax returns.” (Stip. 36.) And UHS-DE provides approved vendors, and orders and purchases some inventory and supplies. (Tr. 1964-65; Ex. C-84 at 2.) See UHS-DE/Pembroke, 2022 WL 774272, at *5 (finding UHS-DE authority over budget and finances as addressed in a management agreement persuasive support for a single-employer relationship).

UHS-DE provides legal services to Cedar Springs, assisting with the compliance of “all laws, rules, regulations, interpretive guidelines,” and other requirements. (Ex. C-84 at 3; Stip. 39; Meloni Dep. 34:9-11, 41:19-20, 55:20-24, 56:1-2; UHS-DE Br. 16, 37-38.) The UHS-DE legal department must review all contracts when Cedar Springs wants to engage someone or some entity for products or services. (Stip. 39; Meloni Dep., 55:20-23, 56:1-2.) UHS-DE administers a hotline to which employees can raise compliance concerns, including those related to safety. (Tr. 920, 922-23; Ex. C-72.) Cedar Springs reported the start of OSHA’s investigation to UHS-DE and UHS-DE was to provide “support” throughout the process. (Exs. C-36, C-76 at 4-5, C-84 at 3.) The UHS-DE legal department wanted to know about all OSHA investigations. (Ex. C-36.)

Contrary to UHS-DE’s assertions, the management agreement supports concluding that the cited entities have integrated and interrelated operations. See UHS-DE/Pembroke, 2022 WL 774272, at *4-5 (finding that a similar management agreement between UHS-DE and a behavioral health facility supported finding that the two entities were a single-employer); UHS-DE/Suncoast, 2023 WL 2388069, at *4-5 (rejecting the employer’s attempt to characterize UHS-DE’s managerial responsibilities at the facility as “resource sharing”).

UHS-DE employed the CEO, CFO and Chief Operating Officer in training(“COO”) for the Worksite.⁹¹ The CEO was responsible for ensuring staff safety and preventing staff injuries. (Tr. 1920-21.) He participated in key Worksite committees. (Tr. 1920; Exs. C-74, C-75, C-76.) He was responsible for hiring, firing, and managing Cedar Springs’ staff. (Tr. 834, 1054, 1816, 1920.) See UHS-DE/Pembroke, 2022 WL 774272, at *4 (considering the CEO’s responsibility for hiring, disciplining, and firing hospital staff to be a key reason for finding the second element of the single-employer test to be met). He also oversaw senior department directors such as the directors of Medical, Nursing, Human Resources, Plant Operations, Performance Improvement and Risk Management. (Stip. 25; Tr. 834, 1054, 1816, 1920.)

The CFO is responsible for managing the Worksite’s finances. (Stip. 26; Tr. 1980-81.) He was part of the Board of Governors and Performance Improvement Committee. (Tr. 1983; Exs. C-76.) UHS-DE’s Regional Vice President of Operations reviewed the Worksite’s budgets, and they were approved by UHS-DE’s Vice President of Operations. (Tr. 1991-92; Ex. C-89.) “UHS-DE has final approval of Cedar Springs’ budgets,” and “must approve any capital expenditure over $5,000.00.” (Stip. 33, 34.) It selects where the Worksite’s revenue is deposited and certifies Cedar Springs’ federal income tax returns.⁹² The CFO received a bonus if Cedar Springs exceeded its predicted profit for the year. (Tr. 1985, 1997; Ex. C-89.)

The CEO and CFO sit on the Board of Governors. (Stip. 27; Tr. 852-53, 1920, 1923, 1983-84; Exs. C-76, C-85 at 4.) See UHS-DE/Suncoast, 2023 WL 2388069, at *5 (CEO’s involvement with Governing Board weighed in favor of finding interrelation and integration of operations); UHS-DE/Pembroke, 2022 WL 774272, at *4 (citing the involvement of UHS-DE employees in an advisory board that sets policy supported finding of interrelation and integration). The Board of Governors (also called Governing Board) is responsible for “planning, management and operational activities,” including approving changes in Worksite policy.⁹³

Additional Cedar Springs’ employees, including the Risk Management Director (Kolln) and the Director of Nursing (Mattson), report to the CEO and other UHS-DE employees. (Tr. 903-4, 1093, 1095-96, 1830, 1856-59; Meloni Dep. 15:7-14.) Ms. Kolln must send the monthly reports and provide other metrics to a UHS-DE Risk Manager. (Tr. 1830, 1857-59, 1874; Ex. C-88.) Ms. Mattson supervised the medical staff and clinical care team along with the Medical Director. (Tr. 923-24, 2072; Stip. 25; UHS-DE Br. 21.) Ms. Mattson attends mandatory monthly nursing calls with the UHS-DE Divisional Director of Nursing. (Tr. 903-4.)

UHS-DE/Pembroke involved a similar structure. 2022 WL 774272, at *4. The Commission considered the Director of Nursing and the Risk Manager reporting to UHS-DE strong support for finding that UHS-DE manages daily operations. Id.

The following UHS-DE employees conducted site visits of Cedar Springs Hospital in 2019:

b. Patrick Wilder, Divisional Director of Clinical Services – 3/20 to 3/21 2019; 4/30 to 5/2/2019; 5/8/2019; 6/27/2019; 7/19/2019; 7/31/2019; 8/19 to 8/22/2019; 10/3/2019; and 12/2 to 12/5/2019;

(Stip. 42.) Their reviews have included making recommendations and advising actions related to the cited hazard. (Tr. 916-17, 1859, 1863, 1867-68; Ex. C-29, C-77, C-78, C-79.) For example, a UHS-DE Loss Control department employee’s visit was to include discussions about OSHA and the Worksite’s WVPP. (Tr. 1867-68; Exs. C-29, C-78.)

Cedar Springs’ employees are required to take certain trainings and follow UHS-DE’s Code of Conduct. (Tr. 908, 1110, 1114-17, 1871, 1933, 2893-94; Exs. C-3 at 49, C-47, C-81, C-82, C-83, C-86, RCS-18 at 13; Meloni Dep. 14:1-5, 22:6-16, 37:19-24, 38:1-3.) Of particular relevance to assessing interrelation and integration for this matter, Cedar Springs’ employees were required to attend UHS-DE’s courses on Preventing Workplace Violence and Corporate Compliance. (Tr. 1113-17; Exs. C-47, C-48, C-82.) This training was generic in that it could be applied to any UHS-DE behavioral health hospital. (Tr. 1115-17; Ex. C-82.) See UHS-DE/Suncoast, 2023 WL 2388069, at *56-57 (describing similar generic workplace violence training). Many training and other materials refer to UHS-DE or UHS. (Tr. 1871-72, 1991; Exs. C-82, C-89, RCS-17 at 2, RCS-18 at 13, RCS-21 at 4, RCS-22 at 39, C-83.) See UHS-DE/Suncoast, 2023 WL 2388069, at *4-9 (finding that using UHS-DE’s policies, forms, and templates, was evidence of a single-employer relationship).

The Human Resources Director attended multiple UHS-DE trainings. (Tr. 1105-10; Ex. C-40.) He was instructed to review and ensure that the Worksite adopted model policies provided by UHS-DE. (Tr. 906, 1105-10; Ex. C-40.) The policies included ones related to workplace violence. (Tr. 1109; Ex. C-40; Meloni Dep. 13:13-24.) UHS-DE specifies certain things to include in Cedar Springs’ restraint and seclusion policy. (Tr. 906; Ex. RCS-22 at 39.) UHS-DE provided nursing admission assessments and daily progress note forms and instructed Cedar Springs to use them. (Tr. 907-8.) Cedar Springs is required to use a UHS-DE approved incident report form, adopt UHS-DE’s HPR (Healthcare Peer Review) Policy, and use UHS-DE’s HPR Dictionary when entering incident reports. (Tr. 1872-74; Exs. C-83, RCS-18 at 13.) UHS-DE required the Worksite to have a WVPP and provided a template. (Tr. 1817; Exs. C-27, C-28.) Once prepared, UHS-DE reviewed the WVPP and provided related training materials. (Tr. 1818-22; Exs. C-27, C-28.)

The Secretary established that Cedar Springs and UHS-DE are interrelated and integrated in their overall operations and in health and safety matters. See 2023 WL 2388069, at *9.

UHS-DE suggests that there had to be parity where the employee served as a leader for both Cedar Springs and UHS-DE. That is not the test. Common management can mean that one entity’s employee oversees the other entity’s financial, executive, or operational aspects. UHS-DE/Pembroke, 2022 WL 774272, at *5-6. The focus is on the relationship between the entities at the relevant worksite for the citation. Id.; C.T. Taylor, 20 BNA OSHC at 1087 n.7.

The cited entities are linked through the CEO, CFO, and other UHS-DE employees.⁹⁴ Key oversight committees were either led by or had UHS-DE employees as members. (Tr. 1701, 1920, 1982-84, 2715.) Both share the same ultimate corporate parent. See Wal-Mart Stores, Inc. v. Sec’y of Labor, 406 F.3d 731, 737 (D.C. Cir. 2005) (relying on the fact that two stores had the same “controlling corporation” to support a repeat characterization).

UHS-DE was integrally involved in the Worksite’s day-to-day management, including with safety and core business functions. See UHS-DE/Pembroke, 2022 WL 774272, at *6 (concluding that similar facts showed that the common management, supervision, or ownership prong of the single entity test meant there was a single-employer relationship); UHS-DE/Suncoast, 2023 WL 2388069, at *9-10; C.T. Taylor, 20 BNA OSHC at 1087 (manager of one company supervising the employees of another supported finding single-employer relationship). As in UHS-DE/Pembroke, there was a direct line of management between Cedar Springs and UHS-DE. The CEO supervised Cedar Springs employees and was, in turn, supervised by UHS-DE employees. This level of oversight suffices. See UHS of Westwood Pembroke, Inc., UHS of Delaware, Inc., v. Sec’y of Labor, 2023 WL 3243988, at *2 (3d Cir. 2023) (unpublished) (common management established when CEO worked for UHS-DE and supervised employees at the worksite).

All three factors support concluding a single-employer relationship existed between UHS-DE and Cedar Springs at the Worksite. See 2023 WL 2343988, at *2 (concluding that UHS-DE and a behavioral health facility had a common worksite, integrated safety operations and common management).

When an authorized government representative asks for certain workplace safety and health records that employers must keep, the employer must provide them within four hours. 29 C.F.R. § 1904.40(a). Of relevance here is the requirement to produce records kept under “Part 1904,” such as the OSHA 300 Log, the annual summary, and the OSHA 301 Incident forms. The OSHA log and OSHA 301 Incident forms are supposed to be promptly completed after qualifying events. As the name suggests, the annual summary is compiled yearly. Employers must timely complete and then maintain these records for five years. 29 C.F.R. § 1904.33. Employers are not required to file the documents with OSHA. Instead, they must promptly produce them upon request from authorized government representatives. See Dole v. Trinity Indus., Inc., 904 F.2d 867, 869 n.3, 872-73 (3d Cir. 1990) (employers must make accident and injury records available for inspection and copying). A representative of the Secretary inspecting a workplace is an “authorized government representative” within the standard’s meaning. 29 C.F.R. § 1904(b)(1)(i).

Respondents contend that there was no failure to comply with the standard. (CS Br. 130.) The record shows otherwise.

During the inspection, the CO requested OSHA 300 logs and the related documents on multiple occasions. First, on November 18, 2019, the CO requested the OSHA 300 logs, 300A forms, and 301 forms for 2015 to 2019. (Tr. 192; Ex. C-3 at 60-63.) Respondents provided incomplete versions of the 300A forms that day. (Ex. C-3 at 4.) No OSHA 301 forms were provided then or in the next two days. OSHA issued a subpoena on November 20, 2019 for updated 300A forms and other documents. (Exs. C-3 at 61, C-6.) Respondents provided some information on November 21, 2019, December 17, 2019, and December 27, 2019. (Exs. C-3 at 61-62, C-6.)

The information provided still did not include the updated 300A forms for three years. (Ex. C-3.) OSHA then delivered another subpoena on January 8, 2020. (Tr. 192-93; Ex. C-3 at 62.) Even without a subpoena these documents need to be provided within four hours. 29 C.F.R. § 1904.40(a). So, by January 8, 2020, the documents should have already been provided. Still, OSHA repeatedly agreed to extend the deadline, ultimately granting Respondents until January 24, 2020. (Tr. 193; Ex. C-3 at 62.) Respondents failed to comply with this well-extended deadline. Id. They did not provide documents responsive to the subpoena until January 28, 2020, days after the extended deadline. (Tr. 193-94; Exs. C-3 at 47, 62; C-7.) That response was also incomplete. (Tr. 193-94; Exs. C-3 at 62-63, C-6, C-7.) It took multiple additional requests before the Secretary received the records Respondents are required to produce within four hours. 29 C.F.R. § 1904.40(a). (Tr. 193-94; Exs. C-3 at 47, 60-63, C-7, C-8.)

The purpose of the recordkeeping obligations is to help OSHA assess the workplace. There is no dispute that workers at the Worksite suffered recordable injuries. These facts show exposure for this recordkeeping violation. There is no contention that Respondents lacked knowledge of the requests. They were made in writing, and some were subject to a formal subpoena. (Tr. 193-94; Exs. C-3 at 62-63, C-7.) Respondents try to deflect attention by arguing that the CO revised the requests multiple times. Any confusion was the result of Respondents’ repeated failure to comply with the valid requests fully. Further, it was Respondents’ provision of incomplete forms that necessitated the CO’s follow-up.

The obligation 29 C.F.R. § 1904.33 imposes is straightforward. Requested records had to be provided within four hours. They were not. The Secretary acknowledges that the violation was a process violation without risk of employee injury. (Sec’y Br. 137.) As such, it is appropriately classified as an other-than-serious violation.

“Once a citation is contested, the Commission has the sole authority to assess penalties.” Valdak Corp., 17 BNA OSHC 1135, 1138 (No. 93-0239, 1995) (citation omitted), aff'd, 73 F.3d 1466 (8th Cir. 1996). The maximum statutory penalty for serious or other than serious violations when the Citation was issued was $13,494.23. Department of Labor Federal Civil Penalties Adjustment Act Adjustments for 2020, 85 Fed. Reg. 2292, 2298-99 (Jan 15, 2020) (to be codified at 29 C.F.R. Part 1903). Section 17(j) of the Act requires the Commission to give due consideration to four criteria in assessing penalties: the size of the employer's business, the gravity of the violation, the employer's good faith, and its prior history of violations. 29 U.S.C. 666(j). Gravity is generally the primary factor in the penalty assessment. See J. A. Jones Constr. Co., 15 BNA OSHC 2201, 2214 (No. 87-2059, 1993).

Starting with Citation 1, Item 1, the Secretary proposed a serious classification and a penalty of $13,494. (Tr. 106-7, 123, 187-89; Exs. C-1, C-4 at 1, 3.) OSHA assessed the gravity as high because of the severity and frequency of the injuries and assaults. (Tr. 123, 187-89; Exs. C-1, C-4 at 3-6, 8-11, 16.) For Citation 2, Item 1, the Secretary proposed an other-than-serious classification and a penalty of $1,928. (Tr. 194-6; Exs. C-1 at 10, C-4 at 19.) The Secretary acknowledged that this violation did not present a risk of employee injury. (Sec’y Br. 10, 137.)

The Secretary argues that the penalty factors do not warrant downward adjustments. (Tr. 187-89; Ex. C-4 at 3, 19.) UHS-DE was previously cited for the hazard of workplace violence. In those matters, the ALJs found that the violation warranted a high gravity penalty, and there were no adjustments for the other penalty factors. See UHS-DE/Suncoast, 2023 WL 2388069, at *1, 177 (ALJ concluded that the violation, coupled with the failure to preserve video evidence warranted the maximum penalty); UHS-DE/Pembroke, 2022 WL 774272, at *1, 14. See also BHC, 2019 WL 989734, at *43 (considering a similar violation to be serious and high gravity and making no downward adjustments for size, good faith, or history); UHS Centennial, 2022 WL 4075583, at *58-59 (considering similar violation to be of high gravity and adjusting only for size).

Over 250 people worked at the Worksite, making an adjustment for small size inappropriate. (Tr. 188, 195-96, Ex. C-4 at 3, 19.) In addition, discounts are inappropriate for history or good faith. (Exs. C-3 at 2, C-4 at 3, 19.) Respondents’ safety and health program for the hazard was not appropriately implemented.⁹⁵ Further, Respondents destroyed evidence related to the hazard. Had the maximum penalty not been warranted for the gravity of Citation 1, Item 1, this failure to appropriately preserve relevant, discoverable information after litigation was reasonably anticipated would have warranted an upward adjustment. It does not appear that the spoliation impacted Citation 2, Item 1. Therefore, an upward adjustment for that item has not been made.

Viewing the record as a whole with the greatest emphasis on gravity, the undersigned assesses the proposed penalties of $13,494 for Citation 1, Item 1 and $1,928 for Citation 2, Item 2.

The foregoing Decision constitutes the Findings of Fact and Conclusions of Law in accordance with Commission Rule 90(a)(1), 29 C.F.R. § 2200.90(a)(1).

Based upon the foregoing Findings of Fact and Conclusions of Law, it is ORDERED that:

1. Citation 1, Item 1 is AFFIRMED as serious, and a penalty of $13,494 is ASSESSED.

2. Citation 2, Item 1 is AFFIRMED as other than serious, and a penalty of $1,928 is ASSESSED.

3. The Secretary’s Motion for Sanctions for Respondents’ Destruction of Relevant Recorded Video Evidence, is GRANTED in part, and DENIED in part.

1 The parties stipulated:

1. Respondent Cedar Springs Hospital, Inc. (“Cedar Springs”) is an employer engaged in a business affecting commerce within the meaning of Section 3(5) of the Occupational Safety and Health Act of 1970, 29 U.S.C. § 652(5).

2. Respondent UHS of Delaware, Inc. (“UHS-DE”) is an employer engaged in a business affecting commerce within the meaning of Section 3(5) of the Occupational Safety and Health Act of 1970, 29 U.S.C. § 652(5).

2 The parties stipulated:

3. Cedar Springs Hospital is an in-patient psychiatric hospital located at 2135 Southgate Road, Colorado Springs, CO 80906. …

19. UHS-DE has a corporate address of 367 S. Gulph Road, King of Prussia, PA 19406.

(Stips. 1-3, 19.) UHS-DE is headquartered in Pennsylvania, and the Worksite is in Colorado. (Stips. 3, 9.) Three Circuits have potential jurisdiction over any appeal of this matter: the Third, Tenth, and D.C. See 29 U.S.C. § 660(a). Generally, Commission judges apply the law of the circuit where it is probable a case will be appealed. See, e.g., Kerns Bros. Tree Serv., 18 BNA OSHC 2064, 2067 (No. 96-1719, 2000).

3 Stips. 1, 2, 45. Cedar Springs noted a numbering error in the parties’ joint stipulations that were filed before the hearing commenced. (CS Br. n.2.) References to stipulation numbers herein will be consistent with the numbering used during the hearing and in the Secretary’s Brief. (Tr. 13-16; Sec’y Br. 2-6.) The undersigned also notes some imprecision in how the stipulations reference the entities. Cedar Springs Hospital, Inc. is shortened to Cedar Springs in Stipulation 1. When referring to the Worksite, as opposed to the corporate entity Cedar Springs Hospital, Inc., the parties frequently use the term “Cedar Springs Hospital,” as in Stipulations 3 and 4, but not always. (UHS-DE Br. 1.) Stipulation 45, in full, is “The Occupational Safety and Health Review Commission has jurisdiction in this proceeding pursuant to § 10(c) of the Occupational Safety and Health Act (“OSH Act”).

6The parties stipulated:

7. Cedar Springs employees at the [W]orksite are exposed to the hazard of workplace violence, defined in this case as physical threats and assaults by patients toward staff.

8. UHS-DE employees at the [W]orksite are exposed to the hazard of workplace violence, defined in this case as physical threats and assaults by patients toward staff. …

21. Cedar Springs’ former Chief Executive Officer (“CEO”) and now UHS Regional Vice President, David Franklin, has been employed by UHS-DE since September 2019.

22. Cedar Springs’ former Chief Financial Officer (“CFO”), Brandon Askew, was employed by UHS-DE during the time of the OSHA inspection.

10 Tr. 840, 511, 2375, 2558, 2563; Ex. RCS-90. The record does not indicate how many therapists were typically at the Worksite. Therapists generally worked one shift during weekdays. (Tr. 395, Ex. RCS-90.) Some units did not have therapists on the weekends or had them for only part of the weekend. (Tr. 1409; Ex. RCS-90 at 1-130.) It does not appear that the therapist supervisors worked on weekends. (Ex. RCS-90 at 1-130.) On the overnight shift, in addition to the unit staff, there was one nurse supervisor and one or two staff for the A&R department. The A&R staff were not nurses or MHSs. (Tr. 397.) The A&R department is in a separate building with no patient care units. (Tr. 840, 2375.)

18 (Stip. 38.) See BHC Nw. Psychiatric Hosp. LLC, No. 17-0063, 2019 WL 989734, at *13-14, 23 (OSHRCALJ Jan. 22, 2019) (discussing the MIDAS system); UHS-DE, Inc., Premier Behavioral Health Solutions of Fla., Inc. d/b/a Suncoast Behavioral Health Ctr., No. 18-0731, 2023 WL 2388069, at *6 (OSHRC Feb. 28, 2023) (describing the facility’s use of the MIDAS system) (“UHS-DE/Suncoast”). Stipulation 38 is: “UHS-DE provides Information Systems support to maintain the MIDAS system, which is used to record incidents that occurred at Cedar Springs Hospital, including some incidents of patient-to-staff aggression.”

20 Tr. 2847-48; Exs. RCS-18 at 5, RCS-33. Francis Martin Mauro, Jr., the trainer, did not indicate Cedar Springs Patient Interaction training was used. (Tr. 2878-80.) After briefly looking at the document, he agreed with counsel’s leading question that the material in the document was “like” the materials he used. (Tr. 2879-80.) He could not remember if the document was provided. (Tr. 2879.) Respondents’ claim that Staff Assistance Codes prevented physical restraints 75% of the time is not directly supported by the record. (CS Br. 37 citing Tr. 2333-34.) In context, the witness appeared to refer to all types of codes, not just Staff Assistance Codes. (Tr. 2322:9 (specifically limiting the question to a Code Green); 2333:21 (question not limited to staff assistance codes); 2334:3 (question not limited to staff assistance codes).) Notably, the minutes discussed by the witness at this point do not include the frequency or success rate of staff assistance codes. (Exs. C-74 at 1, 8, RCS-84.)

23 Ex. RCS-41; Tr. 2831. Mr. Mauro, the trainer and a former MHS, was part of the restraint reduction team. But it is unclear if he worked as an MHS while on that committee. (Tr. 2844, 2846.) During Dr. Lipscomb’s cross-examination, Cedar Springs’ counsel implied that two Milieu Specialists were acting as part of the Restraint Reduction Initiative. (Tr. 1703.) However, the Worksite had no Milieu Specialists until January 2020, when the inspection was well underway. (Tr. 572, 856, 1668, 2557; Ex. C-76.)

24 Tr. 2783; Ex. RCS-3. Respondents cite their expert's report in discussing their processes for reporting and tracking staff injuries. (CS Br. 44 citing Ex. RCS-95.) As detailed below, Vincent Staggs, Ph.D., offered no opinion on the efficacy of the abatement at the time of inspection. (Tr. 2091, 2154-56, Ex. RCS-95 at 1.) Although Dr. Stagg’s report was admitted, the weight given to aspects of the report is tied to the scope of Dr. Stagg’s expertise. (Tr. 2156.)

28 Tr. 2396-98; Ex. C-64 at 9, 23, 59, 70, 72. One debriefing form specifically indicates that Ms. Nolan observed the restraint but did not participate in the debriefing because she was unavailable. (Ex. C-64 at 70.) The rest of the forms for restraints in which Ms. Nolan was involved do not indicate why she did not participate in the debriefing. Id. at 9, 23, 59, 72. Note that a duplicate copy of one form appears on page 71 of Exhibit C-64.

31 UHS-DE, Inc., Premier Behavioral Health Solutions of Fla., Inc. d/b/a Suncoast Behavioral Health Ctr., No. 18-0731, 2023 WL 2388069, at *40-43 (OSHRCALJ Apr. 20, 2022) (“UHS-DE/Suncoast”), aff’d, 2023 WL 2388069 (OSHRC Feb. 28, 2023); UHS of Centennial Peaks LLC, No. 19-1579, 2022 WL 4075583, at *23 n.27 (OSHRCALJ, Jul. 26, 2022) (“UHS Centennial”). UHS-DE was found to be an employer in UHS-DE/Suncoast. 2023 WL 2388069, at *10. UHS Centennial was appealed to the Tenth Circuit. However, the parties notified the Circuit Court of the withdrawal of the petition because they had reached a settlement agreement. In UHS of Fuller, Inc., UHS of Delaware, No. 20-0032 (OSHRCALJ, Jan. 31, 2023) (“UHS-DE/Fuller”), the ALJ also awarded sanctions for the failure to preserve videos. A Petition for Discretionary Review was granted, and this case was directed for review on March 2, 2023, per Commission Rule 92 (29 C.F.R. § 2200.92).

32 Exs. C-19, C-20. The HPR incident logs were compiled from information entered into the MIDAS database. (Stip. 38; Tr. 1837-39, 1846-48, 1853.) They do not include all reported injuries. (Tr. 1825, 1836, 1849.) For example, a worker reported an injury on January 3, 2020, after being kicked and bitten. (Ex. C-44.) There are no corresponding injuries on the log. (Ex. C-20.) While the HPR incident logs have a column labeled “injury,” the information included only relates to whether a patient was injured, not workers. (Tr. 1825, 1836, 1846, 1849, 1863; Exs. C-19, C-20.)

34 Tr. 1900-1901, 1903-9, 1913-14; Exs. C-19, C-20, C-44, C-49 at 2-3, C-75. In September 2020, well after the subpoena was served and litigation commenced, there was a system outage at the Worksite. The cameras were not affected by the outage. (Tr. 1912, 2767-68.) Nor is there evidence it affected any videos retained of past incidents. (Tr. 2768.) However, the outage did temporarily make it less convenient for Ms. Kolln to retain videos. (Tr. 2768-69.) Ms. Kolln did not identify any videos of specific incidents that she was unable to save during the outage. (Tr. 1910-12, 2768.)

35 Sec’y Br. 134. In each of those matters, the Secretary also sought sanctions for a failure to preserve. Sanctions against UHS-DE for the spoliation of video evidence were awarded in UHS-DE/Suncoast, 2023 WL 2388069, at *40-43, UHS Centennial, 2022 WL 4075583, at *23 n.27, and UHS-DE/Fuller, No. 20-0032. Respondents’ counsel for this matter were also counsel in other matters where spoliation of ESI related to workplace violence incidents at behavioral health facilities was found.

41 2019 WL 1142920, at *1. See also Waldon Healthcare Ctr., 16 BNA OSHC 1052, 1059-62 (No. 89-3097, 1993) (finding that the general duty clause applied to the hazard of virus transmission at nursing homes)(“Waldon”); Beverly Enters., Inc., 19 BNA OSHC 1161, 1162 (No. 91-3144, 2000) (consolidated) (finding that the Secretary met the first three elements of establishing a violation of the general duty clause for hazards related to lifting and transferring patients, and remanding for further findings on the issue of feasibility of abatement); Am. Dental Ass’n v. Martin, 984 F.2d 823, 827, 830-31 (7th Cir. 1993) (rejecting the contention that the regulation of medical and dental workplaces was beyond OSHA’s purview); HRI Hosp., Inc. d/b/a Arbour-HRI Hosp., No. 17-0303, 2019 WL 989735, at *1-7 (OSHRCALJ Jan. 22, 2019) (Chief Judge Rooney finding the general duty clause applicable to a provider of psychiatric services) (“HRI”). The Commission has also addressed violations of particular standards at medical facilities. See, e.g., Metwest, Inc., 22 BNA OSHC 1066, 1072 (No. 04-0594, 2007) (upholding a violation against a patient service center), aff’d, 560 F.3d 506 (D.C. Cir. 2009) (upholding citations under bloodborne pathogen standard); Froedtert Mem’l Lutheran Hosp., Inc., 20 BNA OSHC 1500, 1511 (No. 97-1839, 2004) (affirming violations relating to how the hospital handles bloodborne pathogens and communicates hazards); Charles W. Mason, DDS & Assocs., PLLC, 25 BNA OSHC 1792, 1795-96 (No. 10-2313, 2015) (upholding violations related to the handling of sharps used in medical procedures and hazard communication); Loretto-Oswego Residential Health Care Facility, 23 BNA OSHC 1356, 1358 n.2 (No. 02-1164, 2001) (consolidated) (affirming violations related to, among other things, providing timely vaccination shots, making medical evaluations following specific incidents, and training on bloodborne pathogens at a nursing home) aff’d, 692 F.3d 65 (2d Cir. 2012). See also Columbia Presbyterian Hosp., No. 93-298, 1996 WL 18889, at *1 (OSHRCALJ Jan. 2, 1996) (finding a hospital to be an employer and subject to the OSH Act); Am. Dental Ctrs., No. 89-1369, 1990 WL 118162 (OSHRCALJ June 4, 1990) (consolidated) (upholding violations issued to dental treatment facilities).

43 “As a CMS-certified hospital, Cedar Springs is subject to Medicare Conditions of Participation and Conditions for Coverage, which are federal regulations intended to ensure quality standards in hospitals.” (Stip. 15.) The undersigned notes that in their briefs, Cedar Springs and the Secretary appeared to have inadvertently changed the word “ensure” to “endure.” (CS Br. 2; Sec’y Br. 6.) The undersigned relied on the stipulation as stated on the first day of the hearing. (Tr. 12, 15-21.) The parties also stipulated: “As a CMS-certified hospital, Cedar Springs is required to annually submit data to CMS regarding, among other things, its patient days, revenue, number of employees, salaries and benefits paid, and revenue.” (Stip. 16.)

47 Tr. 2148, 2151-52, 2154. Dr. Staggs was accepted as an expert who could provide information related to this method of abatement proposed by the Secretary:

48 Tr. 110-14, 388-89; Exs. C-2b at 7, C-3, C-10, C-94 at 5-6. The parties also reached a stipulation about employees working at UHS-DE’s offices in Pennsylvania. (Stip. 52.) At that location, workers were not exposed to the same hazards as at the Worksite: “The employees of UHS-DE located at the worksite in King of Prussia, PA, are not exposed at that worksite to the same workplace hazards as the employees of UHS-DE and Cedar Springs Hospital who work at Cedar Springs Hospital in Colorado, Springs, CO.” (Stips. 3, 52; Tr. 21.)

49 Stips. 7-11; Ex. C-94 at 6-11. See BHC, 951 F.3d at 563 (behavioral health facility agreed that workplace violence was a hazard present in its facility); UHS-DE/Pembroke, 2022 WL 774272, at *1-2 (affirming a citation for failure to properly abate the hazard of workplace violence at a behavioral health facility); UHS Centennial, 2022 WL 4075583, at *1 (noting that the hazard of workplace violence is “particularly acute” in the healthcare industry). As addressed, the Secretary is entitled to an adverse inference that the destroyed ESI would support finding knowledge.

51 Stips. 7-8, 11; Exs. C-2a at 14, C-2b at 7, C-3 at 8-16, C-10, C-11, C-12, C-17, C-19, C-20, C-43, C-48, C-52, C-55, C-57, C-59, C-60, C-64, C-65, C-66, C-94, C-95 at 5. Cedar Springs alleges that the Secretary stipulated that “2017 was not relevant.” (CS Br. 33.) The parties did not reach a stipulation on this point before the hearing. (Tr. 874-75.) Instead, during the proceedings, the Secretary sought to admit the CO’s narrative of her investigation, which included a list of the documents she sought as part of that investigation. (Ex. C-3.) In context, the Secretary’s counsel was explaining that as part of OSHA’s investigation, the CO requested documents from 2017. (Tr. 77-78, 193.) Counsel acknowledged that when questioning the CO, she did not intend to focus on Exhibit C-3’s references to 2017 events. (Tr. 78-80.) On cross-examination, however, Cedar Springs’ counsel repeatedly brought up injury information from 2017. (Tr. 244-46, 1760, 2605-09.) While 2017 was not the focus of this litigation, the parties did reference historical and background information from 2017. (CS Br. 46, 55, 62; UHS-DE Br. 18; Sec’y Br. 9, 44, 137.) The undersigned focused on the core period of the investigation in reaching this decision.

62 Ms. Kolln’s demeanor and testimony on what data she used from MIDAS to fulfill the WVPP’s requirement were evasive. At one point, she acknowledged that the only information she pulled were incidents coded as “aggression/assault patient towards staff.” (Tr. 1835-36.) That statement is consistent with her deposition testimony and is credited. (Tr. 1841.) A moment later, she disputed that view of what information from MIDAS she used. (Tr. 1839-40.) She was hesitant in her response and appeared to reach her new conclusion only after it was pointed out the flawed approach to using only one category of information from MIDAS.

63 TJC is a non-profit accrediting and standards-setting body. (Tr. 341, 1374, 1607, 1616; Ex. C-100 at 1). It provides guidance to improve safety. (Tr. 1616-17.) The TJC Alert “identifies specific types of sentinel and adverse events and high-risk conditions, describes their common underlying causes, and recommends steps to reduce risk and prevent future occurrences.” (Ex. C-100 at 1.) In response to being notified of such events, the TJC directs that organizations “should consider information in a [TJC Alert] when designing or redesigning processes and consider implementing relevant suggestions contained in the alert or reasonable alternatives.” Id.

65 CS Br. 77-78. In addition, Respondents point to National Realty, a case also addressed in UHS-DE/Pembroke. 2022 WL 774272, at *8. National Realty predates Waldon and the development of the Commission's four-part test for sustaining violations of the general duty clause. See Integra, 2019 WL 1142920, at *4 n.3. In any event, as shown by UHS-DE/Pembroke and UHS-DE/Suncoast, nothing in National Realty requires the Secretary to prove more than one feasible means of abatement.

68 Specifically, in UHS-DE/Suncoast, the Secretary proposed the employer:

Reconfigure the nurses' workstations to include design features that prevent patients from jumping over, reaching into, or otherwise entering into the workstations. Ensure items in the workstations, such as but not limited to scissors, hole punchers, staplers, telephones, cords, pens, computers, computer peripherals, and other items are not accessible by the patients, so they cannot be used as weapons. All scissors at the facility should be replaced with childproof scissors.

2023 WL 2388069, at *93.

70 951 F.3d at 563-65. Specifically, part of the Secretary’s proposed abatement required the employer to:

Provide all affected employees with reliable and readily available means of communication that are effective throughout the facility. Develop a policy to maintain the effectiveness of the communication devices such as walkie talkies. Inform all employees of this policy and train them on the use of the equipment. Enforce the policy as necessary.

BHC, 2019 WL 989734, at *33. The ALJ concluded it was feasible for the employer to address the issues with summoning help and found that this abatement method would materially reduce the hazard. Id. at *35.

71 2022 WL 4075583, at *42. In contrast, in HRI, the ALJ found that because of the facility’s small size and the number of staff in each unit, the Secretary failed to show that the hospital’s phone system, walkie-talkies, fixed panic buttons, and the ability to ask nearby workers for help were inadequate. 2019 WL 989735, at *27-28. A nurse monitored the walkie-talkie traffic to ensure responses to calls for assistance. Id. HRI highlights the inter-relatedness of the Secretary’s proposed abatement. In this matter, staff were frequently alone or in thinly staffed units. Staff responding to a call for assistance sometimes had to come from other buildings. This increased the need for ways to summon help. (Tr. 137-44, 1648-49, 1651; Ex. C-94 at 24.)

77 Tr. 2187-89; Exs. RCS-95 at 5-6, RCS-113. The record includes two articles by Dr. Staggs: “Staffing, RN Mix, and Assault Rates on Psychiatric Units,” which is Exhibit RCS-113, and “Deviations in Monthly Staffing and Injurious Assaults Against Staff and Patients on Psychiatric Units,” which is Exhibit RCS-109. Exhibit RCS-113 looked at the number of RNs (Registered Nurses) versus non-RNs, i.e., licensed practical nurses. It did not compare the number of nurses to MHS type staff. (Ex. RCS- 113.)

78 Exs. RCS-109, RCS-113. Dr. Staggs’ report cites various studies to support the statement that higher staffing levels can be associated with higher rates of violence. (Ex. RCS-95 at 5-6.) These studies were not offered for admission by either party. Dr. Staggs acknowledged that two of the studies cited in his report did not address staffing levels or the other abatement measures on which he opined. (Tr. 2244-46.) He acknowledged significant limitations in the studies he reviewed, conducted, and relied upon. Dr. Staggs was also unfamiliar with how patients are observed at the Worksite. (Tr. 2277.)

81 Ms. Mattson initially claimed that the milieu specialists started on October 9, 2019. (Tr. 847.) However, the position was only being “developed” at that time. (Tr. 2806; Ex. RCS-84 at 26-27.) The specialists did not start working until several weeks after the inspection commenced. (Tr. 572, 856, 2515, 2557, 2614-15; Exs. C-76 at 3, C-94 at 38.) Ms. Mattson gave the report summarized in Exhibit C-76. (Tr. 855.) That report, which indicates that the new Milieu Specialists were not expected to start until January 2020, was closer to the event and was corroborated by other evidence. (Tr. 572, 856, 1017, 1668-69, 2557; Exs. C-76, C-94 at 38.) Equivocal or contrary testimony on when the additional staff began working in the units is rejected. Respondents could have produced evidence of the actual date the Milieu Manager and Milieu Specialists started in the units. (Tr. 846-47; Ex. RCS-90.) Capeway Roofing Sys., Inc., 20 BNA OSHC 1331, 1342-43 (No. 00-1986, 2003) (when one party can produce evidence but fails to do so, there’s a presumption that the evidence would not have been favorable to that party), aff’d, 291 F.3d 56 (1st Cir. 2004). In addition, video evidence of additional events could have clarified when the milieu specialists were fully in place and the extent to which they responded to workplace violence events.

82 Respondents can develop solutions different than the Secretary’s proposal as long as the alternative methods achieve at least as great a reduction in the hazard. See Chevron Oil Co., 11 BNA OSHC 1329, 1334, n. 16 (No. 10799, 1983); Brown & Root, Inc., Power Plant Div., 8 BNA OSHC 2140, 2144 (No. 76-1296, 1980) (“the employer may use any method that renders its worksite free of the hazard and is not limited to those methods suggested by the Secretary”); Pepperidge Farm, 17 BNA OSHC at 2032 (Commission Chair Weisberg concurring and noting that employers are free to develop solutions different than what the Secretary proposes to render their workplace “free” of recognized hazards). The Secretary did not have to prove that Respondents’ post-inspection measures moved the Worksite into compliance.

84 Tr. 1994-95, 1989-90, 2028-29, 208; Exs. C-89, C-90. The Secretary sought judicial notice of the Bureau of Labor Statistics’ average salary for security guards in the Worksite’s geographic area. (Sec’y Br. 127-28.) Doing so is unnecessary. The Secretary showed that Respondents could have hired additional employees to work round the clock. Respondents had significant capacity to remain profitable even after adding staff focused on security and maintaining adequate staffing to address the acuity and number of patients.

85 The proposed abatement was:

Designate specific staff with specialized training in security to monitor patients for potential aggression on all shifts and to assist in preventing and responding to violent events occurring in the units. Designated staff must have the physical capability to effectively respond to aggressive patients. The staff designated to monitor and respond to patient aggression should not be given other assignments such as patient rounds, which would prevent the designated person from immediately responding to an alarm or other notification of a violent incident. Conduct periodic drills for psychiatric crisis/patient aggression (currently known as “Code Grey”) to allow all designated staff to practice and evaluate their skills in real-life settings.

2023 WL 2388069, at *96. Further,

An additional designated staff member with specialized training in security should be available at intake on all shifts. This staff member should have the physical capability to respond to aggressive patients. This staff person should not be given other assignments such as patient rounds, which would prevent the person from immediately responding to an alarm or other notification of a violent incident.

Id.

86 2022 WL 774272, at *52. In UHS Centennial, the ALJ upheld the violation and found the Secretary’s proposal regarding ensuring adequate staffing was a feasible and effective method for reducing the hazard of workplace violence. 2022 WL 4075583, at *49-51. However, he concluded one proposal was “vague” and that the Secretary did not meet the burden with respect to that measure. Id. at *46. The proposal, while related to security, was different than what the Secretary proposes here. Id. In UHS of Denver, Inc., No. 19-0550, 2022 WL 17730964 (OSHRC Dec. 8, 2022), the ALJ relied on an inference to support her finding that an abatement proposal calling for designated security would be feasible. 2022 WL 17730964, at *1. The Commission found that the inference should not have been granted. Id. On remand, the ALJ upheld the citation but concluded that the record lacked sufficient information on the feasibility of one measure. UHS of Denver, Inc., No 19-0550, 2023 WL 4498024, at *16-17 (OSHRCALJ July 3, 2023), appeal docketed, No. 23-9579 (10th Cir. Aug. 31, 2023). Unlike the rejected proposals in UHS Centennial and UHS Denver, here, the Secretary clarified the proposal and more than adequately supported feasibility.

87 2022 WL 774272, at *2; 2023 WL 2388069, at *3. Respondents cite two other Commission decisions on single employer, S. Scrap Materials Co., Inc., 23 BNA OSHC 1596 (No. 94-3392, 2011) and FreightCar Am., Inc., No. 18-0970, 2021 WL 2311871 (OSHRC Mar. 3, 2021). (UHS-DE Br. 4, 28-30, 34, 46, 51.) Both cases are cited and distinguished in UHS-DE/Pembroke and UHS-DE/Suncoast. 2022 WL 774272, at *5-6; 2023 WL 2388069, at *4, 8, 10. The same analysis in those cases is applicable here.

I.Jurisdiction

II.Factual Background

Overview of Corporate Structure

The Worksite

Physical Layout

Buildings

Nurses’ Stations

Staffing

Staffing Grid

Staffing for Acuity

Working alone

Police & Security

Dealing with aggression at the Worksite

Workplace Violence Prevention Plan

Code Greens

Communication Devices

Patient Management

Restraint Reduction Initiative

Injury Reporting & Tracking

Debriefing & Incident Investigations

Contraband & Patient Searches

OSHA’s Investigation

Video Surveillance

Factual Background

Procedural Background

Legal Standard

Failure to Take Reasonable Steps

Restoration or Replacement

Prejudice from the ESI Spoliation

Redress of Prejudice

Intent to Deprive / Bad Faith

III.OSHA Has Authority Over the Worksite and Cited Hazard

Scope of Authority

Other Regulatory Authorities

Due Process

Role of OSHA Guidance About the Hazard of Workplace Violence

IV.Expert Testimony

Dr. Forman

Dr. Lipscomb

Dr. Staggs

The Opinions of the Secretary’s Experts Are Entitled to More Weight

V.Analysis

General Duty Clause

Employee Exposure to the Cited Hazard

Respondents’ Knowledge and Recognition of the Cited Hazard

Serious Physical Harm

Abatement

Existing Abatement

Persistent Occurrence of Injury

Patient Management

Respondents’ Other Flawed Methods to Reduce Workplace Violence

Respondents’ Existing Abatement Was Inadequate

Feasibility and Effectiveness of the Proposed Abatement

Multi-Step vs. Alternatives

Overview of Effectiveness & Feasibility

Assessment of the Effectiveness and Feasibility of the Proposed Abatement Methods

Engineering Controls

Administrative Controls

Feasibility and Effectiveness Established

Single-Employer

Common Worksite

Interrelation and Integration

Management Agreement

Personnel

Training

Common Ownership and Management

Single-Employer Established

Citation 2, Item 1 – Failure to Timely Produce Records

Penalties