Implementing Night Float Systems for Nephrology Fellows: A National Survey of Nephrology Training Program Directors Authors: James J. Rutowski, Christina M. Yuan, Fahad Saeed, Daniel W. Ross, Catherine A. Moore Published: 2026 ABSTRACT Rationale & ObjectiveNight float (NF) systems are widely adopted by residency programs to support trainee wellness and duty hour compliance, yet the extent of NF use in nephrology fellowship programs has not been well characterized. This study examined how NF is implemented in nephrology fellowship programs and how program directors perceive its advantages and disadvantages.Study DesignCross-sectional national survey.Setting& Participants: Training program directors from all Accreditation Council for Graduate Medical Education--accredited adult nephrology fellowship programs.Exposure(s) or Predictor(s)Presence or absence of an NF system.Outcome(s)NF structure, trainee responsibilities, continuity clinic and didactic impact, and program directors’ perceived advantages and disadvantages.Analytical ApproachDescriptive statistics, group comparisons, and thematic analysis of open-ended responses.Results77 programs responded. 28 reported NF use. Programs with NF had more fellows and covered more hospitals at night. Most NF programs suspended continuity clinic and didactic attendance during NF rotations. Program directors most often perceived improved trainee wellness and satisfaction as advantages, whereas reduced continuity clinic and didactic attendance were common disadvantages.LimitationsSurvey was anonymous and nonvalidated, without program-level demographics, duration of NF use, overnight clinical volume, or trainee outcomes.ConclusionsNF adoption varied across nephrology fellowship programs. Program directors not only perceived important wellness benefits from NF but also noted educational challenges, especially regarding continuity clinics and missed didactics. The Accreditation Council for Graduate Medical Education (ACGME) defines night float (NF) as “the episodic coverage of patients just at night.”1,2 NF systems are used in internal medicine residency programs to help address duty hour restrictions, support wellness initiatives, and combat fatigue.3, 4, 5, 6 Different internal medicine subspecialties have implemented NF in various capacities to meet ACGME requirements, optimize patient care, and allow appropriate supervision and education of trainees.7, 8, 9, 10, 11, 12, 13 Similarly to internal medicine, cardiology, and critical care programs, some nephrology programs have implemented NF with reported success and as a tool to address fellow wellness.4,8,13,14 NF offers several potential benefits. In internal medicine residency programs and cardiology fellowships, NF has been associated with improved trainee work-life balance and more stable sleep patterns.4,15,16 Residency programs have adopted NF to meet duty hour requirements, reduce fatigue, and maintain continuity of patient care.17,18 Across both surgical and medical specialties, NF has been reported to improve night-to-night continuity and trainee autonomy without adversely affecting patient outcomes.10,18, 19, 20, 21, 22 More recently, some fellowship programs, including nephrology, have implemented NF to support trainee well-being and mitigate burnout.8,23 Ongoing concerns regarding workforce sustainability and persistent recruitment challenges in nephrology further emphasize the importance of training environments that promote wellness and satisfaction, as highlighted by Agrawal et al.24 NF may be particularly relevant in nephrology fellowships, in which fellows frequently manage urgent, high-acuity issues, such as initiation of renal replacement therapy and consultation for critically ill patients, interventions linked to improved patient outcomes.25,26 Despite these potential benefits, limited data exist on how nephrology fellowship programs implement NF or its impact on educational structure and operations.1,2,8,25,27 Our study characterizes the current use of NF within US nephrology training programs. In a nationwide survey of training program directors (TPD), we sought to define the extent of NF implementation, clarify call responsibilities, assess impact on trainee education, and evaluate TPD perceptions of its advantages and disadvantages. Methods This exploratory survey was developed specifically for nephrology program directors (n = 151), as a tool to query the use of NF in their nephrology fellowship programs, about which little is known. It was developed by 2 program directors (CAM and DWR; one without NF and one with NF) and a nephrology fellow (JJR). All survey creators had designed and overseen call schedules for nephrology programs. Questions were iteratively developed based on the results of and comments from an ad hoc, informational preliminary survey of program directors with 6 or fewer fellows (conducted in January 2025), literature review, and investigator experiences with NF. The survey questions were reviewed, edited, and programmed into the survey platform by a former program director (CMY) and the completed survey was tested by the other investigators before dissemination. Because of the small number of potential respondents, pilot testing of the completed survey was not done. No commercial or industry input influenced survey content. The complete survey instrument is available in Table S1. The survey collected data on program demographics (US census region and number of ACGME-approved fellows), NF structure and responsibilities, perceived disadvantages to implementation, educational content associated with NF rotations, and program directors’ perceptions of the impact on trainee satisfaction. To further maintain anonymity and reduce the risk of inadvertently identifying respondents, we did not collect individual program director demographics such as years in the role, academic rank, board pass rates, or in-training exam results. The survey did not question whether programs were based in academic medical centers or community settings, as this had the potential to identify respondents, given the small number of nephrology fellowship programs nationally. The survey did not collect information on overnight clinical volume, including the number of consults or cross-cover responsibilities per night, which may influence the need for NF systems more directly than the number of hospitals covered. Questions addressing the advantages and disadvantages of NF elicited program directors’ subjective perceptions rather than objective measures of trainee performance, wellness, or educational outcomes. The survey also did not collect information on the duration of NF implementation or the methods programs used to evaluate its perceived effects. Investigators distributed the survey via email to all program directors of ACGME-approved adult nephrology fellowship programs in the United States. Program director contact information was obtained from the publicly available ACGME list of accredited programs.28 The survey was administered using SurveyMonkey.com, a secure, commercially available web-based platform, with IP address blocking enabled to ensure anonymity. The survey remained open for 8 weeks (March 21, 2025, to May 17, 2025). Three reminder emails were sent as follows: 2 approximately 2 weeks apart during the 8-week survey period, and a final reminder the day before the survey closed. Program directors who requested to be removed from further correspondence were excluded from subsequent reminder emails. Owing to the anonymous nature of survey responses, submitted entries could not be withdrawn. The survey and study protocol were deemed exempt by the University of Rochester Institutional Review Board and were conducted in accordance with the principles outlined in the Declaration of Helsinki. Participation was voluntary. Consent was implied by partial or complete survey completion. Statistical Analysis Responses were tabulated and summarized using percentages, means (standard deviation), and medians (interquartile range) as appropriate. Incomplete surveys (n = 1) were included in data analysis. The response rate was calculated as completed (n = 77) and partially completed surveys (n = 1) divided by the number surveyed. The completion rate was calculated as the number of surveys entered and completed divided by the number of completed and partially completed surveys. The mean number of ACGME-approved fellows (t test) and the US census area (χ2) of programs were compared between respondents versus the surveyed population to assess nonresponse bias. Other comparisons were made using the 2-tailed t test (P < 0.05 was considered significant). Qualitative Analysis Question numbers 16 and 23 (Table S1) included open-ended responses. We deployed a thematic analysis to examine the qualitive data. Three investigators independently coded the data to identify emerging categories and themes. The investigators then met twice to discuss their findings, reflect on the data, and reach consensus on the codes. Based on these discussions, an agreement was reached to present the qualitative responses as “advantages” and “disadvantages” of NF.29 Responses were further quantified by counting the frequency of themes, finally organizing the data into the following 2 sections: (1) advantages of NF and (2) disadvantages of NF. Results Survey Response Of the 152 ACGME-accredited nephrology fellowship programs, 1 was excluded owing to invalid contact information and absence of a program website. Of the remaining 151 program directors, 77 (51%) responded to the survey, and survey completion was 96% (74/77). Emails to all 151 programs were validated and an active email was documented for every program. Geographic representation did not differ significantly from national distribution (χ2 test: P = 0.75) (Table S2). Nationally, the mean number of ACGME-approved fellowship positions per program was 7.0 (standard deviation 3.3), compared with 7.5 (standard deviation 3.6) among respondents (Table 1). This difference was not statistically significant (2-tailed t test: P = 0.29; 95% confidence interval, -0.43 to 1.44). The location of the responding programs and their size were not significantly different from those nationally (Table S2).Table 1Characteristics of Training Programs With and Without Night FloatWith Night FloatWithout Night FloatNo. of programs2849Mean no. of ACGME-approved fellowsa10.3 (SD 3.4)(n = 28)6.2 (SD 2.7)(n = 49)Mean no. of hospitals covered while on night call2.2 (SD 1.1)(n = 27)1.5 (SD 0.6)(n = 47)1 hospital covered overnight9/27 (33%)25/47 (53%)2 hospitals covered overnight8/27 (30%)18/47 (38%)3 or more hospitals covered overnight10/27 (37%)4/47 (9%)No. of weeks of night call annually per fellowb7.3 (SD 3.7)(n = 26)9.1 (SD 3.3)(n = 35)Practitioners providing night call coverage Fellows with faculty supervision28/28 (100%)47/49 (96%) Faculty working independently3/28 (11%)10/49 (20%) Nonfaculty nephrologist0/28 (0%)0/49 (0%) Advanced practice providers with faculty supervision1/28 (4%)1/49 (2%) Other0/28 (0%)2/49 (4%)Abbreviations: ACGME, Accreditation Council for Graduate Medical Education; SD, standard deviation.aP < 0.0001; 2-tailed t test.bEighteen programs “without night float” did not have night call assigned on a weekly basis, 10 of which did not supply sufficient information to calculate the numbers of weeks on call per fellow. Nine programs “with night float” had hybrid call schemas, combining night float and traditional night call. Two-tailed t test: P = 0.05. Table 1Characteristics of Training Programs With and Without Night FloatWith Night FloatWithout Night FloatNo. of programs2849Mean no. of ACGME-approved fellowsa10.3 (SD 3.4)(n = 28)6.2 (SD 2.7)(n = 49)Mean no. of hospitals covered while on night call2.2 (SD 1.1)(n = 27)1.5 (SD 0.6)(n = 47)1 hospital covered overnight9/27 (33%)25/47 (53%)2 hospitals covered overnight8/27 (30%)18/47 (38%)3 or more hospitals covered overnight10/27 (37%)4/47 (9%)No. of weeks of night call annually per fellowb7.3 (SD 3.7)(n = 26)9.1 (SD 3.3)(n = 35)Practitioners providing night call coverage Fellows with faculty supervision28/28 (100%)47/49 (96%) Faculty working independently3/28 (11%)10/49 (20%) Nonfaculty nephrologist0/28 (0%)0/49 (0%) Advanced practice providers with faculty supervision1/28 (4%)1/49 (2%) Other0/28 (0%)2/49 (4%)Abbreviations: ACGME, Accreditation Council for Graduate Medical Education; SD, standard deviation.aP < 0.0001; 2-tailed t test.bEighteen programs “without night float” did not have night call assigned on a weekly basis, 10 of which did not supply sufficient information to calculate the numbers of weeks on call per fellow. Nine programs “with night float” had hybrid call schemas, combining night float and traditional night call. Two-tailed t test: P = 0.05. Characteristics of Training Programs With and Without Night Float Abbreviations: ACGME, Accreditation Council for Graduate Medical Education; SD, standard deviation. P < 0.0001; 2-tailed t test. Eighteen programs “without night float” did not have night call assigned on a weekly basis, 10 of which did not supply sufficient information to calculate the numbers of weeks on call per fellow. Nine programs “with night float” had hybrid call schemas, combining night float and traditional night call. Two-tailed t test: P = 0.05. Current NF Practices Of the 77 responding programs, 28 (36%) reported having NF and 49 (64%) did not (Table 1). Programs with NF had significantly more fellows (mean, 10.3 vs 6.2; P < 0.001) and covered more hospitals at night (mean, 2.2 vs 1.5; P < 0.001) than those without NF (Table 1). Among the programs with NF, coverage involved 1 hospital in 33% (n = 9), 2 hospitals in 29% (n = 8), and 3 or more hospitals in 37% (n = 10). Fifteen percent (n = 4) reported nights covered without fellows, either by faculty alone or by advanced practice providers with faculty supervision (Table 1). By comparison, 27% (n = 13/49) of programs without NF reported nights without fellow coverage, using faculty or advanced practice providers (Table 1). One-third (33%; n = 9) of the programs with NF used a hybrid model combining NF with traditional home call, which was most often taken from home rather than in the hospital (Table 1). Fellows in programs with NF averaged 7.3 weeks of night call annually, including 5.5 weeks of NF and 1.7 weeks of traditional night call (Table 1). Most programs (78%; n = 21) used home call, whereas 19% (n = 5) required in-hospital call (Table 1). In all, 96% (26/27) of NF programs did not schedule fellow continuity clinics during NF, whereas 93% (25/27) did not require fellow attendance at didactic sessions (Table 1). To meet educational requirements, 52% (14/27) of programs with NF recorded lectures and used asynchronous access; 33% (9/27) implemented a repeating curriculum throughout the fellowship, allowing material to be encountered multiple times (Table 1). Characteristics of Programs That Have Not Adopted NF Programs without NF relied primarily on traditional home call schedules. In programs with interpretable schedules (n = 35/48), fellows covered an average of 9.1 weeks of night call annually (Table 1). Four programs were excluded owing to uninterpretable responses. Call schedules were structured by days, weekends, or rotation blocks in 35% (n = 17/48) of programs (Table 1). Programs without NF provided coverage for 1 (55%; n = 26/48) hospital, followed by 2 (36%; n = 17/48) hospitals or 3 or more (8%; n = 4/48) hospitals (Table 1). Attitudes Surrounding NF Program director perceptions of advantages and disadvantages reflected subjective impressions rather than formal evaluations, because the survey did not request objective measures or assessment methods. Among the 27 TPD respondents with NF, the most cited advantages were improved trainee wellness (96%) and improved trainee satisfaction (92%) (Table 2). Enhanced recruitment was reported commonly as an advantage (67%) with NF. The most reported disadvantages were a decline in didactic attendance (48%) and reduced continuity clinic attendance (48%) (Table 2).Table 2Advantages and Disadvantages of Night Float as Perceived by Program Directors With Night FloatAdvantagePercentageDisadvantagePercentageReduced work hour violations89 (24/27)Fellows develop a shift-work mentality41 (11/27)Improved trainee wellness96 (26/27)Didactic attendance declines48 (13/27)Improved trainee satisfaction93 (25/27)Outpatient continuity clinic experience is reduced48 (13/27)Improved trainee ITE scores11 (3/27)Negative impact on time for scholarly work4 (1/27)Improved trainee recruitment74 (20/27)Negative impact on time for elective rotations19 (5/27)Other7 (2/27)Other26 (7/27)Abbreviation: ITE, in-training examination. Table 2Advantages and Disadvantages of Night Float as Perceived by Program Directors With Night FloatAdvantagePercentageDisadvantagePercentageReduced work hour violations89 (24/27)Fellows develop a shift-work mentality41 (11/27)Improved trainee wellness96 (26/27)Didactic attendance declines48 (13/27)Improved trainee satisfaction93 (25/27)Outpatient continuity clinic experience is reduced48 (13/27)Improved trainee ITE scores11 (3/27)Negative impact on time for scholarly work4 (1/27)Improved trainee recruitment74 (20/27)Negative impact on time for elective rotations19 (5/27)Other7 (2/27)Other26 (7/27)Abbreviation: ITE, in-training examination. Advantages and Disadvantages of Night Float as Perceived by Program Directors With Night Float Abbreviation: ITE, in-training examination. Among the 47 TPD respondents without NF, 32% (15/47) reported that, to their knowledge, fellows had asked for NF. The most frequently reported reasons for not adopting NF were insufficient fellow numbers (80%), concerns about didactic attendance (57%), and disruptions to outpatient continuity (48%) (Table 3).Table 3Training Program Director Reasons for Having Not Adopted Night FloatReason for Having Not Adopted Night FloatPercentage“Too few fellows to allow it to work properly.”81 (38/47)“I feel didactic attendance will suffer.”57 (27/47)“I feel outpatient continuity experience will be degraded.”49 (23/47)“I feel there will be a negative impact on scholarly work.”30 (14/47)“I feel there will be a negative impact on elective rotations.”36 (17/47)Other28 (13/47)aaFour out of 13 commented that they felt night float would leave fellows ill-prepared for real-world clinical practice, which requires traditional home call. Table 3Training Program Director Reasons for Having Not Adopted Night FloatReason for Having Not Adopted Night FloatPercentage“Too few fellows to allow it to work properly.”81 (38/47)“I feel didactic attendance will suffer.”57 (27/47)“I feel outpatient continuity experience will be degraded.”49 (23/47)“I feel there will be a negative impact on scholarly work.”30 (14/47)“I feel there will be a negative impact on elective rotations.”36 (17/47)Other28 (13/47)aaFour out of 13 commented that they felt night float would leave fellows ill-prepared for real-world clinical practice, which requires traditional home call. Training Program Director Reasons for Having Not Adopted Night Float Four out of 13 commented that they felt night float would leave fellows ill-prepared for real-world clinical practice, which requires traditional home call. Thematic Analysis In open-ended responses, TPD identified several advantages (n = 8) and disadvantages (n = 19) of NF. Of the 27 programs with NF, 7 programs provided qualitative comments. Of the 49 programs without NF, 13 provided qualitative comments. The most frequently mentioned disadvantages were that NF does not reflect real-world practice (n = 7) and that there is a perceived lack of need for NF among fellows (n = 9), owing to an inadequate patient census (n = 2). Less commonly cited reasons included concerns about administrative burden (n = 1), reduced opportunities to provide feedback (n = 1), and the impact of the NF on sleep cycle (n = 1). Three TPD with existing NF discussed advantages of NF implementation at their institutions, including reduced fellow fatigue/improved wellness (n = 3), enhanced daytime continuity of clinical care (n = 1), and promotion of fellow autonomy (n = 1). TPD reported that NF was associated with improved trainee wellness, satisfaction, and recruitment. These positive representative comments from TPD are quoted in Table 4.Table 4Representative Positive Comments of Training Program Directors With Night FloatThemeRepresentative QuoteImproved wellness and morale“It has been a game-changer in terms of fellows’ wellness, morale, and fatigue.”Increased autonomy“Improved autonomy. The fellows enjoy it a lot.”Burden of 24-h call“Our fellows get called very often. It was not actually reasonable to do 24-hour call anymore…” Table 4Representative Positive Comments of Training Program Directors With Night FloatThemeRepresentative QuoteImproved wellness and morale“It has been a game-changer in terms of fellows’ wellness, morale, and fatigue.”Increased autonomy“Improved autonomy. The fellows enjoy it a lot.”Burden of 24-h call“Our fellows get called very often. It was not actually reasonable to do 24-hour call anymore…” Representative Positive Comments of Training Program Directors With Night Float Discussion This national survey of ACGME-accredited nephrology TPD provides insights into current practices of NF implementation. We identified 3 key findings. First, programs with NF differed from those without NF in the size and number of hospitals covered, suggesting that program resources strongly influence NF adoption. Second, NF implementation had significant impacts on continuity clinic and didactic attendance. Third, TPD expressed concerns about whether NF adequately prepares trainees for independent practice. Importantly, reported advantages and disadvantages reflect program directors’ perceptions rather than objectively measured outcomes. As such, findings should be interpreted as perceived effects of NF rather than verified impacts. First, programs with NF were significantly larger and covered more hospitals at night than programs without NF. Larger programs may be able to distribute workload and maintain clinical and educational continuity. This aligns with neurosurgical programs in which larger programs were more likely to adopt NF.17,19 Sufficient program size and personnel may be prerequisites for NF adoption. TPD without NF frequently cited the lack of sufficient fellows as the main reason for nonadoption. Internal medicine programs are much larger, with more residents to share overnight responsibilities.30 Cardiology typically has larger trainee numbers and an additional year of training, making NF more feasible.4,10,16 This highlights the following paradox: NF is a wellness initiative that is most easily implemented in larger programs, potentially leaving smaller programs at a disadvantage. TPD comments suggest NF is a perceived wellness initiative and recruitment advantage. This aligns with reports from surgical residency programs, in which residents strongly support NF rotations, burn out is reduced, there are no differences in the average number of patient encounters, and there is more time for nonclinical activities.21,31 Within internal medicine subspecialities, outcomes are less well studied. Available data indicate that trainees in NF programs report improved satisfaction and wellness.10,32 These findings are consistent with those by Agrawal et al24 and the workforce commentary from the American Society of Nephrology highlighting the need for enhanced fellow wellness support within training programs.14 Second, NF implementation affected educational activities. Nearly all programs with NF suspended continuity clinic and daytime didactics during NF rotations, raising concerns about compliance with ACGME requirements, which specify that outpatient continuity clinics should not be interrupted for more than 1 month.33 Among programs with NF, only 3% maintained the continuity clinic during NF rotations. Programs must ensure compliance with ACGME requirements and implement strategies to transition outpatient care without compromising continuity or patient outcomes. Internal medicine residencies have faced similar challenges and commonly adopted X + Y scheduling models that separate inpatient and ambulatory experiences, an approach associated with improved learning, reduced fragmentation, and enhanced resident satisfaction.34,35 Larger nephrology programs using NF may consider similar strategies to meet ACGME requirements.36 Didactic attendance was disrupted in 93% of NF programs and was most often mitigated through asynchronous or repeated sessions, although the educational impact of these adaptations remains unknown. Third, 25% of TPD who commented on disadvantages noted that NF does not reflect postgraduate nephrology practice, which still relies on home call, raising concern that fellows may be underprepared for independent practice. Conversely, improving sleep may enhance learning, supporting the educational value of NF even if it differs from posttraining practice.37 Existing retrospective and attitudinal studies evaluating the educational impact of duty hour restrictions show mixed results, and the current literature does not clearly define the working practices of nephrologists or whether those practices are optimal or safe.38, 39, 40 Together, these uncertainties raise broader questions about the appropriate duty hour expectations for independent physicians.19,21,41 This study has several strengths. The geographic distribution of responding programs and their program size did not differ from those of programs nationally, supporting the representativeness of our sample. It also is the first exploratory survey to analyze perceived benefits of program directions of NF implementation in nephrology fellowships nationally. As such, we identified several key differences among training programs and considerations of program directors when using NF. We do acknowledge several limitations of this study. The survey was internally developed and not pilot tested. In addition, the survey captured program director perceptions rather than objective outcomes. Interpretation of advantages and disadvantages therefore requires caution. Data on potentially identifying program characteristics, objective educational metrics, and overnight clinical volume were not collected. Finally, thematic analysis of open-ended responses may be subject to investigator bias; to mitigate this, we included a fellow, a program director, and a faculty member in the coding process to achieve consensus. These findings have several implications. Programs adopting NF may need creative scheduling strategies, such as repeated curricula or asynchronous learning, to meet educational requirements, approaches already used in undergraduate and graduate medical education.42,43 Larger programs may consider X + Y scheduling to preserve continuity clinics and maintain compliance with ACGME guidelines, whereas smaller programs may follow pediatric training models, in which X + Y scheduling is less common.35,44, 45, 46 Future studies should prioritize objective, quantifiable outcomes related to NF, including educational performance, patient care, and faculty workload, using both objective measures, such as board pass rates, in-training examination scores, and patient outcomes, and validated subjective measures, including trainee autonomy, readiness for independent practice, and faculty satisfaction, rather than relying on program director perceptions alone.38,47, 48, 49 Incorporating perspectives from fellows and applicants, along with additional program characteristics and more granular assessments of overnight workload, would further strengthen evaluation of educational, wellness, and recruitment implications. In addition, more granular data on the demographics of TPD themselves may help identify characteristics of TPDs willing to initiate change. In conclusion, TPD reported trainee interest in NF systems and highlighted perceived benefits for wellness, satisfaction, and recruitment. However, they also identified important perceived challenges related to program size, continuity clinic, and didactic attendance. The long-term implications of NF implementation on patient care and trainee education remain uncertain, underscoring the need for future studies using objective educational and clinical outcomes.