In the 6 years since implementation of the common duty hour standards on July 1, 2003, programs and institutions have made changes in education and patient care and have achieved sizable gains in compliance, with the percentage of programs cited for duty hour noncompliance currently hovering around 7 percent. Concurrently, the Accreditation Council for Graduate Medical Education (ACGME) has made enhancements to its processes for monitoring and promoting program and institutional oversight of resident duty hours. In 2003, it piloted a resident survey, and between 2004 and 2006, it surveyed all accredited specialty programs and subspecialty programs with 4 or more fellows. The survey was repeated in 2007 and 2008, with half of the programs included each year, and in 2009, the ACGME surveyed all specialty programs, as well as subspecialty programs with 4 or more fellows. During each administration, the survey found a small number of programs (around 5 percent of those surveyed) in which a significant percentage of residents reported noncompliance with several standards. Follow-up has included resurveying, requests for information on how the problem is being addressed, and assessment during the next site visit. In 2008, the ACGME stepped up compliance measures for this cohort, and between October 2008 and March 2009, it conducted site visits of the programs with annually recurring (“continuous”) and multiyear “significant” noncompliance. Program visits entailed review of a full program information form, with added emphasis on duty hour compliance. The concurrent focused institutional reviews for programs in the chronic (successive year) violator group encompassed a review of program-level citations, monitoring of duty hour compliance, documentation of how the program and institution addressed noncompliance problems, and meetings with residents, with oversampling of residents from programs with duty hour noncompliance and those from procedural and other specialties in which residents are known to work longer hours.

Although some of the debate in the community has focused on misinterpretations of the questions in the resident survey, for most programs with a site visit, there was clear indication of duty hour noncompliance at the time the survey was administered. In about 50 percent of the programs with a confirmed compliance problem at the time survey was administered, the site visitor was able to confirm current compliance and report on changes the program had made to resolve previously existing problems. Despite ACGME concerns about the effect of the limits on safety net institutions, many programs with multiyear “significant noncompliance” were located in some of the nation's highly respected teaching institutions.

A benefit of the moved-up site visits to assess resident hours was that program directors and designated intuitional officials became aware that the ACGME took duty hour compliance seriously. This included individuals in programs undergoing the special site visits and others who became aware of the stepped-up compliance process.

The site visits highlighted common problem areas and innovative practices to deal with those problems. Surgery and surgical specialty programs made up most of the programs in the “chronic” violator group, and 2 common causes of noncompliance were identified across programs: (1) rotations to institutions valued for their clinical volume and intensity, but with a heavy service component; and (2) excess hours because residents identify deeply with the culture of their program, with longer hours attributed to their engagement in and commitment to their program and patients. Beyond these 2 common findings, the nature of problems differed among programs, with no additional common factors emerging.

The process also identified “innovative practices,” such as program coordinators calculating the number of hours residents had been on service already in a given week and alerting residents and, more importantly, faculty preceptors to let them know they had a finite number of hours in the given week before they would exceed the limit or require a lower number of hours in the following week.

Although some members of the graduate medical education (GME) community have voiced concern about the seemingly rigid nature of the process by which the ACGME selected programs for an early site visit, the cohort selected in 2008 was a small percentage (4.8 percent) of all accredited programs, and its members represented the “tip of the noncompliance iceberg.” In less than one-fourth of the programs undergoing a moved-up site visit, the interviews did not confirm the survey findings of noncompliance, with disgruntled residents who completed the last survey given as the likely source. Concerned that residents might not have been able to speak freely, field representatives provided contact information for participants who wanted to provide additional information in a confidential fashion. This did not produce any additional information.

During the site visit, when ACGME resident survey data and the results of institutional duty hour monitoring were compared, this found only modest agreement, with differences in survey design, timing of data collection, and data aggregation issues identified as barriers to reconciliation of the 2 sources. Other potential reasons for variance include trust issues reported for both institutional data collection and the ACGME resident survey.

A beneficial aspect of the institutional reviews was the focus on the GME committee's effort to address program-level compliance issues, with ongoing review of changes and improvement activities and documentation in GME committee minutes. Site visitors were able to distinguish between institutions that had regularly discussed and addressed noncompliance issues in institutional committees and dedicated groups, and those where no discussion of duty hour compliance had occurred. For most programs, use of night float and other rotation changes, as well as replacing residents with midlevel practitioners or hospitalists, were common strategies to achieve compliance. A benefit of an institutional focus on duty hour compliance was the ability to involve institutional leaders, particularly leaders above the level of the designated institutional official, such as deans and CEOs. This frequently brought added institutional resources to bear on a seemingly intractable compliance issue.

Residents in programs and sponsoring institutions undergoing duty hour site visits had a sophisticated perspective on instances where remaining on duty would benefit the patient and/or their education. Resident interviews revealed that overcoming rigid or narrow application of the standards at the program or institutional level was important to this group. At the same time, residents distinguished these “added hours important for patient care and learning” from other factors that produced the chronic pattern of noncompliance they and their colleagues had reported in their responses to the resident survey. This included a common observation across a number of programs with significant noncompliance—residents were scheduled for the full 80 weekly hours and 24 plus up to 6 continuous hours. This resulted in residents fairly regularly exceeding these limits when unexpected patient care demands, valuable educational activities, or another type of contingency warranted their remaining beyond the scheduled period. This finding suggests that scheduling below the ACGME maximum to create flexibility for additional patient care demands and/or learning opportunities may produce a favorable climate for compliance. In addition, time and human resources not “constantly subject to measures of short-term efficiency”1 could allow for end-of-shift transfers, educational debriefing and feedback, individual professional development, and organizational learning. Optimizing hours to ensure compliance and promote safe and effective care simply may require scheduling residents below the maximum number of hours to create a “cushion” for residents and their patients.

A second strategy to reduce excess hours entails identification of program- or institution-level “bottlenecks,” focusing on settings and situations that consistently or frequently lead to hours beyond those scheduled, and addressing them. Some bottlenecks likely are year-round phenomena, but others may be seasonal, such as high patient census in the inpatient and observation units during the winter season, or emergency department overcrowding during the summer months. Added planning for these contingencies may further assist in reducing the likelihood of noncompliance with the duty hour standards.

The findings from the ACGME's effort to address serious instances of duty hour noncompliance suggest a need for added education of program directors, institutional officials and, particularly, residents about the standards and their respective responsibilities for compliance. In the coming months and years, the ACGME plans to make further refinements to its process for responding to duty hour noncompliance, including enhancing its ability to learn from programs and sponsoring institutions and to disseminate innovative practices to benefit the educational community.

1
Lawson
,
M.
In praise of slack: time is of the essence.
Acad Manag Exec
2001
.
15
(
3
):
125
135
.

Author notes

Ingrid Philibert, PhD, MBA, is Senior Vice President for Field Activities, Accreditation Council for Graduate Medical Education, and the Managing Editor of the Journal of Graduate Medical Education. Jeanne K. Heard, MD, PhD is the Senior Vice President, Department of Accreditation Committees, Rebecca Miller, MS is the Vice President and Kathleen Holt is Senior Data Analyst in the Department of Applications and Data Analysis, Accreditation Council for Graduate Medical Education.