Abstract
Concerns about sexual misconduct by health care professionals have been highlighted by recent high-profile cases. Professional health monitoring programs (PHPs) offer an additional layer of protection when health care professionals with a history of unprofessional sexual behavior (USB) return to practice; however, little is known about the characteristics or outcomes of clinicians referred to a PHP because of USB. Data were extracted from over 35 years of PHP records involving USB-related referrals (N = 570). The majority of cases were deemed ineligible for PHP monitoring and handled by other entities (e.g., licensing board, legal system). Of the 232 monitored cases (46.84 ± 9.42 years; 95.7% male), most were physicians (n = 156, 67.2%), with 75.9 percent of monitored cases involving USB with at least one patient. Most (74.9%) PHP monitoring outcomes were classified as “successful” or “very successful.” Only three individuals (1.3%) who completed their monitoring were rereferred to the PHP. Monitored professionals exhibited less severe USB and were less likely to experience legal or disciplinary consequences (57.3% versus 69.8%, Cramer’s V = .174, p < .0001) compared with unmonitored professionals. Findings enhance transparency of the PHP process and highlight its utility in safely returning clinicians to practice. Results may inform policies to prevent USB by health care professionals.
Professional sexual misconduct (PSM) is defined by the Federation of State Medical Boards Workgroup on Physician Sexual Misconduct as “behavior that exploits the physician-patient relationship in a sexual way…This behavior may be verbal or physical, can occur in person or virtually, and may include expressions of thoughts and feelings or gestures that are of a sexual nature or that a patient or surrogate may reasonably construe as sexual” (Ref. 1, p 19). Although underreported, sexual misconduct and boundary violations were the third most frequent disciplinary action issued by the state medical boards of 50 states and the District of Columbia in 2019.2 Concerns about such violations or other unprofessional sexual behavior (USB) by health care professionals received significant media attention in 2018 following the conviction of the USA Gymnastics team physician for sexual crimes against young female gymnasts.3 Reports of sexual misconduct to law enforcement have also increased significantly in recent years, particularly since the advent of the #MeToo movement.4
These events spurred the formation of a workgroup charged with improving policies to prevent and respond appropriately to egregious ethics violations in medicine, such as sexual abuse of patients.5 The group published a review of 101 cases of sexual violations by physicians, which identified the following five physician characteristics as primary risk factors for sexual misconduct: male physician, age 40+ years, not Board certified, nonacademic practice setting, and always examines patients independently.6 The results were similar to findings of a study of 1,039 physicians reported to the National Practitioner Data Bank for sexual misconduct between 2003 and 2013, which also found that the majority were not disciplined by their medical board.7 Given the grave risk of patient harm in such cases,7,–,13 more information is needed regarding sexual misconduct by physicians and other licensed health care professionals. Given the range of behaviors that may be characterized as USB, data are needed to better understand the scope of the problem and how best to handle these cases.
Although a number of studies have examined sexual harassment and exposure to USB in the health care setting, most studies have focused on the students, trainees, or professionals who were victimized.14,–,23 There is a dearth of information regarding health care professionals who engage in USB. Specifically, information is needed to facilitate appropriate monitoring of these individuals by professional health monitoring programs (PHPs: also known as “physician health programs”) in cases where an eventual safe return to clinical practice is likely. Currently, 47 of the 50 states in the United States have PHPs that provide monitoring for physicians (and sometimes other health care professionals) with potentially impairing conditions, such as substance use disorders (SUD) and psychiatric disorders.24 These programs were designed to assist health care professionals who completed recommended assessment and treatment for their condition(s) with their return to clinical practice, while ensuring patient safety. About half of the existing PHPs provide monitoring services for those who have engaged in PSM or other workplace sexual boundary transgressions.25
PHPs have an important role in protecting public safety while supporting health care professionals diagnosed with a potentially impairing condition and have been highlighted as a promising resource for monitoring health care professionals who exhibit unethical boundary violations.26 The present study included a review of more than 35 years of referrals related to USB at the Florida PHP. The goal of the present study is to characterize the population of health care professionals referred to the state PHP because of concerns regarding USB, including types of problematic behavior, characteristics of patient survivors, criminal and disciplinary consequences, and final disposition of the referrals.
Methods
Procedures
All study procedures were approved by the University of Florida Institutional Review Board. A query of the Florida PHP database identified all charts related to USB between 1982 and 2018. The population of eligible PHP participants varied over the course of the study because of changes in the program charge. Like many PHPs, this program initially served only physicians but gradually expanded over time. As of 2014, the PHP serves 38 license categories, including pharmacists, dentists, mental health professionals, veterinarians, and massage therapists. Participant charts were individually reviewed by the medical director of the PHP, and extracted data were entered anonymously into a data collection form hosted on Research Electronic Data Capture (REDCap)27,28 at the University of Florida. REDCap is a secure, web-based software platform designed to support data capture for research studies, providing an intuitive interface for validated data capture, audit trails for tracking data manipulation and export procedures, automated export procedures for seamless data downloads to common statistical packages, and procedures for data integration and interoperability with external sources. When all deidentified data had been recorded, the data were exported into SPSS for analysis. All files were saved in a password-protected folder on a secure server.
Records
As seen in Figure 1, the database query returned 572 unique records mentioning USB or PSM. Two cases were excluded from further review after careful reading of the file indicated that the individual was a victim rather than a perpetrator of sexual misconduct. This resulted in a sample of 570 records spanning from 1982 to 2018: 31 referrals in the 1980s, 282 referrals in the 1990s, 152 referrals in the 2000s, and 91 between 2010 and 2018, with 14 records missing a date of referral.
Figure 1. Characterization of records related to unprofessional sexual misconduct referrals. a The types or names of monitoring agreements (contracts) available varied over the course of the 35 years included in the chart review. Some individuals received more than one type of contract, typically if they had multiple presenting concerns or were rereferred to the program at a later date for a different concern. PHP, professional health monitoring program; PSM, professional sexual misconduct.
Review of records indicated that monitoring agreements (contracts) were initiated by the PHP when results of an external evaluation recommended monitoring and the referred individual agreed to participate. A total of 338 referrals did not participate in PHP monitoring. Reasons included: monitoring was not recommended (n = 222) because the individual was ineligible for PHP services (e.g., did not hold a license in the state), there was insufficient justification to recommend evaluation based on information provided to the PHP (e.g., vague, anonymous complaint that could not be verified), or the independent evaluator did not recommend monitoring based on the results of the assessment; the licensing board handled the referral (e.g., denied or revoked the individual’s license) without PHP involvement (n = 47); or the individuals relinquished their licenses (n = 41). In 29 cases, the PHP referred the individual to the licensing board because of a lack of engagement with the PHP when monitoring was recommended. Finally, four records were excluded from further analysis because the individual became incarcerated (n = 1), died by suicide (n = 2), or died by natural causes (n = 1) prior to fully engaging with the PHP.
Data Analysis
Data were analyzed using IBM SPSS version 29.0 (SPSS Inc., Chicago, IL, USA). Success of the monitoring process, in terms of meeting the dual missions of assisting professionals with potentially impairing conditions and protecting patient safety, was determined by the first author, in consultation with the final author, based on a retrospective review of the information available in the chart. “Very successful” cases completed their monitoring agreement without any aberrant behavior, and “successful” cases completed monitoring with modification or extension of the requirements. “Neutral” cases were generally individuals who discontinued monitoring despite no new concerns or complaints (e.g., because they moved to another state or because their license was revoked in response to concerns that predated monitoring). “Unsuccessful” cases involved individuals who were turned over to the Board because of noncompliance that did not involve patient harm, and “very unsuccessful” cases involved new occurrence of patient harm. Chi-squared, Fisher’s exact test (for cell counts less than five), and t-tests were used to examine group differences.
Results
Demographics
Results focus primarily on the 232 cases for whom monitoring agreements were initiated. As seen in Table 1, the majority were married (or remarried) middle-aged men who identified as heterosexual, and most were physicians (n = 156, 67.2%). The most common medical specialties were adult primary care (n = 52, 23.4%), psychiatry (n = 24, 10.8%), and obstetrics-gynecology (n = 9, 4.1%). The monitored group also included 26 (11.2%) nonphysician mental health professionals from a variety of license categories, including psychologists (n = 8), social workers (n = 8), licensed mental health counselors (n = 8), a licensed marriage and family therapist (LMFT; n = 1), and a marriage and family therapist intern (n = 1). More limited data were available regarding race, ethnicity, country of origin, and religious affiliation. The only significant difference between groups was age, with professionals who received monitoring agreements being slightly younger (46.84 ± 9.42 years versus 49.17 ± 10.66 years, t(410.93) = −2.39, p = .0172, d = .231). As seen in Table 2, the most common mental health diagnoses among the monitored group were personality disorders (n = 56, 24.1%), mood disorders (n = 41, 17.7%), paraphilias or other sexual disorders (n = 28, 12.1%), drug use disorders (n = 26, 11.2%), and alcohol use disorder (n = 25, 10.8%). Most monitored health care professionals with drug use disorders exhibited polysubstance use (n = 12), although the others primarily used opioids or opiates (n = 4), cocaine (n = 3), cannabis (n = 3), or sedatives, nitrous oxide, steroids, or amphetamines (all n = 1). In terms of treatment, 119 (51%) participated in individual therapy, 52 (22%) received treatment at a center specializing in treatment of professionals with PSM, 12 (5%) received treatment at a center specializing in treatment of professionals (without specific expertise in PSM), seven (3%) received some “other” form of treatment, 24 (10%) received no treatment, and no treatment information was available in the record for 18 individuals (8%).
Demographics of Healthcare Professionals Recommended for PHP Monitoring
Addictive and Psychiatric Disorder Diagnoses among Professionals Monitored by the PHP
Characterization of USB and PSM
Table 3 displays the types of USB exhibited by the health care professionals who were monitored compared with those who were not monitored (i.e., primarily cases that were handled through disciplinary action by the licensing board or criminal justice system). In the monitored group, there were 176 cases (75.9%) involving USB in the work environment with at least one patient or former patient, 64 cases (27.6%) involving USB in the work environment that did not involve patients (e.g., unwanted sexual contact or affair with a coworker, downloading pornography on a work computer), and 56 cases (24.1%) involving USB outside the work environment (e.g., public exposure, accessing or possessing child pornography, intimate partner violence). Some individuals exhibited USB in more than one of the aforementioned categories, leading to totals greater than 100 percent. The problematic behavior involved purportedly “consensual” sexual behavior between adults (e.g., an affair with a coworker) in 15.1 percent of cases (n = 35). It is important to recognize that, because of inherent power differences in workplace relationships, characterizing these encounters as consensual may be inappropriate. Indeed, the USB involved a strong possibility or likelihood of coercion (e.g., sexual behavior toward a patient or staff member, statutory rape) in 55.6 percent of cases (n = 129). Unwanted sexual attention and communication (e.g., sexual harassment, sexting, stalking, invasion of personal space) was noted in 40.1 percent of cases (n = 93). In addition, unwanted sexual contact either with penetration (e.g., sexual assault, molestation, rape) or without penetration (e.g., groping, fondling, frotteurism) was present in 14.7 percent of cases (n = 34) and 40.5 percent of cases (n = 94), respectively. Other problematic behaviors (e.g., excessive or inappropriate use of pornography, exhibitionism) were present in 12.1 percent of cases (n = 28). The majority of cases related to USB involved adult victims (n = 200, 86.2%), although 36 cases (15.5%) involved postpubescent minors and 17 (7.3%) involved prepubescent minors. In the majority of cases, the victims were only female (n = 197, 84.9%), although there were 28 cases involving only male victims (12.1%), and six cases involving both male and female victims (2.6%). Finally, in three cases, there was no victim or the victim’s age and gender were unknown.
Unprofessional Sexual Behavior Exhibited by Healthcare Professionals Recommended for PHP Monitoring
Professionals who initiated PHP monitoring were more likely than nonmonitored professionals to have exhibited USB in the work environment, either involving patient(s) (75.9% versus 60.4%, chi-square (1.00) = 14.20, p = .0011, Cramer’s V = .162) or not involving patients (27.6% versus 16.0%, chi-square (1.00) = 10.60, p = .0002, Cramer’s V = .141). These cases were also more likely to involve victims who were adults (86.2% versus 69.5%, chi-square (1.00) = 20.26, p < .0001, Cramer’s V = .193) and female (84.9% versus 71.9%, chi-square (1.00) = 12.52, p = .0004, Cramer’s V = .153). Professionals who were monitored had higher rates of USB involving purportedly consensual sexual behavior between adults (15.1% versus 5.3%, chi-square (1.00) = 14.41, p = .0001, Cramer’s V = .165) and sexual behavior with strong possibility or likelihood of coercion (55.6% versus 28.7%, chi-square (1.00) = 40.50, p < .0001, Cramer’s V = .270). By contrast, the monitored group displayed lower rates of unwanted sexual contact with penetration (14.7% versus 22.8%, chi-square (1.00) = 5.29, p = .0215, Cramer’s V = .101).
Legal and Disciplinary Consequences
Table 4 lists consequences experienced by the health care professionals because of their USB, including involvement with the legal system and disciplinary action by their licensing boards. Between-group comparison analyses showed that professionals who initiated PHP monitoring were less likely to report any civil or criminal legal involvement than those who were not monitored (57.3% versus 69.8%, chi-square (1.00) = 8.87, p = .0029, Cramer’s V = .129), and monitored professionals with legal involvement appeared to have less serious cases. Specifically, the monitored group were more likely to report civil settlement outside of court with no criminal charges (19.0% versus 7.4%, chi-square (1.00) = 16.24, p = .0001, Cramer’s V = .174) and to avoid a jail or prison sentence when found guilty of criminal behavior (20.7% versus 11.2%, chi-square (1.00) = 8.86, p = .0029, Cramer’s V = .130). Monitored individuals also had lower rates of legal involvement (i.e., criminal or civil cases) with outcomes unknown to the PHP (6.0% versus 12.7%, chi-square (1.00) = 6.11, p = .0134, Cramer’s V = .110) compared with those who were not monitored. With regard to professional licensure, the monitored group was more likely to receive disciplinary action by the licensing board that did not affect their ability to practice (52.3% versus 24.0%, chi-square (1.00) = 37.31, p < .0001, Cramer’s V = .291) and to experience practice restrictions while maintaining their ability to practice (31.9% versus 13.9% and 9.7% versus 2.2%, p < .0001, Cramer’s V = .287). Conversely, the individuals who did not participate in monitoring agreements were more likely to permanently relinquish their license (24.7% versus 14.0%, p = .0122, Cramer’s V = .135) or to have their license permanently revoked (23.3% versus 7.1%, p < .0001, Cramer’s V = .225).
Legal and Disciplinary Consequences of Unprofessional Sexual Behavior
Participation in the PHP
As seen in Figure 1, various types of monitoring agreements were offered to the PHP participants based on the reasons for referral and results of the comprehensive evaluation. The participants were diagnosed with a variety of potentially impairing conditions and were assigned monitoring requirements appropriate to their individual needs. Potential components of the monitoring agreements included psychiatric treatment, psychotherapy, SUD treatment, a worksite monitor, chaperone reports, staff surveys, patient satisfaction surveys, polygraph testing, participation in a professionally facilitated PHP monitoring group, random toxicology testing, participation in mutual support meetings, and completion of an intensive workplace professionalism course.
Outcomes of Participation in PHP Monitoring
As seen in Table 5, a minority of cases in the monitored group (n = 28, 12.1%) were still being monitored at the time of the study. The majority of cases (n = 128, 55.2%) successfully completed their initial monitoring agreement, including a subset (n = 23) who successfully completed following an extension of the monitoring period related to aberrant behavior or personal preference. A small number of cases (n = 27, 11.6%) discontinued monitoring against recommendation (i.e., relinquished their licenses), and 9.5 percent (n = 22) were turned over to the licensing board because of noncompliance with components of the monitoring agreement. A total of 3.4 percent (n = 8) were denied licensure or had their license revoked by the licensing board during the monitoring period. Finally, three individuals (n = 3, 1.3%) completed their initial monitoring period and were later re-referred to the PHP. A total of 74.9 percent (n = 173) of the monitored cases were classified as “successful” or “very successful,” whereas 13.8 percent (n = 32) were classified as either “unsuccessful” or “very unsuccessful.”
Outcomes and Classification of PHP Monitoring for Unprofessional Sexual Behavior
Discussion
Changes in the organizational culture of health care are needed to protect clinicians, trainees, and patients and to reach the World Health Organization goal of eliminating avoidable harm and improving patient safety.29 Specifically, unprofessional sexual behavior (USB) is a systemic and internalized problem, with sexual harassment and assault being surprisingly common in academic medicine.14,–,19 Indeed, a 2021 national report with 16,611 graduating medical students showed that 40.3 percent of them experienced sexual harassment, mistreatment, or discrimination during medical school, but only 27.3 percent of those experiencing the problems reported it.18 This problem persists in the workplace, where the overwhelming majority of physicians have witnessed sexual harassment, been a target of sexual harassment, or survived sexual assault.20,–,23 Nearly one in five Americans report interactions with a physician who was acting unethically, unprofessionally, or providing substandard care,30 and previous research has suggested an increase in the number of physicians being sanctioned for USB over time,31 which may reflect greater awareness of the problem or willingness to report it. Indeed, the #MeToo movement increased reporting of sexual misconduct to the police,4 although USB by health care professionals likely remains underreported.32
Previous work has demonstrated the positive impact of increasing attention to the taboo topic of USB by clinicians. Specifically, the Roy v. Hartogs33 case led to greater attention from medical licensing boards and raised awareness among mental health professionals regarding the importance of clear professional and ethics guidelines related to clinician-patient contact. These advancements were associated with a decline in USB by mental health providers and a corresponding increase in patient safety.31,34 The present study provides additional insight for forensic psychiatrists regarding how best to intervene in cases of USB to rehabilitate and monitor the offending health care professionals, when appropriate. Victims of USB by health care professionals are rightly outraged and desire justice for their suffering. Yet the outcomes for perpetrators are not confined to a dichotomy of imprisonment with permanent license revocation versus absence of legal or regulatory consequences.7 More often, at this intersection of moral, legal, and ethics breaches of acceptable professional behavior, there are combinations and levels of consequences, which may include referral to a state PHP. Forensic psychiatrists should be aware of these programs and the services they can offer for cases that are deemed appropriate, as PHP monitoring may allow a clinician to return to work while ensuring that patient safety is not compromised. Given the immense time, effort, and resources invested in training a health care professional, as well as the clinician shortage in many areas, the value of saving the careers of health care professionals with good prognosis is apparent.
Notably, the present study demonstrated that, even among those referred, recommendations and disposition varied greatly. Of the 570 cases that were reviewed, fewer than half of the health care professionals ultimately participated in PHP monitoring. The remainder either relinquished their licenses or were deemed inappropriate or ineligible for monitoring, were handled exclusively via the disciplinary arm of their licensing board, or were imprisoned or deceased. The frequency of referrals related to USB also varied across time. This may be partially explained by changes in the number of health profession populations served, variability in interpretation of legislative policy requiring that a practitioner meet diagnostic criteria for a potentially impairing condition to participate in the PHP, and societal evolution regarding the topic of sexual harassment and misconduct.
Importantly, professionals who were monitored by the PHP had lower rates of legal involvement and were more likely to settle out of court and avoid imprisonment, which likely reflects the lower severity of USB among individuals who were offered monitoring agreements. Indeed, in determining whether monitoring is appropriate for a physician, PHPs and state medical boards make a risk assessment that considers the severity of the USB: the number of times the behavior occurred, patient harm, patient characteristics, the character of the physician (including demonstrated remorse), prior professional misconduct or disciplinary actions, ethics violations, and the risk of reoffending.1 Monitored professionals were also more likely to face license restrictions but less likely to have their license revoked. This suggests that protective measures were enacted to allow them to continue practicing without endangering patients. Although some have argued that allowing any physicians who engage in PSM to return to practice after their offense unnecessarily endangers the public,35 the data show that many practitioners can work safely with patients following proper evaluation and treatment and with adequate oversight. A prior review of 120 cases of physicians in monitoring for boundary violations showed that over 85 percent completed monitoring without reemergence of PSM.26
Results of the present study suggest that careful evaluation and case staffing by the PHP (i.e., only offering monitoring agreements to the minority of referrals for whom it was deemed appropriate) may contribute to the positive outcomes that have been observed. Interestingly, a retrospective study demonstrated that physicians redisciplined by medical boards displayed more psychiatric illness, more unlicensed activity, and less sexual misconduct than physicians being disciplined for the first time.36 Thus, PHP monitoring may be an important part of risk mitigation for both patients and practitioners in cases of PSM. It is important to acknowledge that this does not mean that USB or PSM should ever be condoned or that PHP monitoring should be an alternative to appropriate criminal prosecution or civil litigation. In addition, disciplinary action by the licensing board, when warranted, should not be supplanted by PHP monitoring. Rather, monitoring may be appropriate as a concurrent or subsequent approach that helps restore the practitioner to mental and physical health, supports return to practice (if indicated), and provides additional safeguards to protect the public from harm. Our results regarding the potential of PHPs in the case of USB may be valuable for forensic psychiatrists involved in peer review committees and performing fitness-for-duty evaluations for health care professionals in the complex task of differentiating misconduct unrelated to mental illness and disability from that arising from mental illness and substance use.37
Similar to previous studies,6,7,11,13,38 the majority of USB victims in the present study were adults. Still, a concerning number of cases involved minors (22.8%), which was higher than reported in prior studies (i.e., 12.9-14.0%).6,38 As observed among health care professionals disciplined for USB in Canada,38 the United Kingdom,39 and Australia,40 the professionals in the current study were primarily physicians, and consistent with previous research,31,34,40,–,45 the most common medical specialties included adult primary care, psychiatry, and obstetrics-gynecology. It has been hypothesized that these specialties might present a higher risk for USB because of the disclosure of intimate information inherent to the relationship and the existence of longer term one-on-one treatment.40 Cases may be underreported across specialties, however, as these violations often occur in private spaces without eyewitnesses to validate their occurrence and few states mandate reporting if a physician becomes aware of a patient victim of USB.32
Also consistent with prior international studies,6,7,38,–,40,42,46,47 the overwhelming majority of USB exhibited by health care professionals in this study involved male-female dyads. Previous research has demonstrated that male physicians were more likely to initiate social relations and sexual contact with patients than female physicians,48 and male therapists were at higher risk of expressing sexual feelings and behaviors to patients compared with female therapists.49,50 Physicians with disciplinary actions related to USB or PSM are predominantly middle aged or older,6,7,26,31,45 which was reflected in the current results, although the professionals who participated in PHP monitoring were slightly younger (M = 46.84; SD = ±9.42) than those who did not.
Limitations
Results should be interpreted within the context of some limitations. First, the records spanned over 35 years and were of variable completeness and quality, as the standards for “opening a file” on a practitioner have evolved over time to become more stringent. Indeed, in many cases, data extraction involved reviewing multiple handwritten notes or reports that had been scanned into PDFs. Further, the list of health care practitioner license categories that were eligible for monitoring has expanded significantly over the past three decades. As a result, physicians may be overrepresented in the present sample. Notably, nurses were not included in the present study because they are not eligible for monitoring by the PHP involved in this study (there is a separate program specifically for nurses in the state). With regard to practitioner characteristics, demographic information was missing from many files, and it was not possible to compare the monitored and unmonitored groups regarding diagnoses, because these data were unavailable for most individuals in the unmonitored group. Finally, because of data collection methods (i.e., looking for evidence of each variable in the chart rather than obtaining information directly from the participants), the results may reflect an underestimate of some variables of interest.
Conclusions
Unprofessional sexual behavior by licensed health care professionals is a serious offense that merits full investigation and appropriate consequences. In cases where evaluation results and treatment response suggest that the practitioner would be successful returning to practice without recurrent USB, referral to a PHP may be beneficial. The PHP monitoring agreement parameters provide additional safeguards to protect the public from harm while supporting licensed health care practitioners in their return to practice and overall wellbeing. Findings from the current study provide insight into the characteristics of health care professionals who are referred for USB, the types of USB exhibited, the consequences they face, and the outcomes of monitoring. Such information might enhance transparency of the PHP process and inform policies to prevent USB by health care professionals. The success of PHPs as demonstrated by this study might be particularly relevant for forensic psychiatrists who are involved in conducting evaluations regarding fitness to practice. Future research should explore developmental and psychiatric risk factors for engaging in USB, as well as their relation to PHP outcomes. In addition, a deeper analysis is needed to determine the type(s) of treatment and components of monitoring for USB and co-occurring conditions that are associated with successful outcome.
Footnotes
Disclosures of financial or other potential conflicts of interest: None.
- © 2024 American Academy of Psychiatry and the Law
