On July 1, 2003, new standards for resident work hours set forth by the Accreditation Council for Graduate Medical Education (ACMGE) went into effect. These rules limit duty hours in all specialties to an average of 80 hours per week. Key components of these new rules include a 24-hour limit on continuous work time, with up to an additional six hours to participate in transition of care (sign-outs) and didactic activities, and a requirement for ten hours off between shifts.1 These standards are the ACGME's response to the changing environment of health care delivery in the United States and concerns that sleep deprivation among medical residents has harmful effects on patients’ care and safety, medical education, and residents’ well-being.2
Prior to the implementation of the ACGME's work hour rules, there was widespread concern for the impact of the changes on medical education. New York State's residency programs have over a decade of experience with work hour restrictions, however, because the state implemented Section 405 of the State Health Code legislating work hour regulations (also known as the Bell Commission Regulations) in 1989.3 Although New York's rules differ somewhat from the ACGME's (for instance allowing only a three-hour transition period and requiring just eight hours off between shifts), we believe that New York's residency directors can provide valuable insight on how to restructure residency programs to accommodate duty hour changes while maintaining levels of care.
We surveyed residency directors of pediatrics programs in the state of New York to evaluate the impact of the Bell Commission's work hour regulations on pediatrics training and patient care. Previous efforts have shown that many residency programs in New York struggled to meet these work hour restrictions.4 We hypothesized that work hour limitations would have a greater impact on smaller programs compared with larger programs because the smaller programs might have less flexibility in scheduling and that much of the negative impact would result from increases in the logistical work necessary to meet regulation requirements. We also expected to see decreases in outpatient continuity and community activities as programs worked to maintain required inpatient and subspecialty rotations. The study's secondary objective was to assess the strategies used and the programmatic impact of these strategies.
The study's participants were directors at all 32 pediatrics residency programs in the state of New York, which we identified through an online listing provided by the ACGME. We obtained additional data about the programs from the ACGME's Web site, including program size and characteristics of nonrespondents. The questionnaire was designed specifically for this study, and its questions were initially reviewed by faculty from eight programs outside New York. The questionnaire was pilot tested by program directors at three non-New York residency programs to prevent our resurveying program directors who had participated in the pilot test. Feedback from the reviews and the pilot test guided the final revision of the questionnaire. In April 2003, the questionnaire was mailed directly to the residency directors with a self-reply envelope and fax number for return. Telephone call reminders were made three, five, and ten weeks following survey mailing.
The questionnaire began by eliciting background information such as program size and staffing patterns through a set of open-ended and forced-choice questions. The next set of questions focused on the perceived effects of work hours using Likert-type responses on a five-point scale. Strategies to meet work hours requirements were assessed on a rotation-specific basis by asking respondents to check all the methods that applied. Seven methods were defined for respondents, as described in Table 1, and the questionnaire provided respondents with an opportunity to describe other strategies or note that no change was needed on a particular rotation. Finally, respondents were asked open-ended questions regarding best practices, problematic strategies, and advice for other programs. The entire questionnaire is available at 〈www.servingtheunderserved.org〉.
Our analyses were primarily descriptive. We determined the number of programs reporting each strategy. The results of the section relating to methods used to meet work hour requirements were grouped, collapsing all outpatient rotations. To measure the impact on program operations, Likert-type responses in whole integers were coded from –2 to + 2 (–2 = greatly decreased and + 2 = greatly increased). We reverse-coded items such as “logistical work to maintain residents’ continuity clinic,” where a decrease would be a benefit. Thus, for every question, a negative number represents a negative impact, and a positive number represents a programmatic improvement. For each item, we calculated the mean response and the percentage of respondents selecting negative responses.
Further analyses examined the effect of residency programs’ characteristics. For descriptive purposes, small programs were defined as those with fewer than 30 residents. Programs were additionally sorted by those self-identifying as “community based,” “university affiliated,” and “university based,” according to the Fellowship and Residency Electronic Interactive Database (FREIDA) on the American Medical Association's Web site 〈www.amaassn.org〉.Because the university-affiliated and based programs closely resembled the large programs in an initial analysis, self-identified program type is not presented here. Our statistical analysis compares program size (without dichotomizing small versus large programs) with responses to Likert-type questions using Spearman's rho (two-tailed).
A single reviewer (GWC) assessed the open-ended qualitative responses and categorized each response by the theme it most closely matched. Any responses not clearly matching a theme were discussed with a second reviewer (RCS), and the responses were assigned to a theme by their consensus.
The Institutional Review Board of Maimonides Medical Center approved this study.
Demographics and background information
Of the 32 programs, one was excluded due to a voluntary withdrawal of their accreditation, and 21 (68%) completed and returned the questionnaire. On average, program directors had been in their position for 6.4 years (range 0.5–23 years). The mean number of residents for all programs was 47 (range 12–99). The nonrespondent's information available from the ACGME's Web site was not significantly different from that of the respondents.
Four programs increased the number of residents to accommodate work hour regulations, and two program directors reported they did not know whether more residents had been added. Only one program director reported any information on how many residents were added. Eleven programs hired more nonresident staff (ten hired nurse practitioners/physician assistants, six hired more attending physicians, and five hired both). Most of those programs hiring attendings reported hiring one or two more staff, and most of those programs hiring nurse practitioners/physician assistants reported hiring three to five more staff. Only one program director reported hiring more “other clinical staff” to accommodate resident work hours, and none reported hiring “nonclinical staff.”
In comparing the methods used to meet work hour requirements in smaller and larger programs, only larger programs made use of night float and night teams to accommodate work hour regulations in inpatient and intensive care settings. Similarly, only larger programs made use of other staff and changing resident teams. Most programs, regardless of size, reported a considerable use of sending residents home immediately post call and cross coverage in all settings. None of these results were statistically significant. Table 1 shows the aggregate responses across all programs.
In continuity clinic, a variety of methods were used, including canceling continuity clinic post call, moving call days, moving the day of week for the clinic, and having morning continuity clinic. There was little discussion of continuity clinic strategies in the open-ended responses described below.
Perceived impact on program
The program directors reported the impact of work hours regulations was negative, if modest, on most aspects of training (see Table 2). The four aspects that program directors perceived as having made a substantial negative impact were the time residents spend in inpatient settings, patient continuity in inpatient settings, flexibility of residents’ schedules, and the logistical work needed to maintain continuity clinic. The only positive effect reported was the program directors’ perception of residents’ satisfaction with their schedules. In analyzing program size and work hour effects, some responses differed. Directors of larger programs were more likely to answer more negatively about the amount of logistical work needed to maintain continuity clinic (p < .05). There was also a trend toward increased residents’ satisfaction with work hours regulations reported for larger programs (p = .06).
Advice for other programs
The most common advice for other programs related to changes in a program's structure and staffing issues. Night float/night teams was the program directors’ most frequent response to what worked best, but only among large residency programs. Hiring other staff appeared in responses to both what worked best and what was most problematic. Issues of residents', faculty's, and institutional acceptance appeared in what worked best, and residents’ and faculty's resistance was seen in the most problematic responses. General advice to programs from several respondents included suggestions to be flexible and open-minded.
Fourteen years of work hour limitations have had an impact on New York State's pediatrics residency programs in numerous ways. These experiences, as presented here, provide programs throughout the United States with valuable insight on the impact of work hour limitations, such as the ACGME's; informative advice on how residency programs have restructured their programs; perceptions of how these changes affect residents and patients; and the pitfalls that come with those programmatic changes.
Most programs used multiple strategies to meet work hours requirements. Whereas no programs with fewer than 30 residents reported using night float/night teams as an option in any setting, many larger programs not only cited these methods but also described them as being among the best strategies. Almost half of all programs hired additional staff, including attending physicians, nurse practitioners, and physician assistants, but only four programs reported hiring more residents to meet work hour requirements. Hiring additional staff to meet work hour requirements has also been a common practice in other specialties,5 but the small number of programs increasing residents needs to be considered in the context of other forces acting to reduce the number of residency positions. Even before work hour regulations began in New York, many pediatrics residency programs, both in New York and nationwide, had large numbers of unfilled residency spots. Then, beginning in 1997, many of New York's residency programs took part in a federally funded effort, the Health Care Financing Administration Medicare Demonstration Project, to reduce the number of residency positions.6
No program reported the use of additional nonclinical staff. As Drs. Bergman and Moreno persuasively argue in separate commentaries published 15 years apart,7,8 reducing the nonclinical chores of residents could have a large impact on residents’ stress and workloads, yet our study found no indication that this was being done in New York.
There are additional lessons on program restructuring that can be learned from our colleagues in other countries, many of whom have been dealing with work hour restrictions for years and are moving toward even more restrictive regulations. Cass et al. 9 argue in their study of changes in the United Kingdom that simply changing rotation structures will not work as a solution to work hour restrictions. As a more effective way of meeting their own particular regulations, these authors made significant changes to who staffed specific locations and when staffing could change. Klazinga and van Bolderen make similar arguments in their discussion of the work hour issues in Europe.10 Instead of regarding work hour restrictions as a workforce issue, whereby reduced hours require more employees, they argue that the answer is not simply to add more physicians who will work fewer hours but to “redesign... the medical working processes” and “the professional culture regarding teamwork.” These sorts of changes were captured only broadly (if at all) in this study.
Perceptions of the impact on residents
Many of the program directors we surveyed connected work hour regulations to a decrease in the amount of time residents have for education, as has been seen in previous studies.11,12 However, the number of reports of a decrease in the quality of the education the residents received were not significant. This finding was also consistent with other literature where no clear negative consequences were found.13,14 In fact, in a recent commentary by Sectish et al., 15 the authors encourage creative ways to enhance education, such as Web-based learning.
Our hypothesis was that work hour regulations would have a negative impact on the amount of time residents spend in the community, thus decreasing the number of community sites residents were able to visit. However, our results indicate that neither the amount of time residents spend in the community nor the number of sites residents visited changed. In fact, a small proportion of programs even reported increases in both of these areas. We believe an ongoing emphasis on community-based activities in residency programs may reflect the importance of out-of-hospital education. Additionally, the lack of any negative effect may point to how difficult it has always been for residents to do any out-of-hospital work when they are post call. Surveying directors of continuity clinics might provide more details in this area, but this may be analogous to previous work on inpatient night float systems that have shown consistent evening coverage improves patient care.15,16 Having residents consistently available and not sleep-deprived when post call may improve community-based education. As long as this improved resident alertness is not offset by significantly less time in the community it is likely to be a net gain for education in the community.
Lastly, residency directors also believe that residents’ satisfaction with their schedules had improved. Most reported an increase in residents’ satisfaction despite a decrease in the flexibility residents had with their schedules. In fact, one program director even commented on this finding: “Residents have more of a life outside of residency. This is not a bad outcome.” This finding was also consistent with past literature in which residents and physicians both self-reported an improved quality of life.17,18 However, recent research among surgical residents working under work hours restrictions in New York found they were less happy with these regulations than were their peers in New England who had not at that time experienced the work hours rules themselves.19
Perceptions of the impact on patient care
There is a prominent body of literature examining the relationship between work hour regulations and its impact on patient care, but little statistically significant data exists to show a clear negative effect. 8,11–13,18 Most program directors in our study reported no difference in the quality of care provided by residents in both inpatient and outpatient settings. Some even reported that the quality of care had improved, especially in outpatient settings.
Some program directors did report that patient continuity in inpatient settings was affected by residency work hour regulations. Because potential preventable adverse outcomes have been associated with coverage by physicians from other teams, this remains a serious concern.20 Interestingly, a number of programs actually reported an increase in continuity in outpatient setting under work hour restrictions. Although it was beyond the scope of this study to examine this unexpected phenomenon, some possible explanations come to mind. For instance, we believe that eliminating the postcall continuity clinic may be a major improvement. It is possible that residents who were post call were less likely to establish good relationships with families, provide high quality care, and/or make appropriate referrals. Because most programs did not simply cancel the postcall clinic, but rather rearranged schedules in one of various ways, outpatient continuity was not significantly affected besides the logistical problems of moving the clinic around. This is an obvious area for further exploration in a future study.
Effect of program size
Program size appears to have a differential effect on the methods used to meet work hour regulations in various settings and in residents’ satisfaction. The differences in methods can best be seen in the use of night floats and night teams in inpatient wards and intensive care units. Most larger programs used these methods to accommodate work hour regulations while none of the smaller programs did. Similarly, only larger programs changed resident teams and hired more staff to accommodate work hour restrictions. A 1988 survey of pediatrics residents in California found that many residents preferred night float as a mechanism to meet work hour requirements.21 Although many authors have suggested this mechanism as a possible solution to meeting work hour regulations,16,22 our results indicate that the success of this strategy may depend on program size or other program-specific factors related to program size. A study of internal medicine programs by Trontell et al. 23 also found that larger programs were more likely to use a night float. All of these differences are likely related to the increased flexibility available in programs with more residents and the financial differences between smaller and larger programs in the current economic environment. However, it was the larger programs that reported an increase in the amount of logistical work needed to maintain outpatient continuity. It would appear that the price they pay for using the various options is more work logistically.
Because the number of residency programs in New York is fixed, our power to do a rigorous analysis of the data and come up with statistically significant findings was very limited. Thus, our results and discussion focus on the descriptive elements and the qualitative responses we received. A larger study with more power may be worth undertaking now that work hour restrictions are in effect nationwide. Although response rates from small and large programs did not differ, the number of small programs in New York is very limited, similarly limiting analysis of program-size effects. Lastly, this is a survey of current residency directors with various lengths of service in their positions. Their knowledge about past changes and how the current program reflects those changes differs among them depending on the length of time they have worked in the program.
In many ways, the diversity of what has been used in New York State's pediatrics residency programs to respond to work hour regulations points to what many program directors mentioned in their advice to other programs: “Be open and flexible to change. There is no easy or perfect way to go about this. And no one will ever be satisfied all the time.” Creative scheduling was also cited as essential to successfully accommodate work hour regulations in a study of obstetrics and gynecology residencies.22
We all agree that residents’ training is a stressful period in a new physician's life. Although the major factors leading to excessive stress are different for each individual, both acute and, more important, chronic fatigue are significant problems, which the ACGME rules are meant to address. The ACGME's new rules have established a consistent starting point, which we hope can be modified in the future to best meet the needs of our housestaff and our patients. Although accommodating the ACGME's rules is difficult for training programs, hopefully the experience of the New York programs can be used by the rest of the country to start with a little advantage. First steps are always the hardest, but it has been the authors’ intent to promote wiser choices through an understanding of the experiences in New York's pediatrics programs.
This study was funded by the Anne E. Dyson Community Pediatrics Training Initiative. Thanks to the many people who helped review the survey and manuscript, and a special thanks to Anne Duggan for all her assistance.
1Accreditation Council for Graduate Medical Education. Statement of Justification/Impact for the Final Approval of Common Standards Related to Resident Duty Hours; September 2002 〈www.acgme.org
〉. Accessed 18 August 2004.
2Accreditation Council for Graduate Medical Education. Report of the ACGME Work Group on Resident Duty Hours and Learning Environment; June 11, 2002 〈www.acgme.org
. Accessed 18 August 2004.
3Committee of Interns and Residents, A National Affiliate of Service Employees International Union. New York State Resident Work Hours and Hospital Code Regulations; 1989 〈www.med.nyu.edu/resweb/documents/cir_resident_working_hours.pdf
〉. Accessed 18 August 2004.
4Johnson T. Limitations on residents′ working hours at New York teaching hospitals: a status report. Acad Med. 2003;78:3–8.
5Nagel CW PD. The evolution of the nurse practitioner in the neonatal intensive care unit. Am J Perinatol. 2000;17:225–8.
6Centers for Medicare & Medicaid Services Office of Public Affairs. New York Teaching Hospitals Participate in Graduate Medical Education Demonstration; February 17, 1997 〈http://www.cms.hhs.gov/media/press/release.asp?counter=4
〉. Accessed 7 January 2004.
7Moreno MA. Resident stress revisited: a senior pediatric resident's point of view. Pediatrics. 2003;112:411–4. discussion 414-415.
8Bergman AB. Resident stress. Pediatrics. 1988;82:260–3.
9Cass HD, Smith I, Unthank C, Starling C, Collins JE. Improving compliance with requirements on junior doctors′ hours. [Published erratum appears in BMJ. 2003;327:586]. BMJ. 2003;327:270–3.
10Klazinga N, van Bolderen A. Epimetheus’ responsibility: resident working hours and system redesign. Int J Qual Health Care. 2003;15:109–10.
11Holtzman IR BS. The Bell Commission: ethical implications for the training of physicians. Mt Sinai J Med. 2000;67:136–9.
12Bergman AB, DeAngelis CD, Feigin RD, Stockman JA. Regulation of working hours for pediatric residents. J Pediatr. 1990;116:478–83.
13Brensilver JM SL, Lyttle CS. Impact of the Libby Zion case on graduate medical education in internal medicine. Mt Sinai J Med. 1998;65:296–300.
14Conigliaro J, Frishman WH, Lazar EJ, Croen L. Internal medicine housestaff and attending physician perceptions of the impact of the New York State Section 405 regulations on working conditions and supervision of residents in two training programs. J Gen Intern Med. 1993;8:502–7.
15Sectish TC, Carraccio C, Zalneraitis EL. Commentary on the ACGME duty hours standards: improved duty hours with uncertain outcomes. Ambul Pediatr. 2003;3:164–5.
16Yu JN. Night shift call systems in family practice residencies. Fam Med. 1994;26:163–7.
17Ozuah PO, Neuspiel DR, Shelov SP. The pediatric forum: trends in residents′ perceptions of working conditions: 1989-1999. Arch Pediatr Adolesc Med. 2001;155:1073–4.
18Whang EE, Mello MM, Ashley SW, Zinner MJ. Implementing resident work hour limitations: lessons from the New York State experience. Ann Surg. 2003;237:449–55.
19Whang EE, Perez A, Ito H, Mello MM, Ashley SW, Zinner MJ. Work hours reform: perceptions and desires of contemporary surgical residents. J Am Coll Surg. 2003;197:624–30.
20Petersen LA, Brennan TA, O'Neil AC, Cook EF, Lee TH. Does housestaff discontinuity of care increase the risk for preventable adverse events? Ann Intern Med. 1994;121:866–72.
21Cheng TL. House staff work hours and moonlighting: what do residents want? A survey of pediatric residents in California [see comment]. Am J Dis Child. 1991;145:1104–8.
22Foster HW Jr, Seltzer VL. Accommodating to restrictions on residents′ working hours. Acad Med. 1991;66:94–7.
23Trontell MC, Carson JL, Taragin MI, Duff A. The impact of the night float system on internal medicine residency programs. J Gen Intern Med. 1991;6:445–9.