United States of America


1120 20th Street, N.W., Ninth Floor

Washington, DC 20036-3457















Kate O’Scannilain, Solicitor of Labor

            Michael D. Felson, Regional Solicitor

            Paul Spanos, Senior Trial Attorney

        U.S. Department of Labor, Office of the Solicitor, Boston, Massachusetts

                             For the Complainant


Carla J. Gunnin

Jackson Lewis P.C., Atlanta, GA

                                For Respondent


Before:  Covette Rooney, Chief Administrative Law Judge



  Following a complaint, the Occupational Safety and Health Administration (OSHA) commenced an investigation of Arbour-HRI Hospitals Inc.’s (HRI) facility in Brookline, MA. At the end of the investigation, HRI received a Citation alleging it violated 29 U.S.C. § 654(a)(1), a provision commonly referred to as the general duty clause, for exposing employees to the hazard of being physically assaulted by patients. HRI timely contested the Citation and a hearing was held from November 28, 2017, through November 30, 2017. (Stip. 4.) For the reasons discussed, the Citation is vacated, and no penalty is assessed.


  The record establishes that Respondent filed a timely notice of contest and that it is an employer affecting commerce within the meaning of the Occupational Safety and Health Act of 1970, 29 U.S.C. § 659(c) (the Act). (Stip. 1, 3, 4, Ans. at 2, 3.) Based upon the record, the undersigned concludes that the Commission has jurisdiction over the parties and subject matter in this case, and Respondent is covered under the Act. (Stip. 2.)


  HRI is a 62-bed inpatient behavioral health hospital located in Brookline, MA. (Tr. 41, 196; Ans. at 2.) It treats patients with psychiatric disorders and illnesses. (Tr. 41-43, 198.) Patients only are referred to HRI if they are unable to be treated in a less restrictive setting, such as an out-patient clinic. (Tr. 198-200, 269.) Some patients have a history of violence or aggression before they enter the facility. (Tr. 198, 720, 827; Ex. R-24.) Other patients exhibit these behaviors while at the facility even without a known history of violence or aggression. (Tr. 234, 365.)

  After a complaint describing aggressive actions taken by patients against staff was filed with OSHA, the agency commenced an investigation of HRI on July 29, 2016. (Tr. 57, 439-40, 692.) A compliance officer (CO) visited the worksite on August 3, 2016, and one additional time before the close of the investigation on January 19, 2017.1 (Tr. 692.) OSHA’s investigation included an examination of HRI’s approach to the issue of workplace violence, including aggression by patients against staff. (Tr. 693-94.)

After the investigation concluded, a Citation was issued to HRI, alleging that nurses and mental health workers at HRI: “were exposed to acts of workplace violence while working with patients who presented aggressive behavior.” This aggressive behavior allegedly included: “direct attacks involving punches, kicks, scratches, being hit with objects such as a soda bottle that was bounced off a wall,” “being hit with a drawer that a patient pulled out of dresser,” and being bitten, hit and/or kicked by patients. The Citation indicates that HRI failed to adequately protect against the hazard of employees being physically assaulted by patients. To abate the hazard, the Citation sets out several categories of actions to better protect employees from aggressive behavior and mitigate the harm from such incidents.


The Citation alleges a violation of the general duty clause, which requires every employer 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.” 29 U.S.C. § 654(a)(1). (Jt. Pre-Hr’g Stmt. at 7.) To establish a general duty clause violation, 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. Waldon Health Care Ctr., 16 BNA OSHC 1052, 1058 (No. 89-3097, 1993). (Jt. Pre-Hr’g Stmt. at 7.)

According to the Secretary, (1) physically aggressive patients presented a hazard to employees; (2) HRI and the behavioral health industry recognized this hazard; (3) HRI employees suffered serious injuries while providing care to aggressive individuals; and (4) there are feasible and acceptable means of addressing the hazard of patient on staff violence at the worksite. For the reasons set forth below, the undersigned finds that the Secretary met his burden with respect to the first three elements of the test. However, given the abatement methods already in place at HRI, the Secretary failed to establish that the additional steps he proposes would eliminate or materially reduce the hazard. See Mo. Basin Well Serv., Inc., 26 BNA OSHC 2314, 2319 (No. 13-1817, 2018) (finding that in the context of a general duty clause violation when the employer has undertaken measures to address the cited hazard the Secretary must show that such measures were inadequate); Cerro Metal Prods. Div., Marmon Grp., Inc., 12 BNA OSHC 1821, 1825 (No. 78-5159, 1986) (vacating a general duty citation when employer took all requisite steps to abate known hazard).

I.Patient on Staff Violence is a Hazard at HRI

Generally, workplace violence refers to violent acts that happen in the workplace, including assaults and threats of assault. (Tr. 578.) Workplace violence can involve co-workers, or, as the Secretary alleges was the situation at HRI, patients. (Exs. R-2, C-30 at 1.) There is no specific OSHA standard addressing the hazard of workplace violence. This does not mean that employers have no obligation to address the hazard. Rather, if an employer or its industry recognize that workplace violence is an actual or potential hazard that can cause death or serious physical harm, the Act’s general duty clause requires such employers to act to eliminate or materially reduce the hazard. 29 U.S.C. § 654(a)(1); Megawest Fin., Inc., No. 93-2879, 1995 WL 383233, at *6 (O.S.H.C.A.L.J., May 8, 1995) (finding that the Secretary could assert that workplace violence constitutes a general duty clause violation). “[T]here is no requirement that there be a ‘significant risk’ of the hazard coming to fruition.” Waldon, 16 BNA OSHC at 1059-60. Instead, when determining the existence of a hazard, the inquiry is whether there is a significant risk to employees if the hazardous event occurs. Id.; Kelly Springfield Tire. Co., 10 BNA OSHC 1970, 1973 (No. 78-4555, 1982) (focus is on the hazard, not the specific incident that resulted or might have resulted in injury).

  HRI does not deny its employees experienced violence and that the hazard of patient on staff violence remained at the facility despite its abatement efforts. The CEO explained that the facility’s patient population includes individuals who are aggressive and may harm others. (Tr. 198.) Indeed, to qualify for admission to the facility, patients must be considered at risk of harming themselves or others. (Tr. 198-99.) Further, several employees had been injured by patient aggression in the past. (Tr. 48, 95; Exs. C-1, C-9, C-10 at 5.) The Secretary presented evidence of five specific incidents of patient aggression towards staff from May 11 through August 21, 2016, and company records refer to additional incidents. (Tr. 136, 440-41, 804; Exs. C-1, C-8, C-9, C-10 at 5, C-18 at 3.) One such incident involved a patient attacking an employee with a dresser drawer. (Tr. 48-49.) The attack left the employee with a fractured wrist and a damaged tendon in his knee. (Tr. 99; Ex. C-1.) The employee promptly reported the incident, consistent with HRI’s procedures. (Tr. 48-49; Ex. C-1.) HRI managers were aware of this and other acts of aggression against employees. (Tr. 45, 51-52, 584; Exs. C-1, C-9.) Thus, at HRI, “a future violent incident was neither freakish nor implausible.” See Megawest, 1995 WL 383233, at *8-9 (finding that the hazard of workplace violence existed in the cited employer’s facility); Waldon, 16 BNA OSHC at 1059-60. The Secretary established that patient on staff violence can occur at this facility and it presents a hazard to HRI’s employees.

II.HRI and the Behavioral Health Industry Recognized the Hazard of Patient on Staff Violence

The second prong requires showing either that the employer had actual knowledge a condition or activity was hazardous, or that the industry knew the activity or condition was hazardous. SeaWorld of Florida, LLC v. Perez, 748 F.3d 1202, 1208-9 (D.C. Cir. 2014); Waldon, 16 BNA OSHC at 1061 (A hazard is ‘recognized’ within the meaning of the general duty clause if the hazard is known either by the employer or its industry”). Actual knowledge of a hazard by an employer satisfies this requirement. Puffer’s Hardware, Inc. v. OSHRC, 742 F.2d 12, 18 (1st Cir. 1984) (finding that the Secretary showed the hazard was recognized because the employer warned employees its elevator was hazardous). An employer may gain actual knowledge from prior episodes, employee complaints, or warnings communicated to the employer.  Id.; St. Joe Minerals Corp. v. OSHRC, 647 F.2d 840, 845 (8th Cir. 1981) (employer had both actual and constructive knowledge of the hazard when previously warned by at least one employee); Pepperidge Farm, Inc., 17 BNA OSHC 1993, 2004, 2030-31 (No. 89-0265, 1997) (finding recognition based on memoranda and warnings from employees).

HRI provides care for patients with “aggressive or assaultive tendencies” and some patients can be violent. (Tr. 41-44, 198-200.) The patient admissions process incorporates probing for whether a patient previously exhibited violent behavior or if there was a reason to suspect a propensity to violence. (Tr. 212, 453; Exs. R-22, R-24.) Such a history did not preclude admitting the patient but might alter the observation level assigned to the patient or result in other precautions. (Tr. 315, 452-54; Exs. R-22, R-26.) Those patients perceived to be at a higher risk of violence and those who act aggressively receive fewer privileges. (Tr. 303, 315, 454, 457-62, 740; Ex. R-26.) HRI also was aware that even previously non-violent patients may still act aggressively towards staff. (Tr. 437.)

HRI documented, tracked, and trended incidents of violence against staff. (Tr. 51-52, 207, 210, 584.) For example, at its meeting a few days before the CO visited the facility, the Environment of Care Committee discussed the fact that there had been twelve reported employee injuries related to actions taken by patients in the second quarter of 2016. (Ex. C-19 at 2; Tr. 440-41.) The incidents include being hit with objects as well as being punched and kicked. (Ex. C-9; Tr. 45.) Whenever patient aggression results in an employee injury the affected employee must report the incident by completing an Employee Accident Report (EAR). (Tr. 85-86; Ex. C-7.) HRI tracks the EARs and has a process to make management aware of such events. (Tr. 50-51, 53, 350, 584; Exs. C-1; C-9.) And, at the hearing, HRI’s Risk Manager, Marie Kong, acknowledged that she recognized patients could potentially harm staff, possibly seriously. (Tr. 44-45.) She was aware that staff was exposed to being punched, kicked, and attacked with objects. (Tr. 45, 139.)

HRI’s training and written policies also support finding that HRI recognized the hazard of patient on staff violence. See SeaWorld, 748 F.3d at 1206 (training manuals, safety lectures and incident reports provided “abundant” evidence that employer recognized the hazard); Beverly Enters., Inc., 19 BNA OSHC 1161, 1186 (No. 91-3344, 2000) (consolidated) (evidence of precautions taken can support a finding of hazard recognition). HRI provided specific training on workplace violence and how to manage patient behavior. (Exs. R-2, R-20, C-5; Tr. 704.) The behavior management component of HRI’s training program is called Handle With Care. (Exs. C-5, R-12; Tr. 757.) This program’s training manual includes training on personal defense, such as how to defend yourself when someone tries to punch you or attack you with an object, and how to physically restrain a patient. (Ex. C-5; Tr. 758.) In addition to teaching these techniques, HRI also had a system for staff to call for additional assistance when they could not manage a patient's aggression by themselves. (Ex. C-4.) Asking for assistance to manage a patient’s behavior was referred to as calling for a Code Green. (Tr. 66.) The policies related to calling a Code Green were supplemented by several other written policies addressing actual violent events as well as the potential for patient on staff violence. (Tr. 170, 315-16, 410; Exs. C-2, C-6, R-4, R-27, R-29, R-30, R-48, R-49.) See Waldon, 16 BNA OSHC at 1061 (noting that protective measures implemented by an employer can be used to support finding that the employer recognized the hazard).

Like HRI, other providers of behavioral health services are also aware of the hazard of patient on staff violence. The Secretary’s expert, Dr. Lipscomb, was accepted as an expert in workplace violence prevention in healthcare settings, including behavioral health facilities. (Tr. 578.) She has over fifteen years of field research experience, has been involved with many teams focused on reducing workplace violence in the healthcare setting, and published several papers on the topic. (Tr. 539-4.) She explained that the industry has long known of the problem of patient on staff violence. (Tr. 585.) See ACME Energy Servs., 23 BNA OSHC 2121, 2124 (No. 08-0088, 2012) (“Industry recognition may be shown through the knowledge or understanding of safety experts familiar with the workplace conditions or the hazard in question”), aff’d, 542 F. App’x 356 (5th Cir. 2013) (unpublished); Waste Mgmt. of Palm Beach, 17 BNA OSHC 1308, 1310 (No. 93-128, 1995) (expert “familiar with the general workplace condition” established recognition). As support for her position, in addition to her own experience, she also pointed out that OSHA, the Centers for Disease Control (CDC), and the Federal Bureau of Investigation (FBI) each issued guidelines on workplace violence. (Tr. 585; Exs. C-27, C-28, C-29.) OSHA’s 1996 Guidelines for Preventing Workplace Violence for Healthcare and Social Service Workers (OSHA Guidelines) notes that while no diagnosis predicts future violence, inpatient psychiatric patients are among the groups that present the highest risk for such episodes. (Ex. C-27 at 9-10.) The guidelines from the CDC and FBI also specifically discuss the risk of patient assaults against care providers. (Exs. C-28 at 15-17, 29; C-29 at 12-14, 54-56.)

HRI does not refute the Secretary’s position that both it and the industry as a whole recognize the hazard of patient on staff violence. See Pegasus Tower, 21 BNA OSHC 1190, 1191 (No. 01-0547, 2005) (finding that employer recognized the dangers associated with an activity and affirming a general duty clause violation). The company's expert, Monica Cooke, was accepted as an expert on risk reduction in healthcare settings for workplace violence, specifically patient to staff aggression in a behavioral health hospital setting. (Tr. 871.) Ms. Cooke concurred with Dr. Lipscomb that behavioral health facilities generally recognize the hazard of patient on staff violence in their facilities. (Tr. 892, 917.) And, she also agreed that HRI, in particular, recognized the hazard of patient on staff violence at its facility.2 (Tr. 891, 916-17.) For these reasons, the Secretary met the second criterion for finding a violation of the general duty clause.

III.Workplace Violence Was Causing Serious Physical Harm

The third prong of the general duty clause test requires the Secretary to show that the hazard was causing or was likely to cause death or serious physical harm. Waldon, 16 BNA OSHC at 1060. This inquiry does not turn on the likelihood of an accident or injury. Id. Instead, the focus is on whether, if an accident occurred, are the results likely to be death or serious harm. Id.

Ms. Cooke acknowledged that patients at Respondent’s facility could inflict serious harm on employees. (Tr. 892.) Beyond this theoretical risk, the Secretary also established that HRI employees had been injured by patients in the past. (Tr. 48-49, 94, 220-21, 366, 579; Exs. C-1, C-9, C-10, C-19 at 2.) Some of these incidents required medical treatment, including hospitalization and long-term follow up care. (Ex. C-19 at 2-3; Tr. 99, 127, 805-6.) One worker suffered a head and eye injury that required medical attention. (Tr. 94, 948-49.) Another suffered a head and neck injury that required months of physical therapy. (Tr. 805-6.) The Secretary showed if the hazard of patient on staff violence occurs, it can result in serious physical harm or death. See Waldon, 16 BNA OSHC at 1060.


Having found that patient on staff aggression is a recognized hazard at HRI that could lead to a serious injury, we turn to whether the Secretary satisfied the abatement prong of the test. See Waldon, 16 BNA OSHC at 1062. The Secretary concedes that the hazard of patient on staff violence cannot be eliminated but argues that there are feasible means to materially reduce the number of incidents and their severity. See Arcadian Corp., 20 BNA OSHC 2007, 2011 (No. 93-0628, 2004) (finding that the Secretary only has to show that the abatement method would materially reduce the hazard, not that it would eliminate the hazard); Morrison-Knudson Co./Yonkers Contracting Co., A Joint Venture, 16 BNA OSHC 1105, 1122 (No. 88-572, 1993) (same).

In response, HRI argues it took many precautions that collectively adequately address patient aggression towards staff. (Resp’t Br. at 24-44.) And it contends the Secretary failed to establish that the additional abatement steps he proposes would be effective in materially reducing the hazard. Id. at 50-65.

A.Feasible Means to Materially Reduce the Hazard Exist

The Secretary does not propose any action or set of actions which HRI can take to eliminate the hazard of patient on staff violence. (Sec’y Br. at 43.) Nonetheless, he argues that multiple measures can materially reduce violence against staff. Id. at 44. Such a showing would be sufficient to find a violation. See Arcadian, 20 BNA OSHC at 2011 (Secretary satisfies the abatement requirement by demonstrating feasible actions that can materially reduce a hazard).

Both sides presented expert testimony regarding abating the hazard of patient on staff violence. Dr. Lipscomb testified on behalf of the Secretary and argued that steps can be taken to materially reduce the cited hazard. She cites a peer-reviewed randomized study led by Judith Arnetz, Ph.D. (hereafter, Arnetz Study)3 which concludes that by implementing certain measures suggested by OSHA’s Guidelines medical facilities were able to reduce the rate of increase in patient on staff violence. (Tr. 585; Exs. C-27, C-30.) In her view, the Arnetz Study was rigorous and conducted by well-recognized researchers. (Tr. 562.) It involved forty-one units at seven different hospitals being randomly assigned to either intervention or non-intervention groups. (Ex. C-30 at 1.) The intervention group implemented various strategies set forth in the OSHA Guidelines. (Tr. 608-9; Exs. C-28, C-30.) Each unit did not have to implement the same strategies. (Tr. 611-12.) They were given a “menu” of options and could select what they considered to be “the most appropriate strategy” based on the unit’s unique features. (Tr. 611.) The Arnetz Study concludes that the interventions were effective in decreasing risks of patient to worker violence and related injuries. (Tr. 608-12; Exs. C-27, C-30.) Dr. Lipscomb argued that the Arnetz Study supported her view that there were feasible steps HRI could take to materially reduce the hazard of patient on staff violence. (Tr. 585.)

Respondent replies that the results of the Arnetz Study were published after the inspection date and should not be considered. (Resp’t Br. at 3.) However, the Secretary does not rely on the Arnetz Study to support finding HRI or the industry recognized the hazard. Rather, the papers relating to the Arnetz Study were admitted as something Dr. Lipscomb relied on when she assessed whether actions can reduce the hazard of patient on staff violence at HRI and what those actions should be. (Tr. 563-64; Exs. C-30, C-31.) The interventions reviewed as part of the Arnetz Study were based on the OSHA Guidelines first published in 1996 and updated in 2004 and 2015, well before the inspection. (Tr. 467, 624; Exs. C-27, C-30, C-31.) And, Respondent did not identify information in the Arnetz Study that was unavailable or technologically or economically infeasible prior to the issuance of the Citation. (Tr. 566.)

Besides the timing of the publications, Respondent also points out limitations with applying the Arnetz Study to HRI. (Resp’t Br. at 45-47.) First, none of the units in the intervention group were standalone facilities focused on providing psychiatric care, like is the case at HRI. (Tr. 620; Ex. C-30 at 7.) Second, because the facilities in the intervention group could select from a variety of environmental, administrative, and behavioral strategies, the Arnetz Study does not conclusively establish the effectiveness of any one intervention. (Tr. 611-12; Ex. C-30 at 7.) Dr. Lipscomb agreed with this assessment, conceding that the Arnetz Study did not show there was a single “silver bullet” that could be used to address the hazard. (Tr. 623.) Rather, in her view, the study shows the need for a comprehensive program that considers a facility’s particular features. (Tr. 622-23, 686-87.) Third, Respondent argues that prior to the inspection, it implemented some of the intervention strategies noted in the Arnetz Study, particularly safety monitoring, consults with psychiatry, de-escalation training, and behavior management classes. (Resp’t Br. at 45-46.) So, in Respondent’s view, taking some of the other actions on the available “menu” of interventions would not necessarily reduce the hazard further. Id.

  The undersigned accepts the Arnetz Study as something Dr. Lipscomb relied on to formulate her opinion but agrees with Respondent that it does not answer whether HRI’s abatement was sufficient. (Ex. C-30; Tr. 563-66.) Nor does the Arnetz Study establish that the Secretary’s proposed methods of abatement would materially reduce the hazard beyond what has been accomplished by HRI’s methods. See Beverly, 19 BNA OSHC at 1190 (proposed abatement measures must be “capable of being put into effect” and “effective in reducing the incidence of the hazard”). However, the Arnetz Study does bear on the threshold issue of whether there are feasible means to address the hazard of patient on staff violence. And on that issue, Respondent’s expert, Ms. Cooke, appeared to agree that there are many ways to mitigate the risk of patient to staff aggression, such as through employee training and by searching for contraband. (Tr. 886-87, 920.) Such measures were part of HRI’s program to manage workplace violence. (Exs. R-2, R-6, R-20; Tr. 681, 695, 918.) HRI argues that the steps it took were more than adequate to address the hazard. (Resp’t Br. at 25-28.) By arguing its approach was effective, HRI implicitly acknowledges that providers of behavioral health services can take steps to mitigate the hazard of patient on staff violence. Further, as the company has already taken these steps, there is no basis for arguing they could not be implemented. (Exs. R-2, R-6; Tr. 681, 695.) See SeaWorld, 748 F.3d at 1215 (finding that evidence of post-citation methods of abatement taken by employer support finding that the proposed means were feasible). While it may not be possible to eliminate patient on staff violence at a behavioral health facility, both experts indicate there are ways to reduce such incidents and limit their severity. (Tr. 585, 886-87.)

B.Sufficiency of HRI’s Abatement

Having found there are steps that can be taken to reduce the hazard of patient on staff violence, we turn to the adequacy of HRI’s approach. See Mo. Basin, 26 BNA OSHC at 2319; U.S. Postal Serv., 21 BNA OSHC 1767, 1773-74 (No. 04-0316, 2006) (Secretary must show that the methods the employer used to address the hazard were inadequate). HRI maintains that everything it does in its day to day operations contributes to reducing patient aggression towards staff. (Resp’t Br. at 25; Tr. 54-55, 205.) It contends it has a robust workplace violence prevention program and argues that the Secretary failed to show the measures it took were inadequate to prevent the hazard of patient on staff violence. (Resp’t Br. at 24-43.) The Secretary disputes this, contending that: (1) HRI’s workplace violence training was insufficient, and (2) it lacked adequate policies and procedures to materially reduce the hazard of patient on staff violence. (Sec’y Br. at 48-51.)


  The Secretary argues that HRI’s training was inadequate and asserts that HRI could materially reduce the hazard by adopting more effective training. (Sec’y Br. at 50.) HRI does not deny that it can train employees on workplace violence prevention strategies and that doing so is an effective means of reducing the hazard of patient on staff violence. (Tr. 310, 493.) The company has a training program for all staff, including those without direct care responsibilities, before they begin work. (Tr. 171, 204, 445; Ex. R-20.) The training includes a presentation directly about workplace violence as well as on subjects related to the hazard. (Ex. R-2; Tr. 309-12, 718, 722.) For example, HRI trains employees about how to identify potentially aggressive patients, how to reduce the likelihood a patient will become agitated, and how to de-escalate situations if the patient is already agitated.  (Tr. 310-11, 314, 325, 441, 682, 718-19, 723-24, 729-745, 750, 755; Exs. R-12, R-13.) The training includes a full day on de-escalation techniques. (Tr. 750-56; Exs. R-13, R-20.) Employees are tested at the end of the training and any wrong answers are discussed. (Tr. 755.) HRI trains staff on what it refers to as “trauma-informed care.” (Tr. 707-8; Ex. R-20 at 1.) Under this approach, employees are taught to anticipate that patients may act aggressively even if their medical history does not indicate trauma or past aggression.  (Tr. 707-8.) HRI teaches employees behavior management techniques, including when and how to restrain patients. (Tr. 77, 211, 310, 314, 325, 593, 749, 757; Exs. R-12, R-20.) They are trained to identify and remove contraband, including by using metal detectors. (Tr. 742-43, 748; Exs. R-20 at 2, 4, R.61.) The training program includes both classroom instruction and hands-on learning. After receiving instructions from a nurse educator, new employees “shadow” staff before taking on direct care responsibilities.4  (Tr. 761, 777.) They then must demonstrate competency in topics related to workplace safety before they can engage in direct patient care. (Tr. 761-63, 772-78; Exs. R-10, R-11.)

In addition to the new employee training, HRI also requires annual re-certification on certain topics, such as verbal de-escalation and restraints. (Tr. 756, 758.) Re-training is also provided if called for based on an incident or if requested by an employee. (Tr. 722, 756-59.) For example, the Restraint Elimination Committee will refer people for additional training if they think it is needed based on what they see when reviewing camera footage of an incident. (Tr. 512-15.) Every month, HRI tracks how many trainings are offered and reviews training records to make sure employees are receiving the necessary training.  (Tr. 513-15; Ex. R-57.)

According to Respondent’s expert, Ms. Cooke, HRI’s staff received several more hours of annual training on managing patients than what is typical in the behavioral health industry.  (Tr. 887-90.) Dr. Lipscomb does not refute this. In terms of the training’s substance, Dr. Lipscomb agreed that the information provided on de-escalation was good. (Tr. 642.) She acknowledged that HRI’s training on de-escalation and patient aggression was “well received by the workers.” (Tr. 604-5, 681-82.) She did not suggest that any mental health worker or nurse was not doing his or her job adequately. (Tr. 667.)

Still, Dr. Lipscomb called for “additional” and “more effective training.” (Tr. 604-5.) She suggested HRI could correct the alleged deficiency by ensuring that the skills covered in HRI’s verbal de-escalation and restraint training were taught in the context of a “larger program” that defines patient on staff violence and indicates what the process is for reporting incidents and calling codes. Id. Specifically, she argued that HRI’s training did not: (a) define workplace violence, (b) explain when to call a code, or (c) include adequate information about incident reporting. (Tr. 604-5, 684.) But, as explained below, the Secretary did not show that HRI’s program inadequately addressed these topics.

a)Defining Workplace Violence

First, with respect to whether HRI’s training includes a definition of workplace violence, HRI produced a collection of slides from a presentation to new employees. (Ex. R-2.) The slides expressly define workplace violence as “any physical assault, threatening behavior or verbal abuse occurring in the work setting.” (Ex. R-2; Tr. 635.) And they provide various examples of workplace violence events, including violence by patients. (Ex. R-2; Tr. 636, 640-74.) The slides include information about risk factors for workplace violence, describe the warning signs of violence, and suggests precautions employees should take. (Ex. R-2; Tr. 637-40, 642-43.) Consistent with HRI’s other training, the slides include a reminder that employees should report all incidents of workplace injuries, even if minor. (Tr. 639-40, 643-44; Exs. R-2, R-4, C-6, C-7.)

The CO acknowledged receiving these slides before the close of her investigation and believed they were part of HRI’s orientation program for new employees. (Tr. 698.) Still, she noted that employees did not mention whether this presentation in discussions about their training. Id. But, HRI’s nurse educator explained that the presentation was part of orientation for new employees.5 (Tr. 703, 709-10.) He himself was trained by another employee using the slides and he now conducts training on workplace violence using the same slides. (Tr. 709-22; Ex. R-2.) His account of the training is supported by the agenda for new employee training, which explicitly refers to the topic of “workplace violence.” (Ex. R-20; Tr. 709.)

Dr. Lipscomb did not appear to recall reviewing the presentation and acknowledged that she did not ask any of the former employees she spoke to about whether they were trained on the workplace violence prevention slides that make-up the exhibit. (Tr. 632-35, 644; Ex. R-2.) She agreed that the presentation contained an appropriate definition of workplace violence. (Tr. 635.) And she considered the information included in the written training materials on preventing and diffusing aggressive behavior contained in the training to be good advice. (Tr. 641-42, 645.) Accordingly, the Secretary fails to prove that employees were not trained about what workplace violence is.

b)Training About Calling for Assistance (Code Greens)

The second deficiency alleged by the Secretary relates to teaching employees about when to call for assistance. HRI calls its procedure for getting assistance in the event of a psychiatric emergency “Code Green.” (Tr. 67, 316; Ex. C-4.) An actual or threatened assault on a staff member may trigger a Code Green. (Tr. 319.) The Code Green policy is part of HRI’s workplace violence procedures.6 (Tr. 316.) To initiate a Code Green, a staff member uses a telephone intercom system to announce a Code Green and identify the location of the emergency within the facility. (Tr. 70-71.) Employees can also call out “Code Green” verbally to their co-workers instead of using the phone. (Tr. 318.) Various staff, including mental health workers and a social worker, are required to respond to the Code Green and perform assigned roles.7 (Tr. 70, 277, 317-19; Ex. C-4.) Code Greens can involve attempts to engage in further verbal de-escalation or restraining a patient if necessary. (Tr. 443.) Staff is made aware of any Code Greens even if they were not on the unit when it occurred. (Tr. 323, 415.) HRI’s Code Green policy was in effect at the time of OSHA’s inspection and is periodically reviewed. (Tr. 69, 419.)

According to Dr. Lipscomb, a former employee expressed some confusion about when to call a Code Green. (Tr. 603, 605.) This employee was attacked by a patient after he responded to some loud noises. (Tr. 603.) As part of the post-incident de-briefing, a manager explained to the employee that he should have gotten assistance before investigating alone. Id. After this incident, in the words of Dr. Lipscomb, the employee “overcompensated” during a subsequent incident and called an unnecessary code. Id.

There is no debate that HRI trains employees on when to call a Code Green and how to do it. (Tr. 424, 447, 886; Ex. C-4.) HRI’s Risk Manager acknowledged that knowing when to call a Code Green was important. (Tr. 428-29.) She made clear HRI wants employees to call a Code Green when there is an emergency and that management would rather have employees err on the side of calling a Code Green. Id. She explained the company’s desire for employees to call a Code Green early in the process of addressing an aggressive patient before they engage in any restraint technique. (Tr. 330.)

The Secretary does not identify what in HRI’s policies or procedures was unclear or confusing. Like Dr. Lipscomb, Ms. Cooke also reviewed HRI’s training program, including its Code Green procedures. (Tr. 873.) She discussed it with current employees, who indicated that training was very well done, informative, and helpful on a daily basis. (Tr. 873-74.) As noted above, the training included classroom instruction, written policies, shadowing workers, and testing. (Tr. 761, 777; Exs. C-4, R-10, R-11, R-20.) At least one member of senior leadership is also required to attend the code if it occurs during business hours and HRI tracks all incidents to ensure this occurs. (Ex. R-57.) In addition, HRI conducts video reviews of codes and sends employees for further training if it identifies deficiencies in how they responded to a code. (Tr. 510.) The Secretary neglected to sufficiently articulate how training on HRI’s Code Green policy could be structured differently so as to materially reduce the hazard.8 See Cerro Metal, 12 BNA OSHC at 1825 (finding that the Secretary failed to show there was a “specific, feasible additional step” the employer could have taken to improve communication of the work rule).

c)Training About Incident Reporting

The third area of deficiency in training the Secretary cites relates to incident reporting. (Sec’y Br. at 49-50.) HRI trains employees on reporting both staff injuries and patient aggression.  (Tr. 708-9; Ex. R-20.) It instructs employees to report all incidents, including those involving patient aggression towards staff, “right away.” (Tr. 708-9, 715, 722; Ex. R-2 at 8.) Accidents must be investigated, and employees are instructed to attempt to remedy any unsafe condition or practice.  (Exs. C-7, R-2 at 5; Tr. 708, 717, 351.) Employees are also told that reports of violence will be treated seriously and confidentially. (Tr. 717; Ex. R-2 at 5.) The procedures for reporting accidents are posted in several places throughout the facility. (Tr. 347-48; Ex. C-6.) All nurses and mental health workers must demonstrate competency in understanding incident reports before they can begin their regular duties. (Tr. 761, 774, 780; Exs. R-10, R-11.) There is no evidence of employees failing to report incidents, or not understanding that they needed to be reported.9 Thus, the Secretary failed to show that HRI’s training about incident reporting was inadequate.

d)Burden Not Met

Ms. Cooke did not agree with Dr. Lipscomb’s view that HRI needed to provide more effective training. (Tr. 887-88.) She argued that both the amount and scope of HRI’s training went beyond what she typically encounters in behavioral health facilities. (Tr. 888-90.) The Secretary does not establish any topic relevant to the hazard on which HRI employees received no training. Nor does the Secretary show HRI spent an insufficient amount of time on training. Although it may well be true that there are certain actions HRI could take to improve its training, the Secretary needed to identify deficiencies in HRI’s approach and show that the suggested changes would materially reduce the hazard.10 See Ala. Power Co., 13 BNA OSHC 1240, 1245 (No. 84-357, 1987) (Secretary failed to show safety rules were inadequate).

2.Policies and Procedures

In addition to alleging deficient training, the Secretary also argues that HRI’s policies and procedures were insufficient. (Sec’y Br. at 45-46, 48-51.) He asserts that HRI needed: (1) a comprehensive written workplace violence prevention program that included a workplace violence prevention coordinator, incident reporting, incident tracking, and staff involvement; and (2) adequate staffing. Id. According to the Secretary, having a comprehensive workplace violence prevention program and adequate staffing were necessary and feasible to materially reduce the hazard. Id.

a)Comprehensive Written Workplace Violence Prevention Program

  HRI first suggests that the Secretary failed to establish that having any program materially reduces the hazard. (Resp’t Br. at 63.) But, Dr. Lipscomb and HRI’s CEO have a different view. Dr. Lipscomb explained that a written plan was an important mechanism to ensure everyone in the organization had the same understanding and was working toward the same goal of reducing workplace violence. (Tr. 606-7.) Her view is supported by her experience in the field as well as general guidance documents on workplace violence published by OSHA, the CDC, the FBI, and the American National Standards Institute (ANSI). (Exs. C-27 at 11, C-28 at 23, 27, C-29 at 55.) HRI’s CEO agreed with Dr. Lipscomb about the importance of having policies and procedures related to workplace violence. She argued that the written policies that make-up HRI’s workplace violence prevention program “prevents injuries to patients or to staff.” (Tr. 170.) In her view, the company’s policies work together to reduce patient aggression and improve safety. (Tr. 170-71.) Thus, the company’s own experience corroborates Dr. Lipscomb’s view that written policies and procedures reduce the hazard.

Although the Secretary showed that a written workplace violence prevention program can reduce the hazard, under Commission precedent, the Secretary also needed to show that HRI’s approach was inadequate. See Mo. Basin, 26 BNA OSHC at 2319. Cf. Indus. Glass, 15 BNA OSHC 1594, 1604 (No. 88-348, 1992) (recognizing that heat can be a hazard but concluding that the company’s effective measures to reduce exposure and alleviate its effects precluded finding a violation of the general duty clause). And as to that issue, HRI argues it has a comprehensive written program to address workplace violence both directly and indirectly. (Tr. 55, 170, 986, Ex. C-2.) HRI’s “Workplace Violence” policy states that the facility has “zero tolerance for violence of any kind” and goes on to indicate that management is “committed to providing an environment free from all forms of violence including … physical threats or acts of physical assault.” (Tr. 169; Ex. C-2 at 1.) Under this policy’s terms, all reports of violence are to be “treated seriously and fully investigated.”11 (Ex. C-2 at 1.) The Workplace Violence policy is supplemented by others addressing: (1) handling psychiatric emergencies (i.e., Code Green); (2) reporting employee accidents; (3) identifying and assigning different precaution levels for at-risk patients, such as increasing observation levels and/or limiting privileges; (4) staffing; (5) escorting patients outside of the unit; (6) conducting searches of patients, patient areas, and patient possessions for prohibited items (contraband); (7) using restraints with patients and debriefing staff after any such incident; (8) visitors; and (9) patient responsibilities.  (Exs. C-4, C-7, R-22, R-26, R-27, R-29, R-30, R-35, R-38, R-48, R-49; Tr. 170-71, 308-9, 315-16, 410, 499.) HRI maintained the various written policies that collectively constitute its written workplace violence prevention program in a binder, as well as online where they were all accessible to staff through a shared drive.12 (Tr. 54-58, 314-15, 410, 443-44.) These policies were regularly reviewed and updated.13 (Tr. 69, 85, 452, 467, 472, 485.) According to HRI’s CEO, these policies work together to mitigate the hazard of patient on staff violence. (Tr. 169-71.)

Beyond the written policies and procedures HRI’s workplace violence prevention program also included: (1) a management commitment to reducing patient aggression against staff; (2) employee participation in addressing safety, including workplace violence; (3) analyzing for things that could contribute to the hazard of patient on staff violence; (4) preventing and controlling patient aggression when it occurs; (5) maintaining adequate staffing, (6) comprehensive training, including on how to prevent the hazard of patient on staff violence and ways minimize harm from such incidents if they occur. (Tr. 170-71, 209, 708-22; Exs. R-19, R-20, R-57, R-61; Resp’t Br. at 25-40, 63-65.)

The Secretary argues that the program was inadequate because it: (1) lacked a workplace violence coordinator, (2) did not include standardized incident reporting, tracking, and analysis, and (3) did not provide for sufficient employee involvement. (Sec’y Br. at 17-20, 50-51.)

(1)Workplace Violence Prevention Coordinator

HRI does not dispute that management needs to be committed to workplace violence prevention and that people must be assigned responsibilities to implement and ensure compliance with the policies and procedures. (Resp’t Br. at 64-65.) What it disagrees with is the Secretary’s contention that it must have one person with the title of workplace violence coordinator. Id. HRI believed that everyone was responsible for addressing workplace violence. (Tr. 54-55, 204-5, 251, 451.) The risk management group along with the human resources department oversaw the HRI’s program. (Tr. 49-50.) Ms. Kong tracked, trended, and reported to the leadership team about workplace violence trends. (Tr. 50-51, 207.) The clinical director and physicians also tracked and trended incidents. (Tr. 207.) On the units, staff tracked incidents, such as contraband and altercations. (Tr. 780-81.) Multiple managers had specific responsibilities related to workplace violence prevention and intervention when such situations occurred. (Tr. 75, 207, 277, 293; Exs. C-2, C-4, C-6.) These managers, including the CEO, reviewed all incident reports, such as those related to the hazard of patient on staff violence, in their daily Flash meetings. (Tr. 211-13.) These Flash meetings also address incidents of patient aggression, the reasons for any staff injuries, and the need for any corrective actions related to thereto. (Tr. 213-14.) Besides those discussions, HRI also has three committees tasked with responsibilities related to mitigating the hazard of patient on staff violence. (Tr. 100, 104-5, 171.) Notes from these committees indicate that patient on staff violence was routinely discussed, mitigating measures were evaluated and various actions were taken. (Tr. 440-41; Ex. R-55.)

The Secretary does not allege that HRI needed a full-time position focused solely on workplace violence prevention. Dr. Lipscomb only called for someone with “dedicated time” to this area. (Tr. 607-8.) The Secretary does not identify any task a workplace violence coordinator would perform that was not already being done by HRI employees at the time of the inspection. He did not explain how one individual in the role of workplace violence coordinator was superior to HRI’s approach of having multiple managers, including the CEO, dedicating time to mitigating the hazard and having specific responsibilities for it. See Pelron Corp., 12 BNA OSHC 1833, 1838 (No. 82-388, 1986) (finding the Secretary failed to prove that addressing the lack of formality in the employer’s safety program would materially reduce the hazard).

(2)Incident Reporting and Tracking

The Citation suggests a feasible method of abatement is for HRI to: “Implement incident reporting and prompt investigations to ensure all reports resulting in injury are promptly investigated with written findings, recommendations, and a plan for corrective action.” As discussed above, HRI already has an incident reporting policy requiring employees to immediately report all accidents and injuries to their supervisor, regardless of their severity. (Tr. 84, 348-49, 722; Exs. C-6, C-7, R-2 at 8.) Unless the employee needs medical attention, the employee must complete a written report, the EAR. (Tr. 349-50.) Threats of violence, as well as incidents of patient aggression that do not result in a staff injury, are also to be reported. (Tr. 177, 296-301.) Management promptly reviews any EARs and they are discussed at the daily Flash meeting. (Tr. 350.) In addition to discussing the EAR, HRI completes a review of any available video footage of the incident. (Tr. 350-51.) HRI uses this camera review and the EARs to facilitate specific follow up as well as to track patient on staff violence at the facility over time. (Tr. 385; Exs. C-6, C-7, C-9, C-10.) Doing so is part of its workplace violence prevention strategy. (Tr. 385.)

HRI also tracks occurrences related to the hazard of patient on staff violence, such as aggressive behavior against other patients and the presence of contraband, even if the occurrences do not result in actual injuries to staff.14 (Tr. 258-61, 383, 396; Ex. R-55.) Managers acknowledged that requiring incident reporting and tracking incidents of violence allows HRI to learn and take both corrective and preventative actions. (Tr. 262, 349-51, 385.)

So, the issue here is not whether reporting and tracking incidents is a feasible means of abatement or part of addressing workplace violence.15 The debate is whether the Secretary showed that HRI was deficient in implementing this workplace violence prevention strategy. To make this showing, the Secretary relies on a document titled HRI Near Miss Log. (Sec’y Br. at 3-5, 12-13; Tr. 92; Ex. C-8.) HRI considers a near miss to be an incident that did not result in staff injury but had the potential to do so. (Tr. 91.) There is no evidence HRI failed to record any such near miss. However, it appears that the HRI Near Miss Log included four incidents involving actual injuries as opposed to “near misses.” (Tr. 92-97; Ex. C-8.)

The Secretary failed to show that this error resulted in any workplace violence risk. The HRI Near Miss Log was not the primary record of workplace violence incidents and there is no evidence the document was relied on to formulate workplace violence prevention decisions. Instead, managers used the more detailed EARs, which provide substantive details and clearly indicate when injuries occurred. (Tr. 95-96; Ex. C-1.) There is no evidence of an actual injury for which no EAR was prepared. Beyond reviewing the EARs, follow up on employee injuries includes reviews of security camera footage if available and employee interviews. (Tr. 215, 351-52, 510.) Three committees also review incident reports, look for trends, and assess potential corrective actions. (Tr. 171-72, 257-62, 284-85, 440-41, 510-19; Exs. R-55, R-57.) Thus, although it appears that the HRI Near Miss Log was over-inclusive, HRI had methods to capture and review information related to both the potential for patient on staff violence and actual injuries. (Tr. 348-52.)

HRI trains employees on how and when to report incidents. (Tr. 708, 722; Exs. R-20, R-2.) It tracks incidents in writing, requires managers to promptly review all incident reports, and makes employees aware of incidents that occur. The CEO explained that she participates in the reviews of incidents and follows up with the committees to make sure they are aware of trends and continuing to review the company’s approach to workplace violence. (Tr. 209-10, 213-14.) She also testified that even if an instance of aggressive behavior did not result in any injuries it would still be discussed at the daily Flash meeting and the need for corrective action would be considered. (Tr. 212-15.) She emphasized the importance of these discussions. (Tr. 214.) Despite some discrete errors in its internal tracking, HRI showed that incidents were reported, analyzed, and followed up on. (Tr. 215; Ex. C-1.) The Secretary did not identify instances where an incident went unreported or where management did not review an instance of patient on staff violence. Nor did he establish that HRI failed to examine the causes of such instances or act on the information learned.

(3)Employee Involvement

The Secretary also argues that HRI should include non-management “employees in the planning, development, and implementation of workplace violence prevention efforts.” (Sec’y Br. at 48-49.) OSHA, the CDC, and ANSI all call for “employee involvement” in workplace violence prevention programs. (Exs. C-27 at 13, C-28 at 21, 26, C-32 at 18.) Dr. Lipscomb explained that “employee involvement” was an important workplace prevention strategy.16 (Tr. 595-96.) She was not asked to define “involvement” but indicated that it required the attendance of direct care staff and non-management individuals at committee meetings and making minutes of the meetings available. (Tr. 596-99.) In his brief, the Secretary argues that to have sufficient employee involvement requires HRI to: (1) schedule meetings at convenient times, (2) provide staff coverage so employees can participate in meetings, and (3) allow “employees to participate in developing the plan and setting the agenda.” (Sec’y Br. at 49.)

Looking first at HRI’s committees, HRI does not have a committee focused solely on in preventing and mitigating workplace violence. (Tr. 104.) Instead, workplace violence is within the responsibilities of three standing committees: (1) the Performance Improvement Committee; (2) the Restraint Reduction Committee; and (3) the Environment of Care Committee. (Tr. 100, 104-5.) The Performance Improvement Committee meets monthly to evaluate patient and staff incidents, look for trends from incidents and EARs, and to consider potential corrective actions. (Tr. 257-62, 515-16, 518-20; Ex. R-55.) Members review trends and audits to ensure that staff are taking appropriate actions with respect to safety. Id. The Restraint Reduction Committee meets monthly to review incidents that involved restraining a patient. (Tr. 151, 500-2, 510-15.) It confirms that all such incidents are documented and promptly reviewed by senior leadership as HRI’s policies require. Id. This committee also reviews training and makes sure it is completed. (Tr. 515.)

The third committee with responsibilities for workplace violence is the Environment of Care Committee. It meets quarterly and focuses on care, safety, and infection control. (Tr. 395-97; Exs. C-18, C-19.) As part of its work, committee members conduct monthly Environment of Care rounds to look for safety and other issues at the facility. (Tr. 397-98, 403.) Findings from these rounds are discussed at committee meetings. (Tr. 405.) Committee members also discuss employee injuries, training needs, and various risk areas, such as problems related to contraband. (Tr. 395-97, 402, 404-6, 440-41.) The results of camera audits and injury trends are also discussed to identify corrective actions and make sure they are implemented. (Tr. 395, 400-2, 404, 444; Exs. C-18 at 3, C-19.) For example, the committee discussed moving to “live feed monitors” in the two most acute units. (Tr. 402; Ex. C-18.) It determined that this type of monitoring would provide additional safety to staff and then this change was made. (Tr. 403.) The committee also discussed providing mandatory training to help better manage patient aggression. (Ex. C-19 at 2-3.)

These three committees are composed of individuals with fewer responsibilities for direct patient care.17 (Tr. 106-15, 124, 501.) Still, minutes and the testimony show that supervisors of mental health workers and nurse managers, both of which have direct care responsibilities, do attend these meetings. (Tr. 108-9, 114, 387-93, 940-42; Exs. C-11, C-12, C-13, C-14, C-15, C-16, C-18, C-19, C-20.)

Further, all direct care staff are invited to attend the Restraint Reduction Committee meetings as well as the town hall meetings. (Tr. 125, 388, 673, 939-40.) And HRI takes several steps to encourage their attendance at these meetings.18 (Tr. 859-60.) Committee meetings are scheduled at various times to accommodate different shifts and some meetings are held more than once. (Tr. 193, 216, 674, 783, 859.) The meeting times are visibly posted and communicated to staff in multiple ways. (Tr. 216-17, 783, 859-60.) In addition, HRI indicated it was willing to schedule them at different times when staff expressed an interest in coming. (Tr. 216, 885.)

The town hall meetings are an “open forum to discuss anything that will impact staff,” including safety and workplace violence. (Tr. 218, 783.) They are held monthly and on an emergency basis if needed. (Tr. 125, 215-16, 782-83.) For example, HRI held an emergency town hall meeting after the employee was attacked with the dresser drawer. (Tr. 218, 220.) Staff can (and do) make comments and ask questions during the town hall meetings. (Tr. 217, 221-22, 783.) Employees are paid for their time at town hall meetings. (Tr. 234, 673.) Employees can also anonymously provide input about workplace violence through a locked suggestion box anytime or in annual safety surveys. (Tr. 191, 193, 215-19, 668, 674; Ex. C-24.) The CEO explained that there have been very few suggestions placed in the box. (Tr. 219.) She personally reviewed each suggestion and indicated how she followed up on the concerns raised. Id. In addition to these avenues, staff can raise safety concerns directly with the CEO and some have done so. (Tr. 217.) HRI provided examples of when it made changes because of feedback on safety it received from direct care staff. (Tr. 193.)

Besides open forums for discussions with direct care staff about workplace violence, HRI also has a process to inform staff when there is a known potential for aggression and after incidents occur. (Tr. 214-15, 323-24, 478-79; Exs. R-22, R-24.) The intake process includes looking for aggression and violence towards others. (Tr. 827; Ex. R-24 at 26-27, 29.) Intake forms for patients considered potentially aggressive are marked and include a different colored cover sheet, called a High Risk Alert Sheet, to make it easier for staff to identify. (Tr. 720-21, 763, 833; Ex. R-24 at 29.) Both the nurse supervisor and charge nurse are required to sign this cover sheet. (Tr. 835.) The charge nurse informs the staff on the unit about the risk of aggression and staff also have access to the patient’s file, including the High Risk Alert Sheet. (Tr. 720-21.)

Once admitted to the unit, if a patient exhibits aggression or the potential for it, this is included in the nursing notes and is part of the shift change meetings.19 (Tr. 129, 747, 765, 767.) The nursing notes specifically include a section focused on whether there is any risk of assault. (Tr. 769.) Managers discuss issues related to patient aggression and any violence against staff at the daily Flash meetings, which are open to mental health workers and all nursing staff. (Tr. 212-14, 304-5.) If a patient is identified as aggressive, there are several precautions and interventions, such as increasing observation, restricting privileges, and requiring additional treatment, that employees put into place depending on the situation. (Tr. 452-54; Exs. R-22, R-26.) The nursing supervisor or charge nurse may place a patient on high-risk status at any time. (Tr. 462; Exs. R-22, R-26.) Each unit has a whiteboard which identifies the observation levels, privilege levels, and precautions for each patient. (Tr. 479, 721-22.) The whiteboard will indicate whether there has been any violent behavior. (Tr. 722; Ex. R-61.)

Staff are made aware of any Code Greens that occur and are debriefed after any incident that leads to a patient restraint or an employee injury. (Tr. 323, 329, 331, 486-88; Exs. R-30, R-38.) This debriefing process includes considering what actions could have been handled differently, determining if there are any corrective actions that should be taken, and to learn if there are ways to reduce the likelihood of such incidents occurring again. (Tr. 214-15, 487-96, 514; Ex. R-38.) They are asked whether a specific staff member was the targeted by the patient. (Tr. 73, 318-19; Ex. R-38.) When that occurs, the staff member is directed not to be involved in the restrictive intervention. Id. In addition, when a patient targets an attack at a staff member, corrective actions can include re-assigning the person to another unit. (Tr. 321.) Staff are asked about whether the response to the code was fast enough, whether there are needs for additional training, and what suggestions the employees involved in the code have. (Tr. 494-96; Ex. R-38.) If available, camera footage of the incident is reviewed, and staff are asked to provide further input after they review the footage. (Tr. 493; Ex. R-38.) Management promptly reviews information from the debriefings.20 (Tr. 299-300, 510.)

If the incident involved a restraint, then the Restraint Reduction Committee will conduct its own additional review. (Tr. 510; Exs. R-38, R-57.) This committee confirms that incidents are reviewed by management the next business day; that the required debriefing of the staff involved occurred, and that both the incident and debriefing was appropriately documented. (Tr. 510-15; Ex. R-57.) HRI uses the information gathered from the debriefings and other reviews to assess what corrective actions are needed, what improvements can be made, whether there is a need for additional training, and to determine whether any trends exist in the units. (Tr. 299-300, 490-97, 514.)

Dr. Lipscomb argued that HRI’s efforts at involving direct care staff were insufficient. She indicated that HRI failed to debrief an employee after an injury and denied a request to schedule a meeting at a more convenient time. (Tr. 596, 602, 674.) As support for this assertion, she relied on interviews she conducted over the telephone with three former employees after her deposition and the completion of her expert report. (Tr. 553, 614-17.) Her expert report was not supplemented after these interviews. (Tr. 615-17.) At the hearing, HRI timely objected to the introduction of the hearsay evidence and Dr. Lipscomb’s failure to update her report. (Tr. 614-17.)

As an expert, it was permissible for Dr. Lipscomb to rely on hearsay. Fed. R. Evid. 703. Nonetheless, when an opinion is based partially on hearsay this may impact the weight given to any such opinion. See Newell P. R., Ltd. v. Rubbermaid Inc., 20 F.3d 15, 20-21 (1st Cir. 1994) (when the factual underpinning of an expert opinion is weak, it affects the weight and credibility of the testimony); Owens v. Republic of Sudan, 864 F.3d 751, 789 (D.C. Cir. 2017) (a party cannot use an expert simply as a conduit for introducing hearsay). The Secretary did not explain the position of the person or persons who made the statements, what time period the assertions relate to, and whether any of the individuals were employed by HRI during the time period referenced in the Citation. Given the limited information about the declarants, it is difficult to assess the credibility of the statements made to Dr. Lipscomb. See Hamilton Fixture, 16 BNA OSHC 1073, 1085 (No. 88-1720, 1993) (recognizing that the problem with hearsay is that the objecting party has no opportunity to cross-examine the declarant), aff’d, 28 F.3d 1213 (6th Cir. 1994). Thus, Dr. Lipscomb’s testimony about insufficient employee involvement does not outweigh the other evidence regarding how HRI solicited and responded to employee concerns and suggestions about patient on staff violence.

With respect to whether debriefings occurred, the record includes EARs signed by injured employees. (Ex. C-1.) The EARs describe the incidents and each one explicitly indicates that the event was discussed with a supervisor. (Ex. C-1; Tr. 817-18.) The Secretary did not present evidence of any employee injury that was not reported in a written EAR. One witness felt that her debriefing after she was hit with a soda bottle focused only on her condition, not any “lessons learned.” (Tr. 806.) Despite her perception of the debriefing, HRI produced her signed statement indicating the debriefing occurred and explained the specific policy change it made after her injury. (Tr. 139, 818-19; Ex. C-1.) HRI also presented other evidence about its review and follow up process for all EARs. (Tr. 214-15, 349-52; Ex. C-7.)

As to the scheduling of meetings, the CEO and nurse educator testified about when the committee meetings and town hall meetings were scheduled and the steps HRI took to give employees opportunities for providing feedback. Meetings are repeated, and employees were paid for their time, as discussed previously, infra. (Tr. 193, 783-85.) Thus, the evidence about the follow up that occurred after employee injuries and the testimony about the scheduling of meetings is credited over the hearsay statements Dr. Lipscomb cites. See Beta Constr. Co., 16 BNA OSHC 1435, 1442 (No. 91-102, 1993) (assessing credibility in light of the whole record).

As with some of the Secretary’s other proposed abatement strategies, HRI essentially concedes the need for employee input and agrees it is feasible to have employees, including those with direct care responsibilities, involved in the planning, development, and implementation of workplace violence prevention programs. (Tr. 940; Resp’t Br. at 58-59.) HRI offered evidence of multiple actions it takes to encourage all employees to be involved with workplace violence prevention efforts. (Tr. 215-16, 673-74, 940.) HRI solicited feedback and when employees communicated concerns it followed up on them. (Tr. 161-62, 191, 216.) With respect to communication, HRI informs employees about incidents and concerns when new patients arrive, during shift changes, through meetings, a newsletter, and internal emails. (Tr. 216-17, 769, 783.) Employees can raise safety concerns with management through post-incident debriefings, annual surveys, at the monthly town hall meetings, through committees, anonymously through a locked suggestion box, or directly with managers (including the CEO). (Tr. 161-62, 167, 191-93, 215-20, 488, 782-83; Exs. C-24, R-38.) HRI encourages feedback through these channels and developed action plans as a result of the information conveyed. (Tr. 192-93, 216, 783-84.) Agendas and minutes indicate that addressing staff safety was a significant responsibility for three committees and the topic was discussed at town halls. (Ex. C-18, C-19, C-20, C-21.) Importantly, HRI also has a process to follow up on the safety concerns raised by employees and it showed the actions it took to address safety concerns. (Tr. 192-93, 218-19, 222, 783.) The Secretary failed to show that HRI’s efforts at employee involvement were inadequate or that his proposed abatement would further increase employee participation.


(1)Staffing for Code Greens

As to adequate staffing, there is no debate that appropriate levels of staffing are critical at any behavioral health facility, including HRI. (Tr. 910, 920, 928; Ex. C-28 at 29.) But the parties disagree as to whether HRI’s approach to staffing led to adequate coverage for purposes of materially reducing the hazard of patient on staff violence.

HRI follows a staffing grid approach that sets forth a minimum patient to staff ratio. (Tr. 160, 181.) The minimum ratio complies with the requirements and recommendations of a state authority, the Massachusetts Department of Mental Health (DMH), which licenses the facility. (Tr. 179, 229, 882.) HRI explained the steps it takes to make sure every shift has sufficient staffing that at least meets the minimum recommendations set by the DMH. (Tr. 150, 213, 228, 237, 728-729; Ex. R-31.) These procedures are in writing and preclude any unit from operating without a Registered Nurse (RN).21 (Tr. 150, 180-81, 228; Exs. R-31, R-45.) Staff escort patients eligible to leave the unit outside for fresh air breaks, but even in such circumstances HRI maintains a ratio of staff members to patients on the unit. (Tr. 482.) The amount and sufficiency of the staff are reviewed at the morning Flash meeting. (Tr. 213.)

Ms. Cooke evaluated HRI’s approach to staffing and discussed the issue with five staff members. (Tr. 874, 881-82.) In her interviews with staff, the employees did not raise any concerns about inadequate staffing. (Tr. 874.) Based on her review and site visit, she concluded that staffing at the facility was adequate. (Tr. 920.)

As support for its claims of inadequate staffing for the hazard, the Secretary relies on (1) a staffing report from a single day, (2) various comments made on a 2016 safety survey, (3) the testimony of one employee, and (4) Dr. Lipscomb’s opinion. (Sec’y Br. at 10-12, 45-46.) First, the Secretary alleges that a staffing report provided to the CO at the start of the investigation shows HRI failed to staff an RN on one shift on August 3, 2016. (Sec’y Br. at 10-11; Ex. C-23.) Ms. Kong explained that the staffing report the Secretary cites was not a record of who worked the shift but was a planning document relating to who was going to work the shift. (Tr. 144, 147, 152, 187, 195.) Units were never operated without at least one RN and HRI had a protocol for assigning staff if there were no volunteers for the shift. (Tr. 144-45, 147, 150, 190.) The former interim CEO explained how she confirms the sufficiency of staffing. (Tr. 195.) No one refuted the evidence indicating that units always had a least one RN and that the staffing report could not be relied alone to determine the number of workers on any given day. (Tr. 181, 195, 228.) Thus, the staffing report does not establish that HRI lacked adequate staff. (Tr. 407-8.)

Second, in the survey responses, some employees praised HRI and others were more critical. (Ex. C-24.) None of the responses provide much detail. Id. All are brief, and the record does not reflect what questions were asked, what time period any of the responses refer to, the positions of the responders, or whether each comment is from a different person. (Tr. 162; Ex. C-24.) The few responses directly referring to understaffing do not indicate whether the concern relates to the hazard of workplace violence or some other concern.22 (Ex. C-24; Tr. 165-67.) The survey responses are not sufficiently persuasive as to whether changing HRI’s staffing approach would materially reduce the hazard of patient on staff violence at the facility.23

Third, a current employee explained that there is less staff at night. (Tr. 807.) She indicated that not everyone could respond to a Code Green because each unit always had to have a minimum number of employees. Id. But she did say that the Code Green responses were inadequate because of this or that the amount of staff contributed to the hazard of workplace violence. Further, staff had fewer patient related tasks at night and did not leave the units to take patients for fresh air breaks during the night shift. (Tr. 803.) In some cases, the employee’s concerns, while framed as being about staffing, appeared to relate to other issues. For example, she referred to the difficulty of moving a heavy patient who was restrained to a board. (Tr. 812-13.) Her suggested solution for this issue was not more staff, but a different piece of equipment that could be rolled rather than carried.24 (Tr. 812, 820.) Thus, although the witness advocated for more staffing, the Secretary failed to show how staffing contributed either to the occurrence of workplace violence at HRI or the seriousness of any such incident there

Fourth, as to Dr. Lipscomb testimony, she argued that staff levels should be “safe” and criticized strict adherence to a single ratio of patients per staff member, which she referred to as a grid-based approach. In her view, grid-based approaches to staffing fail to sufficiently account for variability in the risk level. (Tr. 601, 666-67.) While she believed that staffing levels at HRI were inadequate, she acknowledged that there was disagreement among HRI employees as to whether this was the case. (Tr. 599-600, 669.)

HRI appears to recognize the limitations of a strict grid approach that only looks at the number of patients per staff member. So, it adopted measures to increase staffing beyond the grid when necessary. The grid is just the minimum threshold. (Tr. 160, 228, 935.) Staffing needs are reviewed during the daily Flash meeting and at other times as well. (Tr. 213, 728-29.) The number of staff relative to each patient can be increased under various scenarios, including challenging patient behavior.25 (Tr. 181-82, 229-30, 728.) In addition, if a patient refuses to allow a search of his or her body a staff member is automatically assigned to continuously monitor just that patient. (Tr. 258-59.) Staff other than on-unit nurses and mental health workers are trained to respond to situations involving aggressive patients. (Tr. 204, 207, 311.) Supervisors, socials workers, and “floats,” which are staff not assigned to any single unit, can be called upon to assist with patient aggression. (Tr. 183-84, 207, 229-30, 277, 320.) Additional nurses or mental health workers are added to shifts when needed. (Tr. 728-29.) Staff, both nurses and mental health workers, can ask the charge nurse at any time for additional staffing. (Tr. 182-83.) The former interim CEO was not aware of such requests being denied, and the record does not reflect any examples of denied requests contributing to the cited hazard.26 Id.

The Secretary alleges that “adequate staff coverage” would be a feasible means of abatement but fails to establish how HRI’s staffing was inadequate to address the hazard. The Citation itself does not directly call for a higher number of employees. Instead, it indicates that “a feasible and acceptable means to abate the hazard” includes:

Ensuring adequate staff for Code Green incidents to ensure that these responses can be fully functional, and that the required patient acuity and staffing ratios can be maintained on each shift, including when there are multiple admissions during a shift, and to ensure continued adequate coverage for remaining patients such as those requiring one-to-one care.

The Secretary did not show that the criteria HRI uses to assess when it is appropriate to go beyond the minimum staff to patient ratio called for by the grid was inadequate.27 Nor did he show that HRI failed to implement increased staffing when needed for staff safety. HRI explained how additional staff were available to respond to Code Green incidents or in the event of increased admissions. (Tr. 229-30.) While there were fewer workers at night, staff did not need to be off the unit for fresh air breaks and they did not have responsibilities related to discharges or certain patient care tasks during that shift. (Tr. 803.) The Secretary did not refute HRI’s evidence that there were multiple employees per shift in each unit and that it has mechanisms in place to increase staffing when there is a greater potential for patient on staff violence. (Tr. 214, 229-30.) There is no evidence that requests for additional staff were denied or that there were insufficient numbers of people to respond to workplace violence incidents. While the Secretary showed that it was necessary to have adequate staff, he failed to establish that HRI’s staffing or its system for adding additional workers was inadequate for addressing patient on staff violence.

(2)Security staff

The Citation also calls for HRI to “Employ security staff when necessary to assist in responding to violent events.” The Citation does not elaborate on when it would be necessary, nor did the Secretary clarify this at the hearing or in his brief. (Sec’y Br. at 46-47.) Further, the Secretary failed to make clear what he meant by the term “security staff.” Id.

HRI had a safety officer but did not employ dedicated security personnel. (Tr. 64, 204.) The decision not to have security personnel appears to have been a considered decision. Ms. Cooke explained that managing patients through intimidation could be counter-productive and increase, rather than reduce, patient aggression. (Tr. 860-61.) She did not agree with having security personnel in a behavioral healthcare setting.28 (Tr. 937.) As an alternative, HRI trained staff on how to prevent aggressive behavior, de-escalate situations when they occur, and, should the preventative efforts fail, safely restrain patients when necessary. (Tr. 593, 938; Exs. R-2, R-12, R-30, R-61.) Ms. Cooke argued that HRI’s staff was trained and competent to manage patient behaviors, including aggression. (Tr. 938-39.)

One employee indicated that she would feel safer if the facility had a dedicated security person because some of the patients are a few inches taller than she is. (Tr. 809-11.) She was not asked why a “security” person would be better than HRI’s approach of having mental health workers, social workers, and nurses address patient aggression. She acknowledged that all of HRI’s direct care staff receive training in verbal de-escalation as well as physical intervention techniques and did not suggest that additional security training was necessary for those employees. (Tr. 593-94; Ex. R-20.)

The Citation does not specify whether HRI needed more employees or if it could assign existing workers into a security role. As discussed above, the Secretary failed to show that HRI lacked adequate staff to address the cited hazard. Similarly, the Secretary did not show that workers lacked sufficient training to address patient on staff violence. As the Secretary failed to prove that HRI’s existing staffing and training was inadequate, the undersigned does not reach the issues of whether dedicated security personnel would materially reduce the hazard.

C.Additional Abatement Proposed by the Secretary

The Secretary also alleges that HRI’s workplace violence prevention program was deficient because it failed to: (1) prevent patients from being able to remove dresser drawers; (2) ensure patients used cups rather than bottles for drinks; (3) install panic alarms; and (4) have nursing stations with sufficient barriers.29

1.Dresser Drawers

As noted above, on May 18, 2016, a patient removed a dresser drawer and hit an employee with it, badly injuring him. (Tr. 93; Ex. C-1.) HRI did not know the patient would act aggressively, nor was it aware of any specific risks related to the furniture in the patient areas. (Tr. 234, 360.) After the attack made it apparent that the drawers could be problematic, HRI took various steps to address the issue. (Tr. 361-62.) It revised the patient’s treatment and had him transferred to another facility. (Tr. 232-34, 355.) It also promptly alerted the staff about the incident at an emergency town hall meeting, evaluated why the incident occurred, and assessed approaches to reducing the risk of patient on staff violence. (Tr. 220-21, 234, 355-56, 361-62.) As a first step, it installed screws in the drawers to make them more difficult to remove and had staff look for the issue on their environment of care rounds. (Tr. 131-32, 360-61, 968-69; Ex. C-19 at 2.) Although no one was injured by another removed drawer, HRI learned that some patients were able to remove the screws holding the drawers in place.30 (Tr. 361-62, 967, 972-73; Ex. C-19 at 3.) So, after further research and consideration, HRI eventually replaced all dressers with open storage cubicles. (Tr. 130-32, 134-35, 362, 972-73.)

The Secretary is not proposing HRI take any further action regarding the storage of patient belongings or drawers in general. There is no evidence that any other drawers remain a hazard at the worksite or that the new storage cubicles might contribute to workplace violence. The Secretary’s contention appears to be that HRI responded too slowly to address the use of drawers as weapons. (Sec’y Br. at 6.) He argues that Respondent recognized the hazard of the dresser drawers before a patient attacked the employee on May 18, 2016, because HRI bolted all dressers to the floor. Id. at 5.

HRI disputes knowing the drawers presented hazard before the attack. (Resp’t Br. at 30.) The furniture had been considered safe for use in a psychiatric institution. (Tr. 360.) It bolted the dressers to the floor not because it knew the drawers could be used as a weapon but because it did not want patients to use the whole dresser to create a barricade. (Tr. 356, 361, 432-33, 630, 961-62.) The evidence does not show that the dressers in use at HRI were inherently dangerous. A patient had never dismantled a drawer to use it as a weapon before. (Tr. 354, 361, 432.) Nor did any other patient repeat the behavior, even in the interim period before the dressers were replaced. After the attack, HRI knew it was possible for a patient to hit someone with a drawer but the likelihood of such a situation was unclear. (Tr. 135, 361-62, 437; Exs. C-1 at 1, C-9 at 1.) The patient involved in the incident failed to disclose to HRI a substance he was taking and so his behavior was not consistent with his medical history.31 (Tr. 232-34, 355, 437; Ex. C-9 at 1.) Still, HRI took steps to reduce the possibility of a future incident. (Tr. 962, 966.) It held a town hall and provided training on patient aggression and unpredictability. (Tr. 437, 441; Exs. R-2, C-4, C-5.) Then it added screws to the drawers and eventually removed the drawers altogether after it appeared that the screws were not completely effective at preventing patients from being able to remove a drawer. (Tr. 134-35, 361-62, 967-68; Ex. C-22 at 2.)

Certainly, this incident shows that patient on staff violence had not been fully abated before OSHA commenced its investigation. But, the fact that the incident occurred does not establish that HRI failed to adopt feasible means of abatement. See Nat’l Realty & Constr. Co., Inc. v. OSHRC, 489 F.2d 1257, 1265-66 (D.C. Cir. 1973) (noting that the general duty clause does not impose strict liability). The Secretary did not show that HRI should have known before the attack that the dressers presented a hazard or that its initial approach of tightening the screws on the drawers would be ineffective. (Tr. 131-32, 354, 361, 432, 961.) Even fully crediting the argument that HRI should have addressed the issue with the drawers sooner, it is not clear this would have contributed to materially reducing the hazard beyond what HRI’s other abatement measures achieved. While taking prompt action is generally preferable, when alleging a violation of the general duty clause, the Secretary must show that its proposed abatement will materially reduce the hazard.

2.Refuse patients bottled drinks

The Secretary calls for HRI to have a rule which precludes patients from having beverage bottles. (Sec’y Br. at 45.) The Secretary argues that the need for this rule is demonstrated by an incident on June 18, 2016, during which a patient threw a soda bottle. (Tr. 103-4, 139, 365, 368-69; Exs. C-1, C-9.) The bottle hit a wall and then hit an employee, injuring her. (Tr. 103-4, 127, 368-69, 434; Exs. C-1, C-9.) No patient had used a soda bottle aggressively before this incident and Ms. Cooke explained that in her experience with other behavioral health facilities soda bottles are permissible. (Tr. 369, 883.)

HRI argues it abated this hazard before the Citation was issued by instructing staff to pour drinks into paper cups and not leave their own soda bottles somewhere accessible to patients. (Tr. 139, 369; Resp’t Br. at 41.) HRI reviewed videotape of the incident and determined that the employee did not fully follow the company’s procedures. (Tr. 366.) In its view, the employee should have taken additional steps to de-escalate the patient’s aggression.32 (Tr. 366-67, 434, 493.) Staff are taught that anything can be a weapon and that de-escalating patient aggression can reduce the likelihood a patient will throw an object. (Tr. 141, 170, 177, 203, 369, 752, 790; Exs. C-5, R-2 at 6, R-12, R-13.)

The Secretary argues that despite HRI’s rule precluding soda bottles, the CO photographed one in a patient area. (Sec’y Br. at 7.) However, the exhibit the Secretary relies on depicts what appears to be a one-liter plastic soda bottle on a desk underneath a raised counter. (Ex. C-25 at 5; Tr. 631.) Ms. Kong explained that the location shown in the photograph was the nurses’ station, not a patient room or common area. (Tr. 140.) The photograph shows that on top of the counter there was a stack of paper cups, but no bottles. (Ex. 25 at 1, 2.)

The Secretary fails to explain how HRI should provide access to drinks differently. The proposed abatement in the Citation does not directly refer to soda bottles. In his brief, the Secretary’s states that the proposed abatement is to prevent patients from having soda bottles. (Sec’y Br. at 45.) But, HRI already specifically instructed the staff to do this and directed employees to look for drinks in patient areas during their environment of care rounds.33 (Tr. 139, 369; Ex. C-19.) It told staff not to leave items lying around and trained employees in ways to prevent aggressive episodes that might result in a patient throwing something. (Tr. 790; Exs. C-5, R-2, R-12, R-13.) In his brief, the Secretary does not explain how to implement the no access to soda bottles differently than what HRI has done.34 See Postal Serv., 21 BNA OSHC at 1773-74 (Secretary failed to show the safety vests employer issued were inadequate).

3.Panic Alarms

The Citation states that “silent emergency alerting buttons that are carried by workers throughout their shift to call for help in any location in the event of an escalating situation with a patient” would be among the feasible methods of abatement. In his brief, the Secretary argues that “mobile” or “stationary” panic buttons would be acceptable options.35 (Sec’y Br. at 26-27.) Dr. Lipscomb asserts that panic alarms are reliable, accessible, and facility appropriate method of summoning help when a patient is beginning to act aggressively. (Tr. 587-88.) She believed that there were documents and guidelines supporting the use of panic alarms in the behavioral health setting but admitted she did not reference any such article in her expert report. (Tr. 652.) Further, she explained that “one size doesn’t fit all,” so panic alarms are not appropriate for every facility. (Tr. 653, 686.)

Ms. Cooke agreed that panic alarms would not be effective in every setting. (Tr. 875-77.) She argued that panic alarms can create an additional step before help arrives because the signal from the button calls one person and then that person has to put the call out to others before help arrives. (Tr. 877.) The Citation seems to anticipate the “two-step” situation Ms. Cooke raises concerns about because it calls for staff to monitor the alerting system and location indicators. Ms. Cooke also indicated that the current staff she talked with did not view the buttons as necessary to allow them to receive help with a patient. (Tr. 877-78.) She agreed with them that an additional method of summoning help was unnecessary at HRI given its phone system, walkie-talkies, small size, and the number of staff assigned to each unit. (Tr. 877-78, 882.) See Cerro Metal, 12 BNA OSHC at 1822-23 (abatement must be capable of being implemented and necessary to reduce the hazard). The small size made it possible to verbally ask for help as there was always more than one person in each unit of the facility. (Tr. 318, 882.) And when staff was with a patient off the units, they had walkie-talkies to summon help. (Tr. 482-83, 877.) A nurse on the unit monitors any communication from the walkie-talkie to ensure a response. (Tr. 482.)

The Secretary alleges HRI’s protocol was deficient because it required employees to locate a phone to initiate the process. But, the Secretary failed to establish that employees had difficulty locating phones to summon help. (Tr. 431-32.) The Secretary did not provide information about the number or locations of phones in each unit. Further, employees could initiate the process in other ways if needed: they could say “Code Green” rather than using the phone system, or they could ask a co-worker to use the phone on their behalf.36 (Tr. 318.)

As for mobile panic alarms suggested by the Secretary, the Citation itself notes that such a system may not be “functional throughout the facility due to poor signals.” So, if HRI adopted this method of abatement, it would have to train staff, as the Citation suggests, to “know the areas where the panic buttons are functional and where they are not.” Thus, rather than saying “Code Green” to a co-worker, over the phone or by using a walkie talkie, with mobile panic alarms, the staff member would still have to determine their location and then consider whether they could use the mobile panic alarm or had to summon help another way. The Secretary did not address the possibility of signal issues or explain how training staff about signal issues would be superior to a telephone system not subject to such a limitation. See Kokosing Constr. Co., 17 BNA OSHC 1869, 1875 n. 19 (No. 92-2596, 1996) (noting that in the context abating a hazard with personal protective equipment, “if a proposed abatement method creates additional hazards rather than reducing or eliminating the alleged hazard, the citation must be vacated for failure to prove feasibility”).

Turning to stationary panic buttons, HRI already had these in the admissions area and outpatient sections of the facility. (Tr. 587; Ex. C-20.) The Secretary contends that adding panic buttons to the units would be superior to “shouting” for assistance because loud noises can increase agitation and panic buttons do not require explanations about what the problem is. (Sec’y Br. at 27; Tr. 588.) Ms. Kong indicated she was not aware of staff being unable to utilize the phones to initiate a Code Green, so it was not necessary to shout. (Tr. 431-32.) As for being able to request assistance without saying why, HRI’s protocol directed employees to say “Code Green” to trigger the call for assistance. (Tr. 70-71; Ex. C-4.) An explanation was not necessary to activate the assistance process. Id.

Dr. Lipscomb advocated for a “very reliable, very accessible” way of summoning help. (Tr. 587.) That conclusion is sound. But, the Secretary failed to present sufficient evidence on the next part of the equation—could additional panic buttons and/or mobile alarms summon help faster and safer than HRI’s current methods such that it would materially reduce exposure to the hazard. See Cerro Metal, 12 BNA OSHC at 1822-23 (“the Secretary must not only describe specific, feasible measures for reducing the hazard, but must present evidence that knowledgeable persons familiar with the industry would regard the steps as necessary”). If a patient became agitated or aggressive, an employee could say Code Green to a co-worker or use the phone to commence the process. (Tr. 67, 318; Ex. C-4.) When staff take patients off the unit, such as during fresh air breaks, they are required to have walkie-talkies to call for help if needed.37 (Tr. 483, 877.) The Secretary did not show that employees were unable to initiate or respond to Code Greens. Nor did he show that the responses were delayed or ineffective because of the reliability of the system for summoning assistance.38 To sustain his burden, the Secretary must do more than provide another way to address the cited hazard. Postal Serv., 21 BNA OSHC at 1773-74 (finding that the Secretary failed to establish that different clothing would be more effective in reducing the hazard than what the employer already provided).

4.Evaluate the design of the nurses’ station

Another abatement method the Secretary suggests is to evaluate the design of the nurses work stations. About a year before the inspection, HRI undertook an evaluation of its nurses’ stations. It determined that more open nurses’ stations might decrease patient frustration and improve monitoring, so it removed barriers and lowered the counter heights. (Tr. 138.) Cheryl Grau, who was previously Respondent’s interim CEO, indicated that the stations were redesigned so nurses could better observe and monitor the patients. (Tr. 263.) HRI was concerned that higher enclosed stations may increase patient frustration by making them feel isolated from staff. (Tr. 262.) It believed that opening up the nurses’ station would increase safety for both patients and staff. (Tr. 263.) For example, before their removal in 2016, patients repeatedly banged on the plastic barriers around the nurses’ station causing them to break several times. (Tr. 138-9, 262.) Ms. Cooke explained that, in her view, enclosed nurses’ stations can make it more difficult for staff to observe, monitor, and assess patients on a regular basis. (Tr. 883.) She advocated for low counters that do not create a barrier between staff and patients. Id.

Despite these benefits of an open design, HRI was also aware that lower counters present different challenges because they may make it easier for patients to grab objects from the nurses’ station. (Tr. 136, 591.) So, HRI took steps to address the negative aspects of the design by, among other things, instructing staff not to leave items on the nurses’ station where patients could access them. (Tr. 373.)

Although the Secretary showed it was technically possible for the nurses’ station to be higher and have an additional barrier, the Citation only calls for an evaluation of the nurses’ station. The Secretary neither sufficiently described the changes he seeks nor established that changing the design would offer superior safety. Dr. Lipscomb was aware of “controversy” about the height of nurses’ stations and was not asked to provide a specific minimum height or depth for the counters. (Tr. 590, 592.) She also acknowledged the need for patients to access staff and recognized that higher counters make this more difficult. (Tr. 590-91.)

Adverse consequences may render a possible abatement method infeasible. CSA Equip. Co. LLC, 24 BNA OSHC 1476, 1478 (No. 12-1287, 2014) (noting that the judge must consider whether the proposed abatement method will result in adverse consequences that render it infeasible). HRI already undertook an evaluation and had procedures in place to limit access to items that could be used as a weapon. The Secretary failed to show that another evaluation was a necessary and valuable component to materially reduce the hazard at this facility. See Cerro Metal, 12 BNA OSHC at 1822-23 (abatement actions must be regarded “as necessary and valuable for a sound safety program”); Postal Serv., 21 BNA OSHC at 1774.

D.Secretary fails to meet the abatement prong of his duty

In sum, the Secretary’s proposed abatement methods essentially fall into one of two groups. The first group consists of measures HRI has already taken: remove dresser drawers, preclude soda bottles, have sufficient staff, involve employees in workplace violence prevention, require incident reporting and tracking, train workers about how to prevent and respond to workplace violence, limit access to items that could harm staff, have a comprehensive written workplace violence prevention program, implement, monitor and assess compliance with that program. For this group, the shortcomings the Secretary alleges are either not supported by record evidence or the Secretary fails to show that the proposed additional action would materially reduce the hazard beyond what HRI already achieved. See Ala. Power, 13 BNA OSHC at 1245 (finding no violation of the general duty clause, when the Secretary did not show that employer’s measures were inadequate or that additional efforts would have been more effective); Beverly, 19 BNA OSHC at 1190 (Secretary must show that the abatement “would be effective in materially reducing the hazard”).

The second category of proposed measures consists of efforts that the Secretary failed to establish would reduce the hazard of patient on staff violence at HRI, e.g., install panic alarms and change the height of the nurses’ station. For this employer, the Secretary did not show these steps were necessary when evaluated in the context of the precautions already in use. See Cerro Metal, 12 BNA OSHC at 1822-23 (Secretary must show the actions are necessary and valuable for the worksite’s safety program); Pepperidge Farm, 17 BNA OSHC at 1995 (vacating a general duty clause violation when the Secretary failed to show that further abatement was required considering what was already undertaken).


The foregoing Decision constitutes the Findings of Fact and Conclusions of Law in accordance with Rule 52(a) of the Federal Rules of Civil Procedure.  Based upon the foregoing Findings of Fact and Conclusions of Law, it is ORDERED that:

1. Citation 1, Item 1 is VACATED, and no penalty is assessed.





__/s/Covette Rooney__________________


Chief Judge, Occupational Safety and Health Review Commission



DATED:  January 22, 2019

Washington, D.C.






1 Neither party identified the exact date of the second site visit.

2 At the hearing, HRI’s counsel also acknowledged that she did not think HRI was “disputing there’s a recognized hazard.” (Tr. 918.)

3 The results of the Arnetz Study were published in two related papers, Judith E. Arnetz, Lydia Hamblin et al., Preventing Patient to Worker Violence in Hospitals: Outcome of a Randomized Controlled Intervention, 59 Am. C. of Occupational and Envtl. Med. 18 (2017), and Lydia E. Hamblin, Judith E. Arnetz, et al., Worksite Walkthrough Intervention, 59 Am. C. of Occupational and Envtl. Med. 875 (2017). (Tr. 608, 611, 623; Exs. C-30, C-31.) The later paper provided further details on the interventions used in the initial study. (Ex. C-31; Tr. 611-12.)

4 Direct care staff includes anyone that provides care to patients, including, mental health workers, mental health supervisors, nurses, nurse managers, and social workers. (Tr. 941.)

5 The presentation materials are dated August 2016. (Tr. 644; Ex. R-2.)

6 HRI's Code Green policy was effective as of August 2007 and was most recently reviewed in December 2016. (Tr. 69; Ex. C-4.)

7 The nurse manager could re-assign or alter the roles of the staff responding to Code Green depending on the nature of the situation. (Tr. 328, 418.)

8 The Citation also states that a feasible method of abatement for the hazard include ensuring “that all staff know which staff members are designated to respond to Code Green incidents on each shift.” The Secretary did not establish that employees did not know who was to respond to a Code Green.

9 As discussed below, HRI did have a “Near Miss Log” that seemed to include actual injuries as well as “near miss” incidents. (Tr. 93; Ex. C-8.) The Near Miss Log was not part of HRI’s regular incident reporting procedures or tracking system. (Tr. 91-93.) So, its over-inclusive nature does not appear to be connected to any deficiencies in HRI’s training program.

10 In his brief, the Secretary frames some of the alleged deficiencies slightly differently than how Dr. Lipscomb did during the hearing. The Secretary argues that effective training “would include training on a comprehensive workplace violence prevention program, incident report procedures and the specifics of what qualifies as a ‘psychiatric emergency’ justifying a Code Green.” (Sec’y Br. at 50.) To the extent that this formulation implies anything different than what Dr. Lipscomb suggested at the hearing, the undersigned finds that the Secretary failed to put forth sufficient evidence supporting such a contention. HRI had a workplace violence prevention program that included training on incident reporting and calling a Code Green.

11 HRI explained that the document was focused on violence among co-workers even though it refers to “non-employees” and “individuals.” (Tr. 64, 160; Ex. C-2.)

12 At the time of the initial inspection, this binder was present at the facility but was not provided to the CO. (Tr. 54-55, 57-58.) However, HRI subsequently provided it to the CO at her follow up visit. (Tr. 693.)

13 The policies list the date the policy was first effective and the date of the most recent review or revision. (Tr. 452, 467, 472, 480-81, 485, 741.)

14 HRI was concerned about patient access to items that could harm staff. (Tr. 469, 642, 743.) So, HRI trained staff about contraband and searches all patients as well as their belongings in the admissions area before the patient enters the unit. (Tr. 467-69, 742-43; Exs. R-27, R-61.) Additional searches occur on the unit when employees suspect that contraband may be present and when ordered by a physician. (Tr. 467, 469-70.)

15 The Secretary does not allege that HRI violated any OSHA standards related to recordkeeping.

16 There is no evidence that individuals not employed by HRI played a regular role in developing or implanting workplace violence prevention strategies at HRI. The argument appears to be that direct care employees were not sufficiently encouraged to participate in HRI’s workplace violence prevention program. (Sec’y Br. at 48-49; Tr. 596-97.)

17 A direct care staff member is a part of the Environment of Care committee, but it is not clear how often he attended the meetings. (Tr. 114, 120; Exs. C-18, C-19, C-20.)

18 The Secretary does not argue that HRI should compel employee participation.

19 The charge nurse completes a shift report that documents whether there is a risk for assault or whether precautions for aggression are needed. (Tr. 769, 771; Ex. R-42.) During a 30-minute meeting at the start of each shift, the charge nurse informs the on-coming staff of relevant information, including any specific risks related to violence or aggression. (Tr. 721.) While all staff is encouraged to attend these meetings, some people coming off shift do not attend. (Tr. 129-30.) But, even if an employee does not attend the meeting when they are coming off their shift, they still have access to the written report. (Tr. 769, 772.)

20 Incidents related to Code Greens or restraints are reviewed the next business day. (Tr. 299, 510.)

21 The DMH also provides recommended guidelines for staff other than RNs, which HRI follows. (Tr. 178-81.)

22 Some survey responses called for “better” or “proper” staffing but do not elaborate on what that means or whether it relates to the hazard of workplace violence or patient health. (Ex. C-24; Tr. 166.) Others suggest more experienced staff, but do not indicate that the number of employees is inadequate or that more experienced staff will reduce the hazard of workplace violence. (Tr. 165; Ex. C024.)

23 HRI considered the practice of surveying direct care employees and developing an action plan after the survey part of its efforts to provide a safe environment and encourage employee feedback. (Tr. 191-92.) In finding that the survey responses were not sufficiently persuasive on the issue of whether HRI had adequate staffing, the undersigned is not determining that surveys were not relevant to whether HRI’s abatement was effective.

24 The witness also described an incident where a patient threw a soda bottle at a wall and the bottle hit her. (Tr. 804-5.) Three employees provided immediate assistance and then additional people came after the Code Green was initiated. Id. The witness did not indicate that additional staff would have prevented or mitigated her injury. Id.

25 One witness explained that within the first twenty-four to forty hours of treatment, patient needs may be higher so they may staff above the grid to address that. (Tr. 181.)

26 Also, if a request for an additional worker cannot be immediately fulfilled, other staff such as supervisors and doctors will go to the unit to assist until the additional staff arrives. (Tr. 182-83.)

27 On cross-examination, the Secretary’s counsel questioned Ms. Cooke about a published statement from the American Nurses Association (ANA) indicating that there should be “optimal” staffing to address workplace violence. (Tr. 921-28.) The Secretary failed to explain how implementing optimal staffing would require any different action than HRI’s approach.

28 Dr. Lipscomb also stated that a “police officer who only has that skill set” should not be in a therapeutic environment. (Tr. 594.)

29 The Citation sets forth nine main methods of abatement, some of which include multiple actions. The Secretary did not address all of the proposed actions listed in the Citation at the hearing or in his brief. The undersigned finds that the Secretary failed to establish the need for of any of the other methods of abatement discussed in the Citation, but not addressed at the hearing or in the briefing. See Peacock Eng’g Inc., 26 BNA OSHC 1588, 1592-93 (No. 11-2780, 2017) (focusing only on the method of abatement discussed by the Secretary); Ala. Power, 13 BNA OSHC at 1246.

30 Ms. Kong indicated that there “was an occasion or two” when patients removed the screws. (Tr. 361.) The notes from the July 29, 2016 Environment of Care Committee meeting are consistent with this testimony. (Ex. C-19 at 3.)

31 The Secretary does not allege that HRI should have handled the patient’s medical treatment differently. (Tr. 231, 234, 354.)

32 HRI does not allege its de-escalation procedures alone was sufficient to abate the hazard.

33 A July 29, 2016 committee report indicates: “food and drinks in patient bedrooms is consistently an issue.” (Tr. 141-42; Ex. C-19 at 1, 3.) The report does not refer to patients having access to any bottles. Id. Nor is there evidence of glass bottles ever being accessible to patients at the facility. (Tr. 631.) The report suggests HRI took the issue of drinks seriously—employees were looking for drinks in patient areas, reported it, and shared the information when violations of the rule were found. (Tr. 141-42; Ex. C-19 at 3.)

34 Although a complete ban on drinks might lead to fewer soda bottles at the facility, the Secretary neither proposes this or nor showed it would be feasible. HRI explained that such an approach would be less therapeutic and violate state treatment requirements. (Tr. 370-72, 857.)

35 Dr. Lipscomb explained that stationary panic alarms are attached in a fixed location while personal alarms are something an employee wears and can hit to trigger an alert. (Tr. 588-89.)

36 As discussed above, the record shows that HRI had adequate staff to respond to such requests.

37 Patients are not permitted to leave the unit if they are aggressive, considered assaultive, or at night. (Tr. 768, 803, 858.)

38 Respondent also alleges that there is no evidence that panic alarms reduce or eliminate the hazard of patient aggression because they are not activated until after a patient has already started to behave aggressively. (Resp’t Br. at 51.) However, the hazard here is patient on staff violence, not just aggressive behavior. Dr. Lipscomb explained how summoning help as quickly as possible can reduce staff injuries. (Tr. 589.)