Residents’ personal health and health care represent an important yet understudied aspect of physician development. During training, residents’ own physical and mental health needs may intensify (1), as they face numerous challenges that may influence their personal health and health care-seeking practices (2, 3), such as intense schedules, extensive workloads, and increasing levels of responsibility in making complex medical decisions. Heightened stress, fatigue, and a sense of imbalance in life are common (4–6) and may adversely affect residents’ performance, patient care, professionalism, and empathy (7–9). Such stressors also confer increased vulnerability to depression, anxiety, burnout, substance abuse, and other types of distress (2, 9–13). The personal health habits of physicians affect the degree to which physicians counsel their patients in healthy life styles (14, 15). Yet, minimal previous research has focused on residents’ personal health and health care.
Residents have multiple overlapping roles—as trainees, physicians, employees, colleagues, and supervisors—so it is possible that they have unique views and concerns related to obtaining personal health care at their training institution. In one study, residents cited confidentiality, privacy, and being seen by someone known to them through their work as a resident, as barriers to obtaining care (1). Yet, the relationship between where residents seek care (or would, if they needed to) and their concerns about various aspects of receiving care (e.g., confidentiality, quality of care, financial issues, or other factors) has not been previously investigated. In addition, particular types of symptoms or illnesses (e.g., those that are more highly stigmatized, such as mental health issues) may be associated with residents preferring to seek outside care, yet we know of no previous study that has examined this question directly.
To better understand these issues, we examined attitudes and behaviors related to the personal health and health care of residents in a variety of training programs at one institution. Findings related to residents’ views and concerns (e.g., confidentiality) related to obtaining care at or outside their own training institution are reported here. We hypothesized that residents would express concerns about confidentiality when receiving health care at their own training institutions, and that they would prefer care outside of their institution for stigmatizing illnesses.
A self-report survey on special issues in personal health care of residents during postgraduate training was developed and pilot tested. The survey was based on prior work by two of us (LR, KAGH) (16, 17); it included 241 items with scaled or counted responses, 11 demographic questions, and 2 open ended questions. Survey content domains included personal health and health care behaviors, health care preferences, experiences as trainees with confidentiality and overlapping roles, attitudes concerning compassion and empathy toward ill patients, and 6 vignettes addressing stigma and fears of professional jeopardy associated with becoming ill or seeking treatment. The survey was distributed to all 217 second and third postgraduate year residents at the University of New Mexico Health Sciences Center (UNMHSC) in 2001 (interns were excluded because the survey was administered very early in the training year; interns’ experiences as residents therefore would have been extremely limited). The largest academic medical center in New Mexico, UNMHSC provides care through the UNM School of Medicine clinics, the UNM Hospital with a Level I Trauma Center, as well as other centers including the Cancer Research and Treatment Center, UNM Psychiatric Center, and Sleep Disorders Center. The survey was voluntary and confidential; respondents received $50 in compensation for their time and effort. This study, funded by the Arnold P. Gold Foundation, was approved prospectively by the UNMHSC Human Research Review Committee.
Responses to personal health and attitude items were from nine-point labeled rating scales appropriate to item content. Responses for conceptually related sets of items were subjected to repeated measures multivariate analysis of variance (MANOVA) with Items as within-subjects repeated measures. The full MANOVA analysis structure for item sets was Item (within subjects repeated measures) X Gender (between subjects) X Residency Area (primary care versus specialty, between subjects) X Training Level (PGY 2 versus 3 or 4, between subjects). Single items were subjected to Gender X Residency Area X Training Level ANOVAs. No effects of Training Level were found. For the subset of respondents who had obtained health care during the past year, sets of attitude items were subjected to repeated measures Item X Gender X Residency Area X Care Site (at versus outside of the training institution) MANOVAs, and single items were subjected to Gender X Residency Area X Care Site ANOVAs. Correlations among personal health ratings and attitudes were examined.
Of the 155 residents who returned surveys (response rate 71%), 14 were excluded for substantially incomplete responses to items reported here. Table 1
provides respondent characteristics. Eleven percent of respondents were psychiatry residents. Gender groups did not differ by training level or area of residency. Residents reported good physical and mental health (means = 7.5 and 7.3 on a scale of 1=“poor” to 9=“excellent”); men reported slightly but significantly better mental health than did women (respective means = 7.5 versus 7.1, p<0.01).
Concerns About Seeking Health Care at the Training Institution
Residents rated the concern (scaled from 1=“no concern” to 5=“some concern” to 9=“great concern”) they experience when seeking health care at their training institution pertaining to 9 issues (Figure 1
). Respondents reported moderate concern regarding being seen by other residents whom they know and about personal information from their medical record being seen by others at the institution (means=5.71–6.11). They were also concerned (less so, but still at a moderate level) about being treated or observed by attendings who might be their supervisor at some time or who knows them, as well as about being treated by a resident or medical student who is not fully trained, being treated by a medical student whom they supervise, or being observed by residents or hospital staff who know them (means=4.59–5.11). They were less concerned about being observed by their patients when in a clinic for treatment (mean=3.79; Item main effect F (8, 126)=14.15, p<0.0001, maximum Cohen’s d=0.86). Respondents showed substantial diversity of opinion around these issues (pooled SD=2.70).
Women consistently reported greater concern than did men for the 9 issues considered (respective means = 5.39 versus 4.54, Gender main effect F (1, 133)=6.22, p<0.02, d=0.42). Residents in primary care areas (family practice, internal medicine and pediatrics) expressed greater concern than did residents in specialty areas (all other areas including psychiatry) about personal medical records being seen by others at their institution, being observed or treated by other residents or attendings they know, or being observed by staff (means=5.02–6.45 versus 4.17–5.78, Item X Residency Area interaction F (8, 126)=2.17, p<0.04, d = 0.25 to 0.37). For all respondents, poorer self-reported mental health was associated with greater concern about being treated by another resident whom they know or by an attending who is also a supervisor (both r=−0.20, both p<0.02).
Psychiatry residents expressed greater concern than did residents in primary care or other specialties about being observed by their patients when in a clinic for treatment (means=4.88 versus 3.36–3.83, maximum d=0.56) and being treated by a medical student whom they supervise (means=5.62 versus 4.59–4.87, maximum d=0.38), although the differences were not statistically significant.
Preferences for Receiving Care At or Outside of the Training Institution by Health Issue
Residents indicated where they would prefer to receive health care (scaled from 1=“much prefer care at my institution” to 5=“site of care makes no difference” to 9=“much prefer care outside my institution”) for 28 health issues (Figure 2
). Residents would prefer care outside of their training institution for HIV, other STDs, problems with prescription drugs, other drugs, or alcohol, marital or relationship problems and other personal counseling issues, anxiety, depression, and eating-related problems (means=6.97–8.13), and would prefer care outside of their institution, but less strongly, for gynecological or urological problems, sleep-related problems, elective surgery, and pregnancy related issues (means=6.03–6.25). Respondents indicated that site of care would make no difference for fatigue, serious infections, gastrointestinal problems, cancer, dermatological problems, physical injury, chest pain, ophthalmic problems, health maintenance or prevention, pain, diabetes, allergy, arthritis, or minor infections (means=3.99–5.21; Item main effect F (27, 107)=14.85, p<0.0001, maximum d=1.82).
Women preferred outside care more than did men (means=6.33 versus 5.62, gender main effect F (1, 133)=8.16, p<0.01, d=0.48), particularly for pregnancy, cancer, ophthalmic problems, and health maintenance (d=0.48 to 0.57). Residents in primary care areas preferred outside care more than did residents in specialty areas (means=6.41 versus 5.54, Residency Area main effect F (1, 133)=12.28, p<0.001, d=0.59), particularly for pain, fatigue, allergy, arthritis, sleep problems, minor infections, physical injury, pregnancy, gastrointestinal problems, and gynecological or urological problems (d = 0.49 to 0.75; Item X Residency Area interaction F (27, 107)=1.75, p<0.03). There was no consistent pattern of relationships between care site preferences for specific health issues and self-reported current health.
Psychiatry residents preferred outside care for HIV and other STDs (means=8.02–8.34), for the set of mental health issues considered (means=5.97–8.16), and for pregnancy, gynecological problems, and elective surgery (means=5.81–6.80), as did residents in other areas. Psychiatry residents preferred outside care somewhat more strongly than did residents in other specialty areas for elective surgery (d=0.65), anxiety (d=0.49), and eating-related problems (d=0.35) and somewhat less strongly than did residents in primary care areas for sleep-related problems (d=−0.40), although none of the differences reached statistical significance with the low power to detect group differences associated with our small sample of psychiatry residents.
Influences on Preference for Receiving Care At or Outside of the Training Institution
Residents indicated how seven factors influenced preferring care at versus outside their training institution (scaled from 1=“much prefer care at my institution” to 5=“site of care makes no difference” to 9=“much prefer care outside my institution”; see Figure 3
). Residents preferred care at their training institution in terms of expense, knowing how to obtain care, and fitting care into their schedules (means=2.56–3.37), and they indicated that site of care makes no difference in terms of quality and time needed (means=4.43–4.54); residents preferred to receive outside care when responding to influences of confidentiality and prevention of embarrassment (means=7.09–7.68; Item main effect F (6, 128)=87.55, p<0.0001, maximum d=2.38).
Although a trend was identified in which women preferred outside care more strongly than did men, the difference was not significant across the seven influences evaluated (means=4.90 versus 4.53, Gender effect F (1, 133)=2.80, p<0.10, d=0.28). Residents in primary care areas preferred outside care more than did residents in specialty areas across the seven influences evaluated (Residency type effect F (1, 133)=5.50, p<0.03, d=0.39). There was no consistent pattern of relationships between care site preferences in terms of the influences considered and self-reported current health.
Psychiatry residents preferred care at their training institution less strongly than residents in other specialty areas when taking into account knowing how to obtain care (means=4.68 versus 2.70, d=0.92, p<0.05 by post hoc Fisher’s LSD) and when thinking about how to fit care into their schedules, although the latter trend did not reach statistical significance (means=4.15–5.30 versus 2.58–3.62, d=0.73–0.78, both p<0.06).
Respondents indicated that they receive relatively little information from their training institution regarding their rights to confidentiality related to their personal health care (mean=3.07 on a scale of 1=“not at all” to 5=“somewhat inform” to 9=“inform very well”); and residents were less well informed regarding safeguards to protect the privacy of residents when treated for personal health issues at their training institution (mean=2.62; Item effect F (1, 133)=26.94, p<0.0001, d=0.23). A majority of respondents did not know whether their institution had confidentiality policies regarding residents who develop physical or mental health problems, 21% reported that their institution had such policies, and 2% indicated that their institution did not have confidentiality policies.
Psychiatry residents felt somewhat better informed than residents in other areas about their health care confidentiality rights and safeguards to protect their privacy when treated for personal health issues at their training institution—trends that did not reach significance (means = 4.30 versus 2.96 to 3.09, maximum d=0.67, both p<0.07 by post hoc Fisher’s LSD).
A majority of psychiatry resident respondents did not know whether their institution had confidentiality policies regarding residents who develop physical or mental health problems; 25% reported that their institution had such policies for mental health problems, and 6% for physical health problems.
All respondents had health insurance. Insurance allowed care only at the training institution for 37% of respondents, either at or outside of the training institutions for 50%, and only outside of the training institution for 2%. Eleven percent did not know where their insurance allowed care. Residents indicated that available health insurance was affordable (mean=6.53 on a scale of 1=“not affordable at all” to 5=“somewhat affordable” to 9=“very affordable”). Almost two thirds (63%) of respondents could not afford care outside of their training institution using their own funds if necessary (responses of 1, 2, or 3 on a scale of 1=“not at all” to 5=“somewhat” to 9=“completely”), 23% could afford outside care somewhat (responses of 4, 5, or 6), and only 14% could readily afford outside care (responses of 7, 8, or 9). Residents rated having health insurance that allows obtaining care outside of their training institution for any health problem as important (mean=6.59 on a scale of 1=“not important at all” to 5=“somewhat important” to 9=“very important”). Women rated having health insurance that allows obtaining outside care as more important than did men (means=6.98 versus 6.20, Gender effect F (1, 133)=4.43, p<0.04, d=0.36). Psychiatry residents did not differ from residents in primary care or other specialty areas in responses concerning health insurance.
Health Care Use by Residents
Three-fourths (73%) of respondents had visited a clinic or physician for physical or mental health care at least once during the past year. More women than men had obtained health care (81% versus 67%, chi-square (1)=9.61, p<0.06). Overall, 27% of respondents had not obtained health care during the past year, 44% had visited a physician or clinic only at their training institution, 21% had obtained care both at and outside of their training institution, and 8% had obtained care only outside of their institution.
Of the residents who obtained care, 40% went outside of their training institution for some or all visits to a clinic or physician. More women than men who sought care used some outside care (55% versus 25%, chi-square (1)=9.61, p<0.01). Among residents who had visited a clinic or physician during the past year, outside care was used by more than half (56%) of those with health insurance allowing outside care, and by one fourth (26%) of those with health insurance allowing care only at the training institution.
Respondents who had obtained some outside care during the past year did not differ from respondents who had obtained care only at the training institution in self-reported current physical or mental health, degree of concern when seeking health care at their institution (items in Figure 1
), preference for location of care for specific health issues (items in Figure 2
), or influences on preference for location of care (items in Figure 3
). However, residents who had used outside care were better able to afford outside care using their own funds if necessary than residents who had obtained care only at their training institution (means=4.77 versus 3.52 on a scale of 1=“not at all” to 5=“somewhat” to 9=“completely,” Care Site main effect F (1, 95)=6.06, p<0.02, d=0.50).
Residents rated their likelihood (scale of 1=“no chance” to 5=“somewhat likely” to 9=“certain”) of visiting a physician or clinic at or outside of their training institution during the next 12 months if they were to need physical or mental health care. Residents indicated that they were somewhat likely to seek care at their training institution (mean=6.16), and less likely, but still somewhat likely, to seek outside care (mean=4.15, Item effect F (1, 133)=3.33, p<0.08, d=0.83). Women and men indicated similar likelihood of seeking care at their training institution (means=6.03 versus 6.29, d=0.11), but women indicated greater likelihood of seeking care outside of their institution than did men (means=4.66 versus 3.65, Item X Gender interaction F (1, 133)=4.61, p<0.04, d=0.42).
Twenty-five percent of psychiatry resident respondents had obtained health care only at their institution during the past 12 months, 13% had obtained care only outside of their institution, 31% had obtained care both at and away from their training institution, and 31% had not obtained health care. Psychiatry residents indicated that they were somewhat likely to visit a physician or clinic at (mean=5.94) or outside of (mean=4.56) their training institution during the next 12 months if they were to need physical or mental health care.
Over 100,000 residents are in training at any given time in the U.S., and they represent the future of medicine (18). Little is known, however, about the personal health care needs, attitudes, and behaviors of resident physicians. In examining residents’ attitudes toward seeking care from within or outside one’s own training institution, we confirmed our hypotheses that residents: 1) would be concerned about seeking care at their own training institutions due to worries about confidentiality pertaining to their own health issues, and 2) that residents would express a preference for seeking care outside of their home institution for stigmatizing illnesses. Specifically, residents were concerned about being seen or treated by other residents, attendings, staff, or students, as well as about the privacy of their medical records. Diversity in opinions indicated that residents vary in their degree of concern or worry. Women tended to be more concerned than men, as did residents in primary care areas and, in several areas, residents with poorer self-rated mental health showed higher levels of concern. Residents’ responses also indicated that protection of confidentiality and prevention of embarrassment substantially influenced preferences for care outside of their training institution.
Stigma and Health Care Preferences
Stigma appeared to influence the degree to which residents prefer care outside their home institution. Residents’ overall pattern of responses can be seen as a map of stigmatized health issues in society at large. Residents appear similar to the rest of the population in preferring to keep stigmatizing illnesses confidential. Although our respondents included only a small number of psychiatry residents, a recent study of psychiatry residents in Manhattan found that they perceived more stigma associated with psychotropic medication use than with obtaining personal psychotherapy (19). It remains unclear under what circumstances, and where, residents in various specialties would seek care for such conditions as depression, anxiety, or substance abuse. This is an important area for further research, given the finding in a large survey of residents that over two thirds reported having observed a colleague working “in an impaired condition;” among causes of observed impairment, “emotional problems” were cited commonly (mentioned by 37% of respondents) (20). Graduate medical educators should thus be attuned both to the potential for impairment as well as the likelihood of stigma related to seeking care for numerous health issues.
A consistent gender pattern also emerged in our findings. Resident women expressed more concern about receiving care at one’s own institution, and they more strongly preferred outside care for specific illnesses. Women also felt more strongly that having insurance allowing outside care was important, and more women than men had actually obtained outside care during the preceding year. Women, finally, endorsed worse mental health compared to male respondents.
These gender results are open to multiple interpretations. While it is possible that women in general are more sensitive to health care privacy and confidentiality considerations, it is also possible that women residents feel more vulnerable overall than their male counterparts, possibly translating into greater preference for care outside one’s training institution. Women’s reproductive health needs may also lead to tendencies to prefer outside care. Our data also build upon prior research on women trainees. In a multisite study examining medical students’ concerns about personal illness and vulnerability, we reported that women were more worried about the possibility of developing a number of health problems (including anxiety and depression), as well as about their risk of academic jeopardy if their Dean’s office were to learn of a current health problem (16). Another study sampling Canadian residents reported that women felt both more stressed and more intimidated than did men during residency (5). A large survey study of second-year residents found that a majority of women reported at least one instance of perceived sexual harassment or discrimination (20). Such findings highlight the particular stressors felt by women residents, which could conceivably affect preferences for seeking care at one’s training institution. Thus, relationships among gender, residency-related stress, and personal health care deserve further attention.
We also found differences in views between primary care residents and those in specialty programs. Primary care residents expressed more concerns about seeking care at one’s own institution. It is possible that residents in primary care specialties receive more training in issues of confidentiality and privacy, or that they are more attuned to these concerns because of the reality that health care most often occurs in the primary care setting, thereby putting primary care residents at a distinct disadvantage, being more likely to see a colleague or supervisor for their medical problem. We did not find any consistent pattern of differences between psychiatry residents and other residents in preferences for outside care or influences on seeking care.
Among respondents who had received any health care in the prior year, approximately half of the women and one fourth of the men had obtained some outside care. These residents were not more ill nor did they express different attitudes toward seeking outside care; they were, however, more able (by self-report) to afford outside care if necessary. Among those whose insurance only covered care at the training institution, 26% had nevertheless obtained outside care. Thus, a substantial minority of residents did obtain outside care—some at their own expense. Given the significant financial stress faced by many residents, the financial aspect of residents’ health care concerns should receive further attention.
Our study was limited by only surveying residents at one training institution. These residents’ views may not be representative of residents nationwide. On the other hand, we had a high response rate (n=155, or 71%) and multiple disciplines involved, reducing the chances that the responses, at least for this institution, were substantially biased. The data are also limited by the somewhat theoretical nature of some of the questions, e.g., asking residents about where they would prefer to seek care if they had a given illness. Whether actual behavior would conform to predicted behavior is unknown. One additional limitation is the absence of data regarding the actual incidence of medical problems requiring care or general health maintenance needs among the residents we surveyed.
Findings in Context of Prior Work
Research examining the perspectives and behavior of residents relating to their own personal health care is scarce. One study reported that residents commonly self-prescribe medications (21), raising questions about why some residents do not seek care from other physicians for symptoms or perceived conditions. The most informative study to date about resident health care surveyed housestaff in four internal medicine programs in the United States (1). A substantial minority (37%) of these residents had no primary care physician or reported being their own primary care physician (12%). The intensity and unpredictability of their schedules represented major barriers to obtaining care, as did confidentiality and privacy concerns, along with a general sense of lack of support from colleagues and programs. Among those respondents providing additional written comments, nearly half (44%) highlighted concerns relating to privacy or confidentiality, access to certain types of care, or financial incentives related to where they obtained care. Our findings build upon these results in suggesting that stigma may also influence residents’ health care views and behaviors, including where they would obtain care.
Our findings echo themes that emerged in an earlier, multisite collaborative study on the personal health care experiences of 1,027 medical students at nine U.S. medical schools (16, 17). In that study, two of us (LWR, KAGH) found that students reported high rates of medical needs—and nearly half of the respondents endorsed at least one issue related to mental health or substance use (16). Many students reported difficulty obtaining care or had informally sought care (i.e., obtained a “curbside” consult), primarily due to time demands, cost, excessive waits, and concerns about confidentiality (17). Medical students in that study reported preferring outside care for certain (generally more stigmatizing) illnesses, were concerned about confidentiality, and expressed a desire for insurance that would cover outside care. Students were concerned about academic/professional jeopardy related to specific health problems, particularly mental health issues.
It appears that the sense of vulnerability related to personal health care postulated here may have its roots in earlier stages of medical training, or at least continues across the continuum of medical training. Residents rely on their supervisors, training programs, and home institutions for their education, evaluations, livelihood, and for professional recommendations. This dependency creates potential vulnerabilities that have been relatively understudied. It is plausible that confidentiality about personal health concerns may be salient to residents precisely because of their status and perceived vulnerability, although this requires empirical validation.
Patienthood in the Physician Experience
Physicians need health care themselves, sometimes for serious conditions; given that physicians’ own experiences as patients may in turn affect their own clinical practices, it is important for us to better understand the roles that stigma and vulnerability play in physicians’ perspectives on personal health care and their own health care behaviors. Recent work on physicians as patients, involving in-depth interviews with practicing physicians who had become patients with serious medical disorders (e.g., HIV, cancer) suggests that patienthood often transforms physicians’ views of their own delivery of care (22). These doctors, for instance, emphasized that their sensitivity and empathetic skills had generally increased as a result of becoming ill. Ensuring that physicians in training have access to adequate, appropriate, and confidential medical care will enhance the positive aspects of patienthood alluded to in this research.
Residents’ concerns, revolving around their simultaneous roles as trainee, employee, student, colleague, and supervisor, and sometimes patient, reflect the “web of relationships” posited by relationship-centered organizational theory (23). These relationships, which can be powerful and positive motivators, can also create a sense of vulnerability—one which has not been sufficiently elucidated with respect to resident health and well-being. When obtaining personal health care, most people feel a sense of increased vulnerability as their physical self and personal information are both potentially exposed. Residents are no different, but their role as resident-patient is unique.
Implications and Recommendations
Our findings have several implications for graduate medical education. Although greater attention is now being paid by medical educators to the well-being and health of trainees (24–26), more work is needed. First, residents’ privacy and confidentiality concerns about seeking care, particularly with regard to more stigmatizing conditions, suggest a need for increased efforts to inform residents about their health care options, including assistance programs (e.g., modeled after Employee Assistance Programs) (2, 26) which are available to all residents but not linked to the University, and for making available insurance plans which allow outside care. Residency training directors, who are often privy to residents’ personal health issues, must become sensitized to the concerns of residents and must be aware of policies governing their own disclosure of information. As the Accreditation Council for Graduate Medical Education does not specifically address privacy issues regarding resident health care in its policies, development of a consensus statement would be an important first step.
Second, the finding that residents feel ill-informed about their rights as patients and safeguards in place to protect them flags an unmet need for graduate medical education—to inform residents more completely about their rights and safeguards, to address concerns proactively, and to develop programs to ensure greater confidentiality and access, regardless of students’ insurance coverage or ability to pay (17, 26).
Finally, the common perception among residents that one’s colleagues will be unsympathetic to personal health needs (1) must be battled on several fronts—by promoting self-care through policies that clearly value residents’ well-being, by educating residents and faculty, not only about physician illness and impairment (27), but also about the wide-ranging benefits of healthy self-care habits, and by establishing policies and programs at the medical student level, to encourage confidential support and care when needed, and to enhance overall student well-being and attunement to self-care (25). Ensuring that trainees have the time, access, support, privacy, and confidentiality to obtain care for themselves is critical to a medical education mission in which each individual’s value is recognized and promoted.