Use of an Opt-Out vs Opt-In Strategy Increases Use of Residency Mental Health Services

Graduate medical education creates high-pressure environments that frequently lead to psychological stress for resident physicians. Given these conditions, researchers have demonstrated a substantial prevalence of workplace burnout, depression, and general psychological distress among residents.^1-7  Despite this, residents report a low rate of mental health services use. In one study, only 23% of residents with depression sought any treatment⁸  compared to 53% of those in the general population.⁹  Yet psychotherapy shows efficacy for many conditions¹⁰  and allows access to an individual who can evaluate, treat, and potentially protect residents from the most severe mental health outcomes, including suicide. The Accreditation Council for Graduate Medical Education (ACGME) formally requires institutions to provide access to mental health counseling¹¹  and to ensure that residents have the opportunity to attend these appointments even during working hours.¹²  These requirements are typically met via access to mental health clinicians coupled with instructions to residents such as, “If you need help, call this number to schedule an appointment.” While expeditious, this “opt-in” strategy has multiple issues that will hamper its efficacy.

Studies of resident help-seeking in this opt-in environment demonstrate low levels of utilization of mental health services.^8,13,14  Residents note confidentiality concerns, lack of time, credentialing and licensing repercussions, and stigma as some of the barriers to initiating contact.^8,15,16  Additionally, studies have demonstrated that when individuals develop greater levels of depression they process information differently.^17-19  These depression-induced changes in processing create biases in thinking that result in negative self-appraisals along with a more pessimistic estimation of the likelihood that help-seeking will actually lead to improvement.^20,21  Thus, paradoxically, the more depressed a resident becomes, the less likely they are to believe psychotherapy will be helpful, leading to less utilization of mental health services among those who may need them the most. Combined, these barriers to help-seeking create a milieu in which the standard institutional opt-in therapy strategy may fail.

Rather than taking an opt-in approach, an “opt-out” approach, where all residents are scheduled and then can call to cancel, may improve resident use of mental health resources. This approach of universal mental health screening has been used successfully in other high-stigma, high-stress professions.^22-24  Thus far, studies that have examined the role of opt-out approaches in residents have found some success.^25-27  Residents generally approved of opt-out approaches, felt they were potentially beneficial, and speculated that they might improve their use of mental health sessions. While helpful, these studies measured subjective outcomes or had methodological issues that leave open the question of whether opt-out approaches would indeed result in actual increased help-seeking among resident physicians. One recent exception included a behavioral outcome measure. Kevern and colleagues²⁸  introduced an opt-out program at a single institution and found that just over half of the residents who participated accepted the opportunity for a therapy session. One-third of those who did partake in therapy scheduled additional sessions. This study suggests that opt-out approaches not only may prove appealing to, and well received by, residents but also may increase the actual utilization of mental health services. The current study will complement and expand on the work of Kevern and colleagues²⁸  by adding a control group to allow additional evidence of the potential value of the opt-out approach.

Accordingly, we sought to determine if an opt-out approach, complemented by efforts to reduce stigma, would result in increased help seeking among residents, as determined by the actual use of psychotherapy sessions, compared to a control condition of a more traditional opt-in approach.

In academic year 2020-2021, Riverside Community Hospital, a major teaching hospital and clinical rotation site for the University of California Riverside School of Medicine in Southern California, developed a bundle of initiatives designed to reduce barriers to help seeking and minimize stigma for its 114 emergency medicine, family medicine, and internal medicine residents. This initiative provided the opportunity to undertake an observational study comparing the use of psychotherapy between the time frames before and after its implementation.

The condition before the initiative (the opt-in condition) occurred between July 1, 2020, and January 31, 2021. During the opt-in period, all residents were eligible to partake in confidential, free teletherapy sessions provided by therapists at a local clinical psychology training program. All participants received an email informing them of the availability of this service. Additional announcements of this service occurred during educational conferences, on computer screensavers, and on fliers placed in resident common areas. These communications included a telephone number that would allow residents to self-initiate a call to schedule an appointment. The initial design called for a 3-month opt-in condition, but due to logistical issues associated with the COVID-19 pandemic, as well as ensuring adequate funding and an adequate number of therapists to cover the potential sessions, the opt-in condition was extended to 7 months.

The period during the mental health initiative (the opt-out condition) occurred between February 1, 2021, and April 30, 2021. Programs notified residents about the mental health initiative via email and virtual meetings, during which residents were told that they would be booked into a session of teletherapy that they could then keep by simply logging onto a link sent to them by the clinic, or that they could cancel or reschedule by calling the psychological services clinic. Those who already were seeing a therapist at the clinic would not be booked but would continue with their previously scheduled sessions. Residents were informed that therapy participation was voluntary and that the residency program and sponsoring institution would not be notified as to whether any given resident did or did not participate in therapy sessions, or if a resident was already seeing a therapist at the clinic. Furthermore, the content of sessions was confidential, with the usual legal exceptions, and was never shared between the psychotherapists and the program or institution.

During this period, the residency programs provided the contact information of all emergency medicine, family medicine, and internal medicine residents to the same mental health clinic to allow clinic administrators to schedule therapy sessions. The clinic reviewed its internal records to determine which residents were currently already in therapy with the clinic, having self-initiated therapy, and removed them from the list of residents needing to be scheduled. The residency program coordinators provided the clinic options for preplanned time slots for the initial therapy sessions. These appointments corresponded to the scheduled residency didactic sessions (academic half days or noon conferences). Throughout the academic year of 2020-2021, all of the residency programs were conducting their educational sessions using computerized video platforms due to the COVID-19 pandemic. Residents were then notified by email of the date and time of their scheduled session along with the option to confirm the appointment, call and reschedule it, or call and cancel it. All residency programs allowed residents to miss a proportion of didactics without reporting a specific reason for why they were absent. Thus, those who attended therapy sessions and missed didactics were not required to report their participation in therapy. At the conclusion of the initial opt-out therapy session, residents could choose to schedule additional sessions. For those who did so, the scheduling was done via the typical procedures of the clinic with jointly acceptable dates and times determined by the resident and the clinic. During this phase, residents could continue to call and schedule therapy sessions themselves, should they choose to do so. The same communication strategies offering counseling sessions at the clinic that were present in the first time frame continued throughout the second. For both the opt-in and opt-out conditions, all therapy sessions were free of charge to the residents. The residency programs compensated the psychological services clinic an hourly rate for all therapy sessions during both conditions, which was covered by a grant from the state of California.

At the beginning of the mental health initiative, program directors notified residents that participation in psychotherapy was an elective part of their educational curriculum. This curriculum was designed to potentially aid them personally, but would also help their patients by improving their own understanding of psychotherapeutic interventions. Additionally, it could help their colleagues by allowing them to better describe the process of psychotherapy to other struggling residents who may harbor doubts about mental health care. Furthermore, characterizing their participation in psychotherapy as fulfilling a curricular component of their training, rather than as a response to a stigmatized mental health condition, could insulate them from the potential negative impact of participation in therapy on licensure or credentialing applications.

At the start of the mental health initiative, the therapists who would team with residents introduced themselves in a virtual session for each of the 3 specialties involved in the study. Each therapist gave their name, their approach to therapy, and a brief introduction at the start of a scheduled didactic session. This exposure was designed to reduce stigma through familiarity with the individuals who may provide the mental health services.

Our primary outcome consisted of the number of unique residents who self-initiated psychotherapy appointments in each of the different time frames. Following the completion of the mental health initiative, the clinic provided anonymized administrative records showing the number of unique residents who called the clinic during each time frame. In the pre-initiative time frame, the clinic noted the number of unique residents who self-initiated a call, scheduled a psychotherapy appointment, and participated in at least one psychotherapy session. In the mental health initiative time frame, the clinic recorded the number of unique residents who, after the completion of their automatically scheduled first session, asked to schedule at least one additional session. Our secondary outcomes were the number of residents who attended the initial prescheduled (opt-out) session during the mental health initiative, and the number of self-initiated therapy appointments the following academic year (2021-2022) after the conclusion of the mental health initiative.

Given that the research design used a sequential observation approach, a competing hypothesis exists for any effects noted in the 2 conditions. The need for therapy may vary as a function of the calendar months studied, rather than the mental health initiative. That is, residents may generally seek out therapy at a higher rate in the months of February through April (the time interval for the mental health initiative) as compared to July through January (the baseline time frame using an opt-in approach). To assess the likelihood of this alternative hypothesis, the same mental health clinic provided anonymized administrative data showing the number of new resident-initiated therapy sessions by calendar month for the academic year starting July 2021 and ending June 2022, the year immediately following our study. The institution returned to a standard opt-in approach to therapy for that academic year.

We compared the rates of self-initiated psychotherapy appointments between the 2 time periods with McNemar’s test and provided descriptive statistics to report participation in the opt-out therapy sessions. For the following academic year (2021-2022), the monthly rate of new resident-initiated therapy appointments in the July through January months was compared to the rate in February through April by a 2-tailed t test. Statistics were calculated by STATA version 15 (Stata Corp).

The institutional review board for the sponsoring institution approved this protocol.

There was a combined total of 114 residents in the internal medicine, emergency medicine, and family medicine residency programs when the initiative began. The Figure shows the distribution of residents throughout the 2 time frames. During the 7 months of the opt-in time frame (pre-mental health initiative) only 7 of 114 residents self-initiated therapy (6%). Thus, subtracting the 7 residents who sought therapy on their own, 107 residents were scheduled for an initial therapy session during the opt-out mental health initiative. No resident self-initiated a call to the clinic to begin therapy during the post-initiative time frame. Of the 107 residents who were scheduled into a therapy appointment, 59 (55%) kept the initial appointment and 48 (45%) did not. Of the 59 residents who kept their initial appointment 23 (39%) self-initiated at least one additional therapy session. Thus, in the traditional opt-in phase, 7 of 114 residents self-initiated therapy (6%), while an additional 23 out of 107 entered therapy during the intervention’s opt-in phase (22%) for a total of 30 residents opting in out of the total of 114 (26%) (P<.001 by McNemar’s test).

During the academic year following the mental health initiative, an average of 2.3 unique resident-initiated appointments occurred per month for the months of July through January, and 2 per month for the months of February through April, with a median of 4 total sessions per resident. There was no statistically significant difference in the rate of unique resident-initiated therapy between these 2 time periods (t (8)=0.25, P=.81). In the academic year following the study (when in-person sessions were an option), there were a total of 135 sessions, and residents opted for the telehealth structure in 111 of these sessions (82%).

Our study, which compared a traditional opt-in structure for mental health support to an opt-out structure, coupled with deliberate efforts to reduce stigma, found a 15% absolute (350% relative) increase in the number of residents utilizing psychotherapy sessions. This increase is even more substantial given that the opt-out time frame lasted less than half the duration of the opt-in time frame.

As noted earlier, Kevern and colleagues recently conducted an opt-out investigation.²⁸  Our study intervention closely resembled that of Kevern et al²⁸  in that both studies used an opt-out approach, utilized teletherapy, and allowed booked sessions during protected didactic time. The resulting utilization of therapy in this approach is surprisingly consistent with that study. In the Kevern et al study, 40% of residents opted out of therapy, while in the current study 45% opted out. Similarly, Kevern and colleagues found that 34% of residents who participated in the initial opt-out session scheduled additional therapy sessions, while we found a rate of 39%. Our study replicates these behavioral results and adds to the foundation of this prior work through the use of a control condition, which allows for greater certainty of the positive impact of institutions transitioning from a standard opt-in approach to an opt-out approach.

Although the current study and Kevern and colleagues’ work indicate the potential success of opt-out programs, such an approach would be problematic if opposed by the residents. Our study focused exclusively on behavioral outcomes, but there is evidence that residents are not opposed to the idea of opt-out approaches and believe they are generally helpful.^25-28  For example, in the recent Kevern et al opt-out study,²⁸  96% of those who participated in therapy, and completed the survey, felt it was worth their time and were satisfied with their appointment, 94% felt it demonstrated that the program cared about their well-being, and 98% recommended that the service be offered to future residents. Thus, residents who opted to participate in therapy through the opt-out program reported significantly positive appraisals of the experience. In addition to generally endorsing the opt-out therapy experience, residents reported it likely would result in actual increases in the use of therapy. In the Kevern et al study,²⁸  81% of those who participated in therapy speculated that the program would increase their willingness to participate in mental health services, which is the ultimate goal of such an intervention.

Regardless of whether residency programs use opt-in or opt-out approaches to scheduling, the issue of competing time demands and priorities may frustrate efforts to improve engagement in mental health services. Mental health appointments conflict with either didactic, clinical, or personal time. Given that lack of time is a primary barrier to seeking mental health services, we advocate for allowing residents to book their appointments during clinical or didactic time should they wish to do so. Given that the median number of sessions for any given resident was 4 over a 12-month period, the loss of didactic exposure or clinical experience is likely minimal, especially compared to the impact of untreated mental health concerns. For those specialties with ACGME-imposed minimum required attendance at didactics, we advocate for the ACGME to allow mental health services as an acceptable excuse for missing a didactics session. Of note, programs are required by the ACGME Common Program Requirements to allow residents to attend mental health appointments during working hours.¹² 

Although the initiative was centered around opt-out scheduling, several other components existed. For example, confidentiality concerns are cited by 57% of residents as a reason they do not attend therapy appointments.⁸  Resident concerns include worries that co-residents, faculty, and/or the program director may discover that they have undertaken therapy. The institution attempted to mitigate some of these concerns by creating a milieu in which all residents were presumed to be undertaking therapy as part of the opt-out approach. Thus, no given resident appeared exceptional when they undertook therapy.

Likewise, nearly two-thirds (62%) of residents cite a lack of convenient access as a barrier to psychotherapy.⁸  Traditional in-person therapy sessions require clients to leave their workplace, travel to an office building, sit in a waiting room (often with other clients who may recognize the physician), complete a session, exit through the same waiting room, and travel back to the workplace. This mental health initiative maximized convenience through the use of teletherapy sessions.

Additionally, using third-party therapists rather than in-house or employment-linked therapists (such as those obtained through an employee assistance program) may have increased resident confidence in confidentiality.¹⁶  Anecdotally, we found that residents were comfortable disclosing to others that they had a therapy session during the opt-out phase. For example, residents would openly state that they needed to leave didactics to attend their therapy sessions. This suggests that the mental health initiative was at least partially successful in reducing confidentiality and stigma barriers.

Our study has a number of limitations. The goal of the initiative was to maximize the use of therapy by residents. The institution was successful in this regard, but the multipronged approach using an opt-out strategy coupled with stigma reduction interventions does not allow for the investigation of individual elements. Moreover, although participants came from 3 different residency programs, they all were under a single sponsoring institution. Our study design did not randomize residents, but rather used an observational design. This allows for confounders such as the calendar months involved in each condition. The pre-initiative time frame occurred between July and January, while the post-initiative time frame occurred between February and April. This leaves the possibility that the effect on help seeking could be due to variables related to the time frames themselves rather than the opt-out or opt-in condition. However, our analysis of help-seeking incidence the year after the initiative did not support that alternative hypothesis, as rates of help-seeking did not differ by the sequential time frames used in our study. Additionally, we used a clinic staffed by doctoral students in clinical psychology supervised by licensed clinical psychologists. This approach allowed for a unified single clinic with adequate resources to see all interested residents. However, the use of clinical psychology students, rather than licensed psychologists, may influence our findings. Finally, our study was limited in the ability to determine the underlying motivational drivers that resulted in the observed increase in therapy. Residents may have opted to participate out of concern for their own mental health, to be a role model for other residents, to learn about therapy so they can better help their future patients, or because of some other factor. Future research may help determine what motivational drivers are best used to encourage resident mental health utilization.

In this study of 3 residency programs at a single institution, we found that a mental health initiative designed to reduce barriers to help-seeking, including a transition to an opt-out strategy for scheduling appointments, resulted in a substantial increase in the number of residents using psychotherapy sessions.

The authors would like to thank Kathy Haimson, Lorra Baldes, Jennifer Wilson, Carol Cahueque, Noel Cooper, the staff at Psychological Services of Riverside, and the Residents at Riverside Community Hospital.

Funding: The cost of the therapy sessions was supported by a grant from CalMedForce and the State of California.

Conflict of interest: The authors declare they have no competing interests.

This work was previously presented as a poster at the virtual Accreditation Council for Graduate Medical Education Annual Educational Conference, March 30-April 1, 2022; the Coalition for Physician Well-Being Joy and Wellness Summit, Park City, Utah, USA, July 10-12, 2023; the American Conference on Physician Health, October 11-13, 2023, Palm Springs, California, USA; and at the American Medical Association ChangeMedEd meeting, Chicago, Illinois, USA, September 27-29, 2023.

Disclaimer: This research was supported (in whole or in part) by HCA Healthcare and/or an HCA Healthcare affiliated entity. The views expressed in this publication represent those of the author(s) and do not necessarily represent the official views of HCA Healthcare or any of its affiliated entities. Jason T. Siegel, PhD, head of the HCA/CGU Physician Wellness Research Lab, receives compensation from HCA Healthcare for time spent working on well-being issues with HCA Healthcare Graduate Medical Education.