Impact of the COVID-19 Pandemic on Resident Physician Clinical and Educational Work Hours at a Large Sponsoring Institution

The COVID-19 pandemic has dramatically affected residency programs, with many program directors restructuring call and rotation schedules in early 2020. Restructuring resident staffing was thought to preserve the health care workforce and reduce COVID-19 transmission risk.¹  Therefore, many programs rotated residents between in-person work and reserve staff working from home.¹ 

Numerous editorials theorized a subsequent decrease in resident work hours, raising concerns about causal deficits in training.^2,3  Currently, the impact of changes to resident scheduling has been studied only through surveys of program directors and residents. When surveyed, three-quarters of surgical residents reported a pandemic-related decrease in work hours.^2,4–6  Research shows that residents and program directors perceive training has been impacted negatively by pandemic-induced schedule restructuring.^2,4–6  Similar perceptions were noted historically after the Accreditation Council for Graduate Medical Education (ACGME) work hour restrictions in 2011 were enacted, as residents reported decreased educational opportunities secondary to a decrease in hours worked, and faculty reported first-year attendings required more supervision after implementation of work hour restrictions.^7–10  This historical information endorses current concerns that work hour decreases secondary to pandemic-related changes to resident clinical schedule may negatively impact resident education.⁷ 

To better understand whether educational quality was affected by COVID-19-related residency schedule changes, we must first determine if a decrease in work hours occurred. However, to our knowledge, no publication has objectively measured the effect of COVID-19 on the resident experience. As one of the nation's oldest and largest graduate medical education sponsoring institutions, with more than 1800 trainees, Mayo Clinic is well-suited to measure the pandemic's impact on resident work hours. This study compares resident work hours across specialties at 3 integrated academic health centers in 2020 to the previous year, 2019, unaffected by COVID-19.

De-identified resident work hour records from the 3 Mayo Clinic campuses (Mayo Clinic in Arizona [MCAZ], Mayo Clinic in Florida [MCFL], and Mayo Clinic in Minnesota [MCMN]) were obtained from an institutional residency management system managed by the Mayo Clinic School of Graduate Medical Education from March to May for the pandemic (2020) and control (2019) periods. All sites are tertiary academic centers. MCAZ and MCFL serve major metropolitan areas (Phoenix and Jacksonville, respectively), while MCMN is in a smaller city (Rochester). Of note, none of the 3 hospitals studied were at or over capacity secondary to COVID-19 during the time periods studied.

Mean work hours were calculated per work period (supplementary data Figure 1), specialty, and site. Mean hours worked from home were calculated by site and year. Residents self-reported work hours in an online system, reporting hours worked from home separately. Hours worked from home included clinical (eg, telemedicine, patient calls, completing notes), educational (eg, conferences, online modules), and administrative work.

Student's t test was used to compare the control and pandemic periods. Work hours were compared within programs by site, as pandemic-related schedule changes differed between specialties and hospitals. Significance was determined with α<.05, and all hypothesis tests were 2-sided. To achieve an anticipated moderate effect size of d=0.5 with a power of 0.8 and 2-sided alpha error of 0.05, 64 residents per group was required.

This study was approved by an institutional COVID-19 Research Task Force and deemed exempt by the Mayo Clinic Institutional Review Board.

Work hours were recorded for 43 residency programs: 22 at MCMN, 11 at MCAZ, and 10 at MCFL. Work hours for 1118 residents in the control period (MCMN: 832, MCAZ: 139, MCFL: 147) and 1149 residents in the pandemic period (MCMN: 850, MCAZ: 148, and MCF: 151) were analyzed. Residents per program was similar during the control and pandemic periods (supplementary data Figure 2).

Governor executive order mandated people to stay-at-home in Arizona (March 30–May 15, 2020), Florida (April 1–May 1, 2020), and Minnesota (March 25–May 18, 2020). During this time, elective procedures were deferred and people limited interactions, excluding essential activities (eg, working as a resident physician).^11–13 

In April, 4 programs had a decrease in average weekly hours worked: anesthesiology, otolaryngology, general surgery, and urology (Figure 1, Table 1). General surgery residents had the largest reduction in hours (48 hours/week vs 63 hours/week, P<.001). Residents at MCAZ worked more hours from home in the pandemic compared to the control period in April and May (April: 5.2 hours/week vs 1 hour/week, P<.001; May: 2.4 hours/week vs 0.9 hour/week, P<.001; Table 2).

In April, 4 programs had a decrease in average weekly hours worked: dermatology, diagnostic radiology, internal medicine, and neurology (Figure 1, Table 3). Dermatology residents had the largest reduction in hours (40 hours/week vs 50 hours/week, P=.013). In March, a decrease in average weekly hours worked was seen in diagnostic radiology (46 hours/week vs 51 hours/week, P=.045) and family medicine (50 hours/week vs 56 hours/week, P=.003; Table 2). Residents at MCFL worked more hours from home in the pandemic compared to control period in March, April, and May (March: 1.7 hours/week vs 0.5 hour/week, P=.001; April: 6.5 hours/week vs 0.7 hour/week, P<.001; May: 3.4 hours/week vs 0.3 hour/week, P<.001; Table 2).

In March, decreased average weekly work hours were seen in neurological surgery (57 hours/week vs 60 hours/week, P=.032), neurology (53 hours/week vs 57 hours/week, P=.014), and urology (52 hours/week vs 54 hours/week, P=.007; Figure 1, Table 4). In April, 11 of 22 programs had reduced average weekly hours, including anesthesiology, diagnostic radiology, internal medicine, neurology, oral and maxillofacial surgery, ophthalmology, orthopedic surgery, physical medicine and rehabilitation, radiation oncology, general surgery, and urology (Figure 1). Anesthesia had the largest reduction in hours worked in April (31 hours/week vs 51 hours/week, P<.001). In May, decreased average weekly work hours were demonstrated in anesthesiology (46 hours/week vs 50 hours/week, P=.026) and orthopedic surgery (50 hours/week vs 58 hours/week, P=.024; Figure 1, Table 3). Residents at MCMN worked more hours from home in the pandemic compared to control period in March, April, and May (March: 2.4 hours/week vs 0.8 hour/week, P<.001; April: 10.2 hours/week vs 0.8 hour/week, P<.001; May: 3.4 hours/week vs 1 hour/week, P<.001; Table 2).

Our results demonstrate that residents worked fewer hours during the early pandemic, in months corresponding to pandemic-related government restrictions. We additionally demonstrated that specialties were affected by a decrease in work hours differently based on the hospital. Finally, we demonstrate that residents worked significantly more clinical hours from home in early pandemic months.

We demonstrate that residents in almost half of all residency programs across 3 hospitals worked fewer hours in April. While regions enacted governmental restrictions differently, April was the only full month impacted by restrictions in Arizona, Florida, and Minnesota. Minnesota's governmental stay-at-home restrictions lasted the longest, which may explain why a larger proportion of programs at MCMN were affected compared to MCAZ or MCFL. Further, the number of COVID-19 cases in the month of April at MCAZ, MCFL, and MCMN were fewer than reported in other areas of the United States (Figure 2).¹⁴  This may explain why we demonstrate no increase in hours worked in certain specialties expected to be impacted heavily by COVID-19, such as emergency medicine, anesthesiology, and internal medicine.

While we demonstrate that residents worked fewer hours in April 2020, not all specialties were impacted, and for those that were, it was not consistent across sites. We offer the following possible explanations. First, as reported in the literature, call and rotation restructuring was approached differently in different locations.^1,3  Mayo Clinic provided program flexibility in their approach to pandemic-related call and rotation restructuring. Each program, in collaboration with specialty leadership, was provided autonomy to restructure schedules to effectively balance patient care and educational activities. As such, the authors are aware of several methods for restructuring, including transitioning to a team-based approach from a preceptorship model; removing residents from off-service/campus rotations or electives; and assigning residents to designed on- or off-call groups that rotated daily, weekly, or biweekly. To the authors' knowledge, at these 3 Mayo Clinic campuses residents were not deployed to aid other departments. More studies are required to characterize the types of schedule restructuring that occurred most frequently and were most efficacious for resident learning.

Second, the transition to virtual work may have differed among specialties. In a descriptive article about psychiatry residents, some residents maintained a stable clinical workload during COVID-19 by quickly transitioning to telemedicine.¹⁵  Mayo Clinic residency programs with no decrease in hours worked may have rapidly pivoted to telemedicine, allowing residents to conduct visits virtually when unable to see patients face-to-face during the early pandemic.

Third, specialties may have experienced differing patient volumes based on their typical practice: inpatient, outpatient, or mixed. We noted that surgical residency programs at MCAZ and MCMN were impacted more than non-procedural specialties, likely secondary to mandated postponement of elective procedures. Further, a decrease in working hours was not demonstrated in mainly outpatient specialties, such as pediatrics, family medicine, and psychiatry.

We further hypothesized that residents would work more hours from home during the pandemic. We saw that, in early pandemic months, residents on all 3 campuses worked from 2 to 10 times more hours from home. While we feel this increase likely reflects residency programs enabling residents to complete clinical and educational work, such as conference attendance or telemedicine, from home, activities completed from home can vary significantly between residents and specialties, limiting interpretation of this increase in work hours. Regardless, several studies have described residency programs' transition to virtual educational conferences during the pandemic.^16,17  The quality of this virtual education remains unknown. Internal medicine and emergency medicine residents, when surveyed, showed a preference for in-person over virtual learning and reported decreased levels of engagement in online compared to in-person lectures.¹⁸  Further work evaluating the effectiveness of virtual education in the COVID-19 era, especially for procedural specialties, is needed.

As demonstrated with regulation of work hours, a decrease in clinical work hours can impact educational and clinical experiences.^7–9  Historically, regulated decreases in work hours were associated with the shifting of responsibilities to senior residents and concerns about graduating residents' competency.^7,10  Similar effects may be caused by this demonstrated decrease in work hours during the early pandemic. As hours worked as a resident may not equate to resident competency gained, future work should determine how hours worked correlates with the quantity and quality of educational experiences during this period.

While this study evaluates only one sponsoring institution, each campus is distinct. This study draws from residency programs of varying sizes located in the Midwest, Southwest, and Southeast United States. This study does not provide the perspective of residents working in hospitals at or over capacity secondary to COVID-19 and therefore may not be generalizable to all resident training experiences. Further, as work hours are self-reported by residents at Mayo Clinic sites, we are unable to confirm the accuracy of these hours. Finally, while the residency programs at Mayo Clinic are some of the largest in their respective fields, the individual size of some programs is relatively small and may be underpowered.

These results mark a step to quantify the impact of the pandemic on resident clinical experience and education. The demonstration of decreased work hours during the period impacted by governmental restrictions may guide program directors to supplement in-person clinical work with virtual educational experiences in future governmental shutdowns. Further, the varying impact on work hours across hospitals and specialties indicate a need for future studies that identify factors that allow residents to maintain work hours during a pandemic.

The COVID-19 pandemic decreased resident work hours in April 2020 across Mayo Clinic campuses in Arizona, Florida, and Minnesota. This impact on hours worked was distributed similarly across medical and surgical specialties. Residents worked more hours from home during the pandemic period, likely attenuating the anticipated decrease in hours worked.