Crowley, Matthew J. MD; Barkauskas, Christina E. MD; Srygley, F. Douglas MD; Kransdorf, Evan P. MD, PhD; LeBlanc, Thomas W. MD; Simel, David L. MD; McNeill, Diana B. MD
The resident work hours limits implemented by the Accreditation Council for Graduate Medical Education (ACGME) in 2003 dramatically changed the landscape of graduate medical education. These regulations arose from concerns that traditional medical education promoted fatigue and errors,1,2 placing both patients and trainees at risk. Responses within the medical community to the duty hours rules have been both positive and negative,3–5 and multiple studies have attempted to evaluate the regulations' effectiveness in terms of adverse events and errors,6,7 patient outcomes,8–10 and resident perceptions.11–13 Although shorter shifts have been associated with reduced medical error in certain settings,14 some researchers have questioned whether the risk posed to patients by interruption of continuity of care may equal that associated with resident fatigue.12,15,16
In addition to necessitating more care transitions, duty hours limits have produced other challenges for residency training programs. Residents must care for increasingly complex patients and manage more administrative tasks within a limited time frame. When duty hours limits are implemented without substantial service redesign, the amount of work a resident needs to do does not change, but the amount of time available to do it is reduced. This “work compression” phenomenon likely contributes to perceptions of resident fatigue.
To combat work compression and minimize care discontinuity, residency programs have redesigned team systems, shifted patients to nonhousestaff providers, and enhanced sign-out procedures.17–19 Compression of the workday has also led to concerns about educational quality,11,20–22 forcing programs to reconsider how best to educate trainees in today's residency environment. The recent Institute of Medicine recommendations for patient safety and resident education23 may trigger continued evolution in medical education.
The 2008 Residency Match
In March 2008, the Internal Medicine Residency Program at Duke University Medical Center fell short of its anticipated quota of categorical applicants in the National Resident Matching Program (NRMP). Failing to “fill the Match” shocked program leaders and residents and prompted a comprehensive root cause analysis. We reviewed our Match process with the NRMP, including a detailed examination of internal data from recent years' Matches. This analysis revealed factors that likely contributed to our unexpected Match result, such as a decrease in internal medicine applicants in 2008 and a tendency toward ranking too few applicants. We felt that these factors alone did not offer a sufficient explanation, however, so we next surveyed current housestaff and 2008 applicants.
Although our resident survey identified multiple program strengths—for example, high-quality clinical training, level of resident autonomy, and nighttime chief resident sign-outs for on-call residents—more than 58% of respondents cited “high clinical volumes/patient load” as a perceived program weakness. The next most popular response was “lack of protected educational time.”24 Resident focus groups confirmed the impression offered by the survey results and also identified the inpatient general medicine service at Duke University Hospital as a primary source of resident dissatisfaction. Many participants indicated that patient volumes were too high for this service to realistically handle, which led to challenges in adhering to duty hours, less time for reading and educational opportunities, and low morale.
Our post-Match survey of 2008 applicants who chose not to rank Duke highly supported housestaff input. Although most respondents complimented our residents, more than 40% identified “residents” as being among their top concerns; free-text comments alluded to “tired,” “burned-out,” and “overworked” housestaff.24 These comments represented a departure from those on prior years' surveys. It had become clear to applicants that our residents had too much work to do within the allotted time frame and that this work compression was affecting their job satisfaction.
After reviewing these data, program leaders acted immediately to reduce patient volumes on the Duke general medicine service, primarily by shifting patients to nonhousestaff services. However, we felt this service required changes to ensure an optimal resident experience. All internal medicine programs work under ACGME rules, so we decided to visit peer programs to learn how they had addressed the same challenges we faced. We describe this process and our findings here. To the best of our knowledge, no previous published report has summarized and compared service models adopted by multiple programs in response to ACGME regulations.
Designing the Resident Site Visit Project
In June 2008, we established a committee to recommend structural changes to the Duke general medicine service. This committee included the program director, three associate program directors, the three chief residents, six residents, three internal medicine faculty members, and two other program staff members. The Resident Site Visit Project (RSVP) was designed with the goal of sending our residents to other sites to learn how peer programs had structured their inpatient general medicine services to (1) provide safe and effective patient care, (2) maximize educational opportunities, and (3) ensure duty hours compliance while imparting flexibility to adapt to future rule changes. We targeted programs that were similar in size to our own and that we perceived as having successful program designs based on prior NRMP results.
We contacted seven academic internal medicine training programs to discuss site visits; we focused on programs in the northeastern and southeastern United States to minimize travel expenses. Six of the seven sites agreed to accommodate a two-day visit during summer 2008. Our chief residents prepared a standardized list of quantitative and qualitative questions to be answered during each visit (see Supplemental Digital Content 1, http://links.lww.com/ACADMED/A18).
Two-person RSVP teams, each composed of a Duke chief resident and senior resident, met with program directors during the site visits to learn about their 2008–2009 general medicine service and to discuss perceived service advantages or challenges. If programs used different service structures at separate hospital sites, teams collected information pertaining only to the primary teaching hospital's general medicine service. The RSVP teams also met with chief residents and interviewed housestaff to explore the practical workings of each service, and they observed rounds and patient handoff procedures when possible. On completion of the visits, teams compiled a narrative description of each program's inpatient general medicine service and tabulated data collected. Before including program-specific data in this project, we asked program directors and department chairs to review and approve the information we compiled. The RSVP teams reconvened to present their findings to our committee and guide us in articulating core principles for the redesign of the Duke general medicine service. We subsequently assessed agreement among the site visit program directors regarding our core principles using a Likert scale.
Findings From the RSVP
Demographic data and service information for our institution and the six programs visited are shown in Tables 1 and 2. Below, we describe our program and then, in alphabetical order, the peer programs at which RSVP teams conducted site visits. All program information pertains to the 2008–2009 academic year.
At Duke University Medical Center, medicine patients are admitted to the general medicine service or multiple subspecialty services. Each general medicine team consists of one intern and one resident, and teams work overnight every fourth night. The intern has an admit cap of 5 new patients, and the resident has an admit cap of 7 patients (5 of whom are worked up jointly with the intern). Each team is responsible for 14 patients; the intern can care for a maximum of 12 patients, so the resident cares for the additional 2 patients alone. Overnight, the on-call intern covers 25 to 40 patients. A day float resident rounds with the postcall team and assumes primary responsibility for patients when the team leaves the hospital. Trainees are guaranteed one day off per week. Two hospitalist services care for nonteaching patients.
Programs visited by RSVP teams
Brigham and Women's Hospital, Boston, Massachusetts.
Medicine patients are admitted to the general medicine service or subspecialty services, and all services use an identical team structure. Each ward team consists of two interns and one resident, and teams take call every fourth night. While interns work overnight, residents admit until 10:00 pm and then return the next morning. A night float resident works with the on-call interns after their primary resident leaves and takes over cross-cover responsibilities after midnight to provide protected time for interns to complete admissions. A day float resident helps cover the departing interns during the postcall day. Each intern has an admit cap of 5 new patients and is responsible for a maximum of 10 patients at any given time; the team cap is 20 patients. Trainees are guaranteed one day off per week and one weekend off per month; coverage for these “Golden Weekends” is provided by day float residents. Physician extenders assist housestaff on multiple services, and there is a separate nonteaching hospitalist service.
Johns Hopkins Hospital, Baltimore, Maryland.
Medicine patients are admitted to general medicine services and subspecialty services. Each general medicine “firm” consists of four interns and two senior residents. There are four inpatient medicine firms; throughout their residency experience, trainees work within the same firm with the same group of residents, attendings, and ancillary staff. Each intern works overnight every fourth night, admits new patients from 12:00 pm to 1:00 am, and presents them to the rest of the team on postcall mornings. Admissions between 1:00 am and 8:00 am are completed by a night float resident; a day float resident covers admissions from 8:00 am until 12:00 pm and helps postcall interns complete their work and leave by 6:00 pm. Each intern has an admit cap of 5 new patients plus 2 overflow patients and is responsible for 6 to 7 patients at any given time. While on call, each intern covers the firm's 20 to 30 patients. Trainees are guaranteed one day off per week, as well as one weekend off per month. One hospitalist service cares for nonteaching patients.
Massachusetts General Hospital, Boston, Massachusetts.
Most medicine patients are admitted to general medicine services, as there are no medical subspecialty teams. There are two separate general medicine services with distinct structures; only one is described here. Each medicine team within this service consists of four interns and one resident and covers patients on one assigned ward. This geographic organization provides hard caps teams cannot exceed, as bed availability on the wards limits each team to 20 to 24 patients. Interns share responsibility for patients on their ward; the team's resident provides oversight for the team's patients. Teams admit daily, with each intern taking overnight call every fourth night. Interns have an admission cap of 5 new patients and provide night coverage only for the patients on their assigned ward. Residents do not take overnight call. Night float residents provide overnight support for on-call interns. Residents have each weekend off, and interns are guaranteed one day off per week. One hospitalist service cares for nonteaching patients.
Mount Sinai Medical Center, New York, New York.
The medicine service is organized into three teaching “firms,” each of which provides care for different groups of general medicine or subspecialty patients and includes five to eight interns, four residents, and one physician assistant. Within each firm there are distinct teams of two interns and one resident. Interns work overnight every eighth night and, while on call, are responsible for covering 15 to 25 patients. Postcall interns sign their patients over to a physician assistant. Residents on the medicine service do not work overnight, but distinct night float residents help admit patients overnight during a 12-hour shift. The maximum number of admissions in a 24-hour period is 4 for interns and 10 for residents. Each team has a cap of 16 patients. Trainees are guaranteed one day off per week. There is a large, nonteaching hospitalist service, which admits throughout the day and also after the medicine team caps. Physician extenders are used on each of the firms, and there is a full-time extender who remains in the emergency department to assist with medicine admissions and help manage patients awaiting beds. Of note, the Mount Sinai program functions under New York State rules that limit consecutive resident work hours to 24 plus 3 hours for transitions.
University of North Carolina Hospitals, Chapel Hill, North Carolina.
Medicine patients are admitted to the general medicine service and multiple subspecialty services. A typical daytime medicine ward team consists of two interns and one resident. Neither interns nor residents on medicine services generally take overnight call, although interns work an overnight call shift every two weeks to help provide weekend coverage, and the cardiology housestaff take call every fourth night. The day teams admit new patients every other day, with each intern admitting every fourth day. Interns have an admission cap of 5 new patients and can care for a maximum of 12 patients at a time; each team cares for a maximum of 24 patients. A night float system consisting of five junior or senior residents performs overnight admissions and provides nighttime cross-coverage on 20 to 40 patients for all medicine services. Each trainee is guaranteed one day off per week. One hospitalist service cares for nonteaching patients.
Vanderbilt Medical Center, Nashville, Tennessee.
Medicine patients are admitted to the general medicine service and multiple subspecialty services. Each general medicine ward team consists of one intern and one resident, who work overnight every fourth night. Interns cap at 5 new admissions and are responsible for 10 to 12 patients, whereas residents cap at 7 admissions, 5 of whom are evaluated jointly with the intern. While on-call, each intern is responsible for covering 20 to 30 patients. There is no night float or day float call system; when the postcall team completes its work, the team hands its patients off to the subsequent on-call intern. Each trainee is guaranteed one day off per week. Physician extenders assist on the subspecialty services, and there is a distinct nonteaching hospitalist service.
Core Principles for Program Redesign Drawn From the RSVP
Following the site visits, we reviewed the information we collected, compared the programs' service structures, and identified common issues and challenges. Our discussion of these challenges helped us articulate six core principles to guide the redesign of our own inpatient general medicine service. Summary data regarding the degree to which the program directors at the sites we visited felt our core principles were relevant to their own program redesign efforts are shown in Table 3. Programs generally agreed with our core principles, which we describe below.
Emphasize patient safety
We identified two primary components of safe patient care that can be affected by medical service design:
Safe care transitions:
Handoffs are unavoidable in residency training programs and represent opportunities for miscommunication. Ensuring appropriate staffing during care transitions should help housestaff communicate vital information and prevent them from overlooking follow-up tasks. We observed several approaches to service design that facilitate care transitions, such as larger teams that share patients, day float residents, and overlapping shifts.
Minimization of resident fatigue:
Although there is no consensus in the literature regarding the effect of duty hours limitations on patient safety via reduction of fatigue, it stands to reason that rested residents function better than fatigued residents. Night float rotations were the most common approach to reducing fatigue among the programs we visited.
Reduce resident work compression
Residency has always been service-oriented; trainees learn while they provide care for patients. There is no substitute for hands-on experience in housestaff education, but to extract maximum learning from patient care, residents need time to think and read about encounters. Our site visits revealed a common sentiment that today's compressed training experience provides less time for introspection and collaborative learning. Although hiring hospitalists or physician extenders requires a significant investment by hospital administrators, nonteaching services facilitate decompression of the resident workday and were used by all of the programs we visited. Sites visited also used other strategies to decompress the resident workday, such as reducing nighttime cross-cover and adding ancillary staff to help with the administrative tasks associated with patient care.
Create educational opportunities
As medical technology evolves and patient care becomes more complicated, residents face more demands on their increasingly limited in-hospital time. Our site visits confirmed that all programs protect time for resident learning during the workday by offering formal didactic sessions, rounds with senior teaching faculty, and time for reading and reflection. These opportunities are an important adjunct to patient-centered learning on the wards.
Ensure automatic duty hours compliance
Programs continue to encounter challenges in ensuring duty hours compliance. Our program's experience has shown that if resident shifts push the boundaries of a duty hours rule, there will be unavoidable violations that arise from a desire to serve patients and that do not represent program expectations or intentional disregard by residents. Many of the residents with whom we spoke at the sites visited alluded to instances when they felt that duty hours limits forced them to choose between patient care and obeying the rules. Using shifts that fall short of duty hours limits may alleviate this tension by ensuring that residents will reliably adhere to these limits even if they must stay beyond their scheduled time.
Preserve essential program attributes
There are different ways to train excellent residents within the framework of the ACGME rules. Each of the programs we visited possesses unique features that contribute to an effective training experience. It is essential that programs maintain their distinctive attributes as they evolve.
Involve stakeholders in the process of change
The housestaff and faculty of the programs we visited indicated that they—the main stakeholders—sometimes perceive change as threatening to their program's identity. Therefore, the process by which a program implements change carries great importance. Residents and faculty must be involved in shaping programmatic change, because proceeding without buy-in from these groups risks alienating the most important members of the program community.
Significance of the RSVP
Training programs must adapt when faced with circumstances such as resident dissatisfaction, a lack of success in the NRMP Match, or evolving ACGME rules. As all programs are obligated to work under the same rules, there is a role for mutually beneficial collaboration. To date, published experiences with program redesign have generally focused on single programs.17–19 The RSVP represents a novel approach to programmatic change: We formally visited six peer programs, systematically collected information on their different service models, and used these data to inform the definition of program-specific core principles to guide our service redesign. Including residents in this process enhanced our understanding of peer-program service structures and enhanced acceptance of the changes to our general medicine service that resulted from the RSVP.
Because residency programs evolve within distinct historical contexts, we were unsurprised to observe significant variability in general medicine service models among the sites we visited. However, we also noted that, despite their differences, programs face many of the same challenges as they attempt both to optimize resident education and to adhere to ACGME regulations. Common challenges we observed included adhering to duty hours restrictions and balancing duty hours adherence against care discontinuity, minimizing the effect of work compression on resident education, and maintaining program identity in the face of duty hours restriction. Identifying common challenges helped us focus our attention on key issues at our institution and led to our core principles for program redesign.
The RSVP core principles directly informed the changes that the Duke University Internal Medicine Program made to the Duke general medicine service for the 2009–2010 academic year. To emphasize patient safety, we focused on transitions of care by allowing for more overlap between the shifts of outgoing and oncoming residents, which should promote shared responsibility and decrease the negative effects of handoffs. In addition, we attempted to minimize resident fatigue by selectively employing night float rotations. Despite some concerns as to whether night float rotations reduce fatigue,25–27 our site visits suggested that they are a reasonable option given current ACGME rules. To reduce work compression, we acquired additional hospitalist support to help relieve resident teams, and we shifted overnight coverage to hospitalists to reduce residents' cross-coverage responsibilities. To create educational opportunities, we emphasized protected time for formal didactic sessions and rounds with teaching faculty. To ensure automatic duty hours compliance, we reduced call shifts to 25 hours so that, even when unexpected patient care situations arise, duty hours adherence could occur smoothly and reliably. To preserve essential program attributes, we retained nighttime chief resident sign-outs, overnight call for interns (now every fifth night in place of every fourth night), and other traditional components of training at Duke. Finally, we will continue to include our most important stakeholders—housestaff and faculty members—in the process of change. Site visits, as well as tools such as surveys and focus groups, represent valuable strategies to involve these groups.
Although program directors at the sites we visited tended to agree with our core principles for program redesign, the principles presented in this article represent our interpretation of the RSVP data. Other programs would likely arrive at alternative core principles, reflecting their unique histories and training goals.
This project has several limitations. We focus on relatively few programs that are similar to our own in size, and we have likely omitted approaches to general medicine service design. There is also a regional bias in the programs visited. Because the goal of this article is to demonstrate a process of collaboration, we do not evaluate the services in terms of efficacy or resident satisfaction measures, and we do not attempt to offer global recommendations for program redesign. We do not intend to propose the services described here as models for how others should structure their own services.
Because all programs have distinct histories, training priorities, and physical limitations, there is likely no single “best practice” model for general medicine services. However, the RSVP approach should be of widespread relevance even if the results achieved through the process differ among programs. Programs seldom have the opportunity to compare educational methods directly; formalizing arrangements to perform external comparisons on a regular basis offers great potential as a means to expand program horizons and aid in the dissemination of successful educational practices. Such information sharing is particularly important given today's centrally regulated resident education environment.
Further research is needed to rigorously examine service model features shared by different programs, such as night float rotations and handoff processes. This project highlights the importance of multicenter studies of these components, which would increase the generalizability of such research and could represent a powerful tool to move resident education in a data-driven direction.
Because all training programs operate under the same rules, each program has much to learn from how others have responded to the common challenges these rules create. The RSVP collaboration helped us develop core redesign principles to restructure our general medicine service in a manner we feel is compatible with Duke's tradition and history. Although core principles and resulting best practice service models would likely vary across programs, the RSVP may demonstrate a “best process” approach to programmatic change. Ultimately, we believe that collaboration among peer programs through resident site visits facilitates innovation, creates a foundation for change that increases stakeholder involvement, and opens the door to further collaborative research.
The authors thank the programs that facilitated this project by hosting resident teams. In alphabetical order by program (program director listed first, followed by 2008–2009 chief residents): at Brigham and Women's Hospital, Joel Katz, MD, Rafael Bejar, MD, PhD, Rebecca Berman, MD, and Tyler Berzin, MD; at Johns Hopkins Hospital, Charlie Wiener, MD, William Fischer, MD, Brian Garibaldi, MD, Emily Sydnor, MD, and Ryan Tedford, MD; at Massachusetts General Hospital, Hasan Bazari, MD, Eugene Rhee, MD, Rachel Simmons, MD, Ryan Thompson, MD, and Rory Weiner, MD; at Mount Sinai Medical Center, Mark Babyatsky, MD, Lauren Peccoralo, MD, and Michael Pourdehnad, MD; at University of North Carolina Hospitals, Lee Berkowitz, MD, Cheryl Davis, MD, and Wood Gibbs, MD; and at Vanderbilt Medical Center, John Sergent, MD, Todd Bauer, MD, Jennifer Clune, MD, Natasha Schneider, MD, and Isaac Thomsen, MD.
The authors thank Matthew Cavender, MD, Neela Goswami, MD, Jay Pyrtle, MD, and Rasheeda Stephens, MD, for participating in site visits. The authors would also like to acknowledge Amy Bus for her invaluable assistance coordinating this project, and Benjamin J. Powers, MD, for his critical review of the manuscript.
The authors thank Harvey Jay Cohen, MD, and the Duke University Medical Center Department of Medicine for supporting this project.
As the RSVP does not include a significant research component, IRB approval was not sought. Program directors and department chairs at the sites visited reviewed and approved the inclusion of all data pertaining to their programs.
1Philibert I, Friedmann P, Williams WT. New requirements for resident duty hours. JAMA. 2002;288:1112–1114.
2Veasey S, Rosen R, Barzansky B, Rosen I, Owens J. Sleep loss and fatigue in residency training: A reappraisal. JAMA. 2002;288:1116–1124.
3Charap M. Reducing resident work hours: Unproven assumptions and unforeseen outcomes. Ann Intern Med. 2004;140:814–815.
4Fletcher KE, Davis SQ, Underwood W, Mangrulkar RS, McMahon LF, Saint S. Systematic review: Effects of resident work hours on patient safety. Ann Intern Med. 2004;141:851–857.
5Skeff KM, Ezeji-Okoye S, Pompei P, Rockson S. Benefits of resident work hours regulation. Ann Intern Med. 2004;140:816–817.
6Mycyk MB, McDaniel MR, Fotis MA, Regalado J. Hospitalwide adverse drug events before and after limiting weekly work hours of medical residents to 80. Am J Health Syst Pharm. 2005;62:1592–1595.
7Vidyarthi AR, Auerbach AD, Wachter RM, Katz PP. The impact of duty hours on resident self reports of errors. J Gen Intern Med. 2007;22:205–209.
8Horwitz LI, Kosiborod M, Lin Z, Krumholz HM. Changes in outcomes for internal medicine inpatients after work-hour regulations. Ann Intern Med. 2007;147:97–103.
9Volpp KG, Rosen AK, Rosenbaum PR, et al. Mortality among patients in VA hospitals in the first 2 years following ACGME resident duty hour reform. JAMA. 2007;298:984–992.
10Volpp KG, Rosen AK, Rosenbaum PR, et al. Mortality among hospitalized Medicare beneficiaries in the first 2 years following ACGME resident duty hour reform. JAMA. 2007;298:975–983.
11Jagsi R, Shapiro J, Weissman JS, Dorer DJ, Weinstein DF. The educational impact of ACGME limits on resident and fellow duty hours: A pre–post survey study. Acad Med. 2006;81:1059–1068.
12Jagsi R, Weinstein DF, Shapiro J, Kitch BT, Dorer D, Weissman JS. The Accreditation Council for Graduate Medical Education's limits on residents' work hours and patient safety. A study of resident experiences and perceptions before and after hours reductions. Arch Intern Med. 2008;168:493–500.
13Lin GA, Beck DC, Stewart AL, Garbutt JM. Resident perceptions of the impact of work hour limitations. J Gen Intern Med. 2007;22:969–975.
14Landrigan CP, Rothschild JM, Cronin JW, et al. Effect of reducing interns' work hours on serious medical errors in intensive care units. N Engl J Med. 2004;351:1838–1848.
15Horwitz LI, Moin T, Krumholz HM, Wang L, Bradley EH. Consequences of inadequate sign-out for patient care. Arch Intern Med. 2008;168:1755–1760.
16Okie S. An elusive balance—residents' work hours and the continuity of care. N Engl J Med. 2007;356:2665–2667.
17Myers JS, Bellini LM, Rohrbach J, Shofer FS, Hollander JE. Improving resource utilization in a teaching hospital: Development of a nonteaching service for chest pain admissions. Acad Med. 2006;81:432–435.
18Ogden PE, Sibbitt S, Howell M, et al. Complying with ACGME resident duty hours restrictions: Restructuring the 80-hour workweek to enhance education and patient safety at Texas A&M/Scott & White Memorial Hospital. Acad Med. 2006;81:1026–1031.
19Wong JG, Holmboe ES, Huot SJ. Teaching and learning in an 80-hour work week: A novel day-float rotation for medical residents. J Gen Intern Med. 2004;19:519–523.
20Lin GA, Beck DC, Garbutt JM. Residents' perceptions of the effects of work hour limitations at a large teaching hospital. Acad Med. 2006;81:63–67.
21Mathis BR, Diers T, Hornung R, Ho M, Rouan GW. Implementing duty-hour restrictions without diminishing patient care or education: Can it be done? Acad Med. 2006;81:68–75.
22Myers JS, Bellini LM, Morris JB, et al. Internal medicine and general surgery residents' attitudes about the ACGME duty hours regulations: A multicenter study. Acad Med. 2006;81:1052–1058.
23Iglehart JK. Revisiting duty-hour limits—IOM recommendations for patient safety and resident education. N Engl J Med. 2008;359:2633–2635.
24Unpublished internal data. Durham, NC: Department of Medicine, Duke University Medical Center; 2008.
25Akl EA, Bais A, Rich E, Izzo J, Grant BJ, Schunemann HJ. Brief report: Internal medicine residents', attendings', and nurses' perceptions of the night float system. J Gen Intern Med. 2006;21:494–497.
26Cavallo A, Jaskiewicz J, Ris MD. Impact of night-float rotation on sleep, mood, and alertness: The resident's perception. Chronobiol Int. 2002;19:893–902.
27Cavallo A, Ris MD, Succop P. The night float paradigm to decrease sleep deprivation: Good solution or a new problem? Ergonomics. 2003;46:653–663.