Abstract
Purpose
In July 2012, in response to residents' concerns regarding the impact of the traditional 24-hour call system on their personal well-being and educational experience, the University of British Columbia Radiology residency program adopted a 12-hour night float system. This shift takes place in the context of increasing concerns, both across Canada and internationally, about resident well-being and the impact of prolonged duty hours on patient care.
Methods
An anonymous survey was distributed to all 25 postgraduate years 2-5 University of British Columbia radiology residents 12 months after the introduction of night float. This study sought to solicit residents' feedback about these changes and to identify potential future changes to optimize the call system.
Results
The response rate was 100%; 96% of residents were in favor of continuing with night float rather than the traditional call system; 72% of residents reported that their judgement was affected secondary to being on night float. Although most residents described varying degrees of impairment, the rate of acute discrepancies between resident preliminary and attending radiologist final reports decreased by more than half, from 2% to less than 1%.
Conclusions
The vast majority of our residents were in favor of maintaining the night float call system. Night float had a beneficial effect on the resident educational experience: by eliminating the pre-call morning and post-call day off rotation, residents gained an additional 24 days per year on other clinical rotations.
In July 2012, the University of British Columbia (UBC) radiology residency program adopted a 12-hour night float system at Vancouver General Hospital, the province's largest hospital and only level 1 trauma center. The impetus for this change came from resident concerns regarding the traditional 24-hour call system. In particular, the residents thought that the continuous adjustment to and from call affected both the amount and quality of time that they spent on rotation. This shift from 24- to 12-hour call was adopted in the context of increasing concerns, both across Canada and internationally, about resident well-being as well as the impact of prolonged duty hours on patient care [1–24].
Under the traditional 24-hour call system, the resident worked a regular 8 AM to 5 PM workday and was on call after 5 PM until 8 AM the following morning. On average, the call frequency was approximately 1 in 5. Under the night float system, the resident works 12-hour shifts from 8 PM to 8 AM for 7 consecutive days, twice each year. Seven consecutive days were chosen to allow sufficient time for the resident to adjust to overnight work and to provide continuous after-hours coverage given the number of available residents. During this week, the resident works exclusively on night float and is excused for his or her regular rotation duties.
On-call responsibilities include protocoling and providing preliminary reports for all urgent imaging studies performed for both the patients in the emergency department and inpatients. The majority of studies are performed by using computed tomography, and common study types include neuroimaging, body imaging, and trauma radiology. In addition, residents are responsible for performing and reporting urgent ultrasounds, which commonly include pelvic, scrotal, abdominal, and deep vein thrombus studies. Plain films are reviewed at the request of the ordering physician. In general, approximately 20-40 studies are performed during a typical 12-hour night float shift. At the time that the survey was conducted, the attending radiologists were in-house, working alongside the on call resident from 5 PM to 8 PM and were available by pager thereafter. Clinical fellows also were available for consultation overnight.
Materials and Methods
This study sought to solicit residents' feedback about the changes as well as to identify potential future changes to optimize the call system. The study population consisted of radiology residents at UBC. An anonymous questionnaire was distributed to all 25 postgraduate year (PGY) 2 to PGY-5 residents. Only the PGY-3 to PGY-5 residents had experience with the 24-hour call system at that hospital site from which to compare. The study was carried out 12 months after the introduction of night float to assess its impact on residents' experience and impacts on their well-being.
The survey consisted of a total of 18 closed and open-ended questions administered by sending an e-mail as a Word document (Microsoft Corp, Redmond, WA). The inclusion criteria included all PGY-2 to PGY-4 residents who had completed 2 weeks of night float and all PGY-5 residents who had completed 1 week of night float. PGY-5 residents do night float for only 1 week because they do not cover call in the last half of their year. The questions were designed to obtain information about the residents' quality of life, quality of educational experience, and perception of factors that influence medical errors. Residents also were asked to input recommendations for changes to optimize the call system in the future. The survey drew from surveys already in the published literature and also included items of interest to the authors [25]– [27]. Data analysis consisted of a simple tabulation of responses.
Results
A total of 25 of 25 eligible residents completed the survey (response rate of 100%). On average, it took residents 2.8 days to acclimatize to night float, and 3.9 days to return to their daily routine after completion of the week. Thirty-six percent of the residents were able to sleep during their night float shifts; on average, they slept just over 1 hour. Eighty percent of the residents reported becoming fatigued at some point during the shift, most commonly during the early morning hours (4 AM to 6 AM). Approximately three-fourths of the residents agreed that they became fatigued and/or overwhelmed during their week of night float, most commonly at day 5 or 6.
Seventy-two percent of the residents reported that their judgement was affected secondary to being on night float. Most described varying degrees of impairment. For example, residents reported decreased sensitivity to imaging findings, spending more time than they normally would on a given study, as well as an overall lack of ability to concentrate as the shift and week progressed. However, a significant minority, 28%, thought that their skills had improved. For example, some residents thought that, due to the overall shorter shifts, they were more alert during the time they would typically “crash,” during the early morning hours.
To explore the impact that this perception had on call performance, we examined the discrepancy rate between resident preliminary reports and attending radiologist final reports. A total of 38 preliminary resident reports were marked as discrepant by staff radiologists, of which 19 pertained to computed tomography (CT) of the abdomen-pelvis, 14 to neuroimaging (CT of the head and CT angiograms), 3 to CT of the cervical spine and 2 to chest imaging. Compared with a study undertaken at our institution under the traditional 24-hour call system [28], the rate of clinically relevant acute discrepancies decreased by more than half, from 2% to less than 1% (38 discrepant reports of a total of 4,190 studies [0.91%]). Ninety-six percent of the residents were in favor of continuing with night float rather than the traditional call system. Regarding the optimal structure of night float, the residents thought that the ideal numbers of consecutive hours were 8-10, with a range of responses from 5-15 hours. The ideal numbers of consecutive days were 5-7, with a range of responses from 5-14 days.
Discussion
This study confirmed that of the vast majority of UBC radiology residents, 96%, were in favor of maintaining the night float call system. Indeed, its adoption has had a beneficial effect on the resident educational experience. Under the traditional call system, residents lost 1.5 days on rotation for every shift of weekday call. By eliminating the pre-call morning and post-call day, night float enables residents to be present more days during a given rotation; as a result, an additional 24 days per year on rotation were gained. Moreover, early feedback from residents suggests that the transition to and from consecutive night floats shifts is preferable to intermittent overnight shifts once or twice each week under the traditional system. By grouping call shifts together, night float enables greater flexibility in scheduling, which allows residents to pursue rotations outside the Vancouver-based hospitals.
To our knowledge, there is only 1 other study in the radiology literature that examined the impact of duty-hour restrictions on resident well-being [11]. In a cross-sectional survey of 20 radiology residents at an academic medical center, it took an average of 2.0 days to acclimate to night float and 2.3 days to return to a normal routine after completing their 6 consecutive 11-hour night shifts [25]. Our residents took longer to adjust to and from the 7 consecutive days of 12-hour overnight shifts, and reported an average of 2.8 days and 3.9 days, respectively. Both the longer shift length and number of consecutive shifts may account for some of these reported differences.
Although three-fourths of the residents subjectively perceived that their judgement was affected secondary to night float, the rate of radiology resident–attending radiologist discrepancy rate had, in fact, decreased compared with the traditional 24-hour call system. Similar findings are borne out in the literature [29–32]. Clinically significant resident–attending radiologist discrepancies were compared between different call systems simultaneously employed at a single academic medical center [29]. The night float system, which consisted of 7 consecutive days of 9-hour overnight call, had a significantly lower discrepancy rate than 24-hour call (P < .01) [29]. Another study of resident–attending radiologist discrepancies at a single academic medical center demonstrated a statistically significant increase in the rate of major discrepancies during the last 2 hours of 12-hour consecutive overnight shifts [30]. Similarly, a retrospective review of resident–attending radiologist discrepancies demonstrated that the rate of both major discrepancies increased markedly over the course of an after-hours shift [31]. Although these studies did not examine the impact of discrepancies on patient outcomes, these results may reflect the culmination of factors, including fatigue and circadian desynchronization, that ultimately impact the cognitive performance of radiology residents.
Survey responses have already provided the basis for change to optimize the structure of night float. For example, when considering that it takes an average of 3.9 days for residents to return to their normal routine after the completion of night float, the start and end days were shifted from a Sunday to Saturday schedule to a Friday to Thursday schedule. This change enables residents to recover over the weekend before resuming their rotations the following Monday. These changes to overnight radiology on call coverage are taking place against a broader backdrop of discussion and legislation about resident well-being and resident duty hours.
In Canada, changes to resident duty hours began in the province of Quebec; effective July 1, 2012, Quebec residents were restricted to a maximum of 16 hours of consecutive in-house call [1]– [3]. Although similar restrictions have not been adopted in other provinces, this topic is attracting attention among medical communities as well the media [4],[5]. In 2012, the Royal College of Physicians and Surgeons of Canada, together with stakeholder groups, created a national steering committee on resident duty hours to explore this issue [4]. Although the steering committee stopped short of recommending national restrictions on the number of consecutive hours, it is conceivable that this issue may surface again. To this end, we believe that the changes adopted at our institution may be of interest to other programs that are considering similar changes.
There is a longer history of duty-hour restrictions in the United States, owing initially to concerns about patient safety. Since 1989, residents in New York State were limited to 24 consecutive hours of in-house call [6]. In 2003, similar restrictions were mandated for all residents nation-wide [7]. Research that demonstrates the adverse effect of prolonged working hours on both patient and resident safety has since led to the introduction of further duty-hour restrictions [8]– [10]. Effective July 2011, all PGY-1 residents are restricted to working no more than 16 consecutive hours.
As a result of duty-hour restrictions, many residency programs made changes to the structure of their call systems, including instituting a night float system. A survey of chief residents in accredited radiology programs in the year after the Accreditation Council for Graduate Medical Education changes found that 85% responded that the new requirements had improved their call experience and that 90% responded that their educational experience had been enhanced [21]. Chief residents reported an improvement of resident morale and improvement in quality of patient care and learning [21]. However, more frequent call and an increased work load among senior residents also was reported [21].
The impact of these duty-hour restrictions on the quality of health care and patient safety continues to be debated [11]. There is a well-established body of evidence that prolonged working hours are associated with deterioration of attention, cognitive function, and motor skills across a variety of medical specialties [12–16]. However, studies after the 2003 duty-hour restrictions by using national data samples did not demonstrate improvement in patient safety and mortality, nor did they seem to increase residents' hours of sleep or reduce fatigue [17–20]. As duty-hour restrictions continue to evolve, more rigorous evaluation of the impact of the changes will clearly be needed.
This study has several limitations. This was primarily a subjective study, with no quantitative verification of residents' responses. The target population was small, limited to 25 radiology residents from a single institution. Although our PGY-2 residents comprise one-fourth of the target population, they do not have experience of the traditional call system with which to compare night float. Finally, the study was limited to the resident experience and did not include the input of staff radiologists, fellows, and referring physicians; metrics about the impact on patient care were not within scope of the study.
Conclusion
The UBC radiology residents were overwhelmingly in favor of keeping the night float call system. Compared with the traditional 24-hour call system, the 12-hour night float call system benefits the resident educational experience and quality of life. By grouping call together, residents were able to spend more days on rotation with greater flexibility in scheduling rotations outside Vancouver-based hospitals. Moreover, this change was accompanied by a significant decrease in the discrepancy rate between resident preliminary and attending radiologist final reports. In the context of increasing concerns about resident well-being as well as the impact of prolonged duty hours on patient care, we believe that this topic may be of relevance to other residency programs that are considering similar changes. Finally, this topic is of crucial importance for practicing physicians [22]– [24]. There is a complex interplay between physician performance and sleep deprivation, which carries implications for the well-being and safety of both patients and physicians.