Prevalence of sexual harassment and sexual assault from patient to provider among women in dermatology and across specialties

Abstract

Background

Sexual harassment (SH) includes unwelcome sexual advances, requests for sexual favors, and hostile conduct that targets someone based on gender and overlaps with some types of sexual assault (SA). SH/SA in health care can occur between providers or between patient and provider. Most studies of SH in medicine focus on SH perpetrated by one health care provider against another, with very few studies examining SH from patient to provider.

Objective

This study aimed to describe the prevalence and impact of SH/SA from patient to provider, with a particular focus on SH/SA experienced by dermatologists and trainees.

Methods

An anonymous electronic survey was sent to professional listservs and an online forum, which included representatives from multiple institutions, practice settings, and medical specialties. Trainees and dermatologists were targeted particularly.

Results

A total of 330 complete responses were included. In all, 83% of respondents reported experiencing SH from a patient. SH from a patient was more frequently reported by women compared with men (94% vs. 52%; p = .001). Behaviors consistent with SA were experienced by 31% of respondents and were more frequently experienced by women (35% vs. 15%; p = .001). Women were more likely to report that patient-to-provider SH contributed to burnout (33% of women vs. 9% of men; p = .002). Female trainees were significantly more likely to have experienced SH compared with female attendings within the past year (94 of 110 trainees [86%] vs. 83 of 127 attendings [65%]; p = .001). There was no significant difference in the proportion of women reporting ever experiencing SH when comparing dermatology and nondermatology specialties.

Limitations

The limitations of this study include the relatively small sample size, oversampling of trainees, and a gender-biased sample.

Conclusion

Patient-to-provider SH/SA is widespread, particularly among women and trainees, and may have a significant impact on burnout.

Background

Sexual harassment (SH) includes unwelcome sexual advances, requests for sexual favors, and hostile verbal or physical conduct that targets someone based on gender, whether or not sexual overtures are involved (Benya et al., 2018, Center, 2018). Sexual assault (SA) is defined by the Centers for Disease Control and Prevention as “a sexual act that is committed or attempted by another person without freely given consent of the victim or against someone who is unable to consent or refuse” (Basile et al., 2009). SA includes rape but also nonpenetrative acts, such as intentional sexual touching or noncontact acts of a sexual nature.

Civil rights activist Tarana Burke introduced the phrase “Me Too” in 2006 to raise awareness of the pervasiveness of SH and SA in our society. In 2017, use of the hashtag “#MeToo” by survivors of SH/SA propelled a global movement, highlighting the breadth of the problem and sparking a renewed discussion of awareness and intolerance (Kearl, 2018). A 2018 #MeToo study found that 81% of women and 43% of men had experienced SH and/or SA in their lifetime (Kearl, 2018); for 38% of affected women and 13% of affected men, SH or SA occurred at their workplace (Kane, 2018).

Most studies about SH in medicine focus on SH perpetrated by one health care provider against another, with very few studies examining SH from patient to provider. Heightened awareness of SH in general has highlighted the potential role of the patient as the perpetrator of SH. A 2018 Medscape survey suggested that physicians are often the target of SH, with 27% of physicians experiencing SH from patients. Dermatology was identified as the specialty reporting the most patient-to-provider SH, with a 46% prevalence (Kane, 2018). We previously reported an institutional pilot study that suggested that SH from patients may be even more frequent than found in the Medscape survey, with nearly 70% of dermatology providers reporting experiencing SH by patients (Notaro et al., 2019).

We sought to further describe the prevalence and impact of SH/SA from patient to provider with the hypotheses that SH from patients affects dermatology more than other medical and surgical specialties and female providers more than their male counterparts.

Methods

Study design

A link to an anonymous electronic survey was sent to professional listservs and an online forum. Groups including trainees and dermatologists were particularly targeted. Responses were collected between December 16, 2018 and December 31, 2018 using Research Electronic Data Capture. No incentives were provided to participants for survey completion. This study was reviewed by the University of Washington Human Subjects Division and determined to be exempt from review by the institutional review board (#STUDY00005548).

Statistical analysis was performed using STATA (version 14.0) and R (version 3.6.1). Fisher’s exact test and the Wilcoxon rank-sum test were used to compare categorical and ordinal responses between groups, respectively. Respondents with missing values were excluded from individual comparisons requiring those variables. P-values <.05 were considered statistically significant.

Sample

The sample included all members of a graduate medical education listserv, comprising resident trainees and faculty from multiple specialties at a single academic institution, a Facebook group of physicians and trainees (Physician Mom’s Group), and physician members of the Association of the Professors of Dermatology, spanning multiple institutions and practice settings.

Survey content

Survey questions included demographics and inquired about respondents’ experiences with and responses to SH or SA from patient(s). SH was defined per the U.S. Equal Employment Opportunity Commission definition as including “unwelcome sexual advances, requests for sexual favors, and other verbal or physical conduct of a sexual nature” (National Women's Law Center, 2018). SA was defined as any of the following behaviors: unwanted, intentional exposure of patient genitals; unwanted exposure to pornography or sexual content; and unwanted, intentional touching of a provider’s genitals, groin, or breasts.

Results

Survey respondents

A total of 362 survey responses were collected, and 32 incomplete responses were excluded from the analysis. Of these responses, 247 respondents identified as female, 79 identified as male, and 4 did not identify as either male or female or did not provide their gender. Other demographics of respondents are summarized in Table 1.

Table 1.

Respondent characteristics.

		Gender*
	All (N = 330)	Female (n = 247)	Male (n = 79)
Age, n (%)
25–29 years	60 (18.3)	46 (18.7)	14 (17.7)
30–34 years	104 (31.7)	79 (32.1)	24 (30.4)
35–44 years	102 (31.1)	84 (34.1)	17 (21.5)
45–54 years	37 (11.3)	23 (9.3)	13 (16.5)
≥55 years	25 (7.6)	14 (5.7)	11 (13.9)
Training, n (%)
Attending physician (MD/DO)	171 (52.0)	127 (51.4)	42 (53.2)
Resident or fellow physician (MD/DO)	147 (44.7)	110 (44.5)	36 (45.6)
Other	11 (3.3)	10 (4.0)	1 (1.3)
Specialty, n (%)
Dermatology	116 (37.8)	85 (37.4)	30 (39.0)
Family medicine/internal medicine/pediatrics	118 (38.4)	92 (40.5)	25 (32.5)
Other	73 (23.8)	50 (22.0)	22 (28.6)

DO, Doctor of Osteopathic Medicine; MD, Doctor of Medicine.

^*Four respondents who did not identify as male or female were excluded from this analysis.

Survey results

Frequency of sexual harassment

A total of 274 of 330 respondents (83%) reported ever experiencing SH from a patient, and 213 (65%) reported experiencing SH from a patient in the last year. Of those who experienced SH from a patient in the last year (n = 213), 135 (63%) reported experiencing SH 1 to 3 times in the past year and 17 (8%) reported experiencing SH more than once a month.

SH from a patient was reported by significantly more women compared with men at any time (231 of 247 [94%] vs. 41 of 79 [52%]; p < .001) or within the last year (183 of 247 [74%] vs. 29 of 79 [37%]; p < .001). The single most frequently experienced type of SH was comments on the provider’s appearance, with 281 of all 330 respondents (85%) and significantly more women than men (230 of 247 [93%] vs. 48 of 79 [61%]; p < .001) reporting this type of SH. The results are summarized in Table 2. Similarly, a high percentage of providers reported being asked about marital status or relationship status by a patient (72%) and being told jokes or stories of a sexual nature (63%).

Table 2.

Survey results.

		Gender*
	All (N = 330)	Female (n = 247)	Male (n = 79)	p-value†
Ever experienced SH, n (%)	274 (83.0)	231 (93.5)	41 (51.9)	<.001
Experienced any of these behaviors, n (%)
Comments on your appearance	281 (85.2)	230 (93.1)	48 (60.8)	<.001
Asked about marital or relationship status	237 (71.8)	200 (81.0)	34 (43.0)	<.001
Asked on a date	77 (23.3)	73 (29.6)	2 (2.5)	<.001
Told jokes or stories of a sexual nature	208 (63.0)	175 (70.9)	30 (38.0)	<.001
Any of the above	300 (90.9)	240 (97.2)	57 (72.2)	<.001
Experienced any of these unwanted behaviors, n (%)
Intentional exposure to genitals	86 (26.9)	76 (31.5)	8 (10.7)	<.001
Exposure to pornography or sexual content	14 (4.4)	10 (4.1)	4 (5.3)	.75
Intentional touching of your genitals, groin, breasts	14 (4.4)	12 (5.0)	1 (1.3)	.31
Any of the above	98 (30.6)	85 (35.3)	11 (14.7)	.001
Reasons for not reporting an incident of SH/SA, n (%)	(N = 218)
Did not think reporting would have productive consequences	146 (67.0)
Was not sure if serious enough	138 (63.3)
Did not have time	96 (44.0)
Was not sure it was SH/SA	91 (41.7)
Did not think patient intended to harass	91 (41.7)
Did not know how	73 (33.5)
Did not want anything to happen to the patient	51 (23.4)
Was afraid of negative patient satisfaction	29 (13.3)
Felt helpless about what happened	28 (12.8)
Was afraid of negative consequences from supervisors	25 (11.5)
Felt ashamed	20 (9.2)
Felt hopeless about what happened	13 (6.0)

SA, sexual assault; SH, sexual harassment.

^*Four respondents who did not identify as male or female excluded from this analysis.

^†Fisher’s exact test (categorical variables) or Wilcoxon rank-sum test (ordinal variables).

SH was much more frequent among female respondents; thus, a subgroup analysis was performed of female respondents (n = 247). Among these women, there was no significant difference in the proportion reporting ever experiencing SH when comparing dermatology and nondermatology specialties (77 of 85 [91%] vs. 136 of 142 [96%]; p = .15) or between trainees and attending-level physicians (121 of 127 [95%] vs. 102 of 110 [93%]; p = .42). When asked about just the last year, female trainees were significantly more likely to have experienced SH compared with female attendings (94 of 110 [86%] vs. 83 of 127 [65%]; p = .001).

Impact of sexual harassment

The majority of respondents (218 of 300 [73%]) who reported experiencing an incident of unwanted sexual behavior from a patient did not report the incident in an official capacity, with a lack of reporting being more common among women than men (184 of 230 [80%] vs. 32 of 67 [48%]; p = .001). When asked if they knew how to report unwanted sexual behavior at work, 153 of 308 respondents (50%) indicated “probably yes” or “definitely yes.” Of those who experienced an episode of SH/SA and did not report it in an official capacity (n = 218), the most common reasons are summarized in Table 2 and included “did not think reporting would have productive consequences” (146 of 218 [67%]) and “was not sure it was serious enough” (138 of 218 [63%]). A total of 214 of 308 respondents (70%) indicated that they would feel supported if they reported an incident by a patient whereas 94 respondents (30%) indicated that they would not or were unsure, with no significant differences between women and men (p = .10). Few respondents (5 of 252 [2%]) sought mental health counseling after experiencing unwanted sexual behavior from a patient.

Seventy-six of 252 respondents (30%) reported that experiencing patient-to-provider SH contributed to burnout, with women (72 of 216 [33%]) significantly more affected than men (3 of 34 [9%]; p = .002). Of those who reported prior SH, 40 of 254 (16%) reported having terminated a relationship with a patient due to SH, and 18 of 253 respondents (7%) endorsed that experiencing unwanted sexual behavior from a patient had affected their career path, with no significant difference between genders (p = .58).

Sexual assault

When asked if they had ever experienced behaviors consistent with SA, 98 of 320 respondents (31%) indicated that they had experienced one or more behaviors, with more women (85 of 241 [35%]) than men (11 of 75 [15%]) experiencing such behaviors (p = .001; Table 2).

Discussion

Our study confirms the pervasiveness of patient-to-provider SH/SA, with 83% of all respondents experiencing SH and 31% reporting behaviors consistent with SA. We found a prevalence of patient-to-provider SH that was significantly higher than reported in the Medscape survey (Kane, 2018). We cannot exclude a contribution from response bias to our results, but we hypothesize that our very high reported rate of SH may indeed be valid because we asked questions about experiencing specific behaviors (e.g., being asked on a date) rather than using the general phrase “sexual harassment.” This was intentional in our study design to lessen the tendency to minimize incidents by those who have experienced SH/SA, which is based on the literature that suggests that a substantial proportion of those who have experienced sexual victimization (including rape) do not label their experiences as such for multiple and complex reasons (LeMaire et al., 2016).

We hypothesized that dermatology may harbor an increased risk for SH/SA from patients because of examinations that frequently involve nudity, disrobing, and examining genitalia, as well as physical contact with patients’ skin. In contrast to the Medscape survey and to our own hypothesis, however, dermatologists in our study did not report more SH from patients compared with those in other specialties. Although more data should be obtained to confirm this finding, we conclude that other factors, such as gender of the provider and perceived power differential (e.g., being a trainee), may play a larger role than medical specialty in patient-to-provider SH.

In our study, female providers reported experiencing rates of SH from patients nearly two times higher than men (94% vs. 52%, respectively). This gender disparity is substantiated by other recent studies of SH in health care (Benya et al., 2018, Fnais et al., 2014, Jagsi et al., 2016). An unanticipated finding was the high proportion of men who reported experiencing SH from patients; we believe this highlights that measures to prevent and address this behavior will benefit providers of all genders.

We also hypothesized that providers who were more junior, including trainees, would report more SH from patients than those more senior. However, we found no significant difference between female trainees and female attendings when asked if they have ever experienced SH from a patient, with both groups reporting high rates. When asked about experiencing SH in the past year, though, significantly more trainees reported experiencing SH from patients compared with attendings, with a trend toward higher rates of burnout and impact on career path reported by trainees compared with attendings. We posit that trainees are still accumulating SH experiences, whereas attendings may experience fewer episodes, perhaps due to higher perceived authority or development of effective avoidance strategies.

Experiencing SH has been linked to symptoms of depression, stress, and anxiety, as well as decreased job satisfaction, productivity, and performance (Dzau and Johnson, 2018). Specifically within medicine, providers who experience SH reported lower mental health, job satisfaction, and sense of safety at work, as well as increased intention to leave a job (Vargas et al., 2020). Our findings reinforce that patient-perpetrated SH is associated with feeling burned out and even, for a small portion of those affected, with an impact career path.

In our study, nearly three quarters of respondents who reported experiencing unwanted sexual behavior from a patient did not report the incident in an official capacity. This parallels the results of a recent study that showed that an overwhelming majority of dermatology residents experiencing SH did not officially report it (DeWane et al., 2020). The reasons for not reporting an incident of SH/SA are complex. Multiple studies have found uncertainty regarding the seriousness of the event to be a common cause for nonreporting of SH (DeWane et al., 2020, Vargas et al., 2020). This is reinforced by our data, which showed that 63% of those who had experienced but not reported an incident chose not to report because they were “not sure it was serious enough.” Furthermore, the majority of respondents (67%) did not think reporting would have productive consequences. Frank discussion and increased educational efforts around what does and does not constitute SH/SA from patients and ensuring that reporting mechanisms are easy to use, productive, and supportive for those who report are essential.

In the current era of medicine, patients can review their providers either publicly or through patient satisfaction surveys, which may be tied to professional advancement or financial incentives. Providers may fear repercussions of reporting patient-perpetrated SH. Despite this, only 13% of our respondents who had experienced but not reported SH/SA cited negative patient satisfaction as a reason for not reporting.

Given possible bias and emotional triggering of respondents with the term “sexual violence” or “sexual assault,” we intentionally did not use this terminology in our survey but instead asked about specific behaviors. We found that nearly one-third of respondents had experienced behaviors consistent with SA; the most frequent was unwanted/intentional exposure to a patient’s genitals. Fewer respondents (4%) experienced unwanted touching of their breasts or genitals by patients. We theorize that although boundaries are often clear with regard to unwanted physical contact, grey area behaviors such as unwanted exposure to genitals may be less clear to perpetrators, particularly in the medical setting.

Limitations

Our study has several limitations. The primary limitation is our sample size and population. Our sample size, although large compared with other studies of this type, is small when compared with the physician group at large. Because of the methods we used to disseminate our survey, we are unable to calculate a response rate; therefore, our data may not be representative of our intended group, and our responses may be particularly subject to selection and/or response bias. We were particularly interested in the impact of SH/SA toward and on trainees, making purposeful attempts at oversampling trainees; therefore, our data are biased toward academic training centers. Our sample was weighted toward those who identify as women (75%), especially when considering that women make up only about one third of the current physician workforce (Association of American Medical Colleges, 2017). Our questionnaire was disseminated nearly 2 years ago, and our findings may differ were the study conducted today. Lastly, we did not specifically investigate the prevalence and impact of SH/SA on providers within minority groups (e.g., gender or racial/ethnic minorities). Based on prior research showing that these groups are at a higher risk of SH/SA in general (Steelfisher et al., 2019), further studies to address this knowledge gap are needed.

In our group, male respondents tended to be older than female respondents (≥35 years old: 62% vs. 39%, respectively; p < .001) and were more likely to be attending physicians (60% vs. 38%, respectively; p = .007). Therefore, our data may be subject to response bias, with those who have experienced SH/SA from a patient more likely to respond. The true incidence of SH/SA from patient to provider may be higher among our undersampled group (older men) and lower among our oversampled group (younger women).

Responding to sexual harassment/assault by patients

Responding to SH/SA by patients must begin with a cultural shift in medicine. We must not tolerate or make excuses for this behavior. Any intervention for responding to SH/SA should be proactive and done in a manner that preserves the safety and personal dignity of the affected provider. Strategies for responding to SH/SA from a patient must be tailored to the situation and can broadly be divided into responses by the individual, responses by a bystander (responding to witnessed SH/SA of another), and institutional policies addressing such behavior. If providers experience SH from a patient, they must first determine if they feel physically safe; if they do not, they should remove themselves from the situation (e.g., leave the room). If they feel physically safe, practical strategies for interrupting the behavior “in the moment” include directly asking the patient to stop, naming the behavior and its effect on the victim, and redirecting. Table 3 provides a language toolbox of sample phrases that can be used.

Table 3.

Strategies for responding to patient-perpetrated sexual harassment/assault in the moment.

Asking to stop/naming the behaviour	“That’s harassment. I want you to stop doing that.”
	“You just touched me. Stop doing that.”
	“That joke isn’t funny to me. Let’s not say that again.”
Appealing to behavioral standards	“Jokes like that one don’t feel right for this professional space.”
	“Would you tell that joke to a family member?”
Redirecting	“Let’s keep this professional.”
	“Let’s keep this appointment focused on you today.”
	“That feels like a very personal comment to me; please focus on my professional role”

We agree with Mello and Jagsi (2020) that physicians have a duty to intervene when they observe SH/SA, similar to physicians’ moral obligation to intervene when they observe an impaired colleague. Bystander intervention by an attending physician in response to witnessed harassment of a trainee is particularly imperative for fostering a positive and safe learning environment. The supervising physician should later debrief and discuss the incident with the trainee and address and/or facilitate reporting. In addition to these strategies, helpful bystander interventions include removing the target of SH/SA from the situation and more indirect approaches, such as discussing the behavior with the perpetrator at a later time or requesting institutional sanctions against the perpetrator (Mello and Jagsi, 2020).

On an institutional level, maintaining a culture that prevents SH/SA by patients must have several facets. Leaders can promote a culture of diversity, inclusion, and respect into institutional policies (Dzau and Johnson, 2018). Having a policy that clearly and visibly (e.g., with signs posted in clinic spaces) states that harmful behavior such as SH/SA will not be tolerated by anyone, including patients, is an important first step in establishing behavioral expectations. An example of such a sign is provided in Fig. 1. Addressing patient-perpetrated SH/SA through formal SH training may be helpful for recognition and skill-building for providers. Formal bystander training may be a particularly effective intervention for SH/SA (Potter and Moynihan, 2011). Lastly, environments with positive role models can lessen sexually harassing behavior, and women who work in gender-balanced workplaces experience less SH (Kabat-Farr and Cortina, 2014). Promoting and retaining women to parity in high-level institutional leadership positions will be important to fostering antiharassment cultures.

Fig. 1.

Example of a sign that can be used in clinical spaces to set behavioral expectations.

Conclusion

Despite some limitations, our data highlight the very real problem of patient-perpetrated SH and SA occurring toward physicians, particularly women. Trainees in particular are on the front line of patient interactions and potential SH/SA exposure. Our findings highlight a critical need to educate all of those involved, but particularly those who work with trainees, about the frequency of patient-to-provider SH/SA, provide training on SH/SA from patients, and provide training on how to support providers when SH/SA from patients occurs. Efforts to address SH/SA in the health care environment need to come from the highest levels of institutional administration, as well as on a local and individual level. These efforts should include acknowledgment of the frequent occurrence of SH/SA from patients, further research into the impact of this behavior on physician burnout, and the ethical issues of physicians reinforcing boundaries that are sometimes blurred in a service industry.