In July 2003, the Accreditation Council for Graduate Medical Education (ACGME) put in place duty hour regulations that were designed to improve patient safety by reducing fatigue. Prior to July, 2003, residents often worked 100 or more hours a week. However, there have been concerns that the duty hours have created problems with continuity of care, given the greater need for hand-offs and cross-coverage.
We undertook a detailed qualitative assessment of residency training to examine how residents cope with duty hours in terms of their work routines, communications, and approaches to patient safety.
To develop observational and interview protocols for use in examination of how duty hour limits in internal medicine and surgery affect provision of clinical care, resident education, and the quality of residents' lives. These protocols will be used to help examine duty hour limits in a larger-scale study.
The first eight weeks of the study will involve the observation of residents and attending faculty on the medical and surgical services at the PVAMC. After that period, ten internal medicine and ten surgical residents will be interviewed along with five internal medicine and five surgical attending faculty. Observation will be conducted in four settings: (1) a general medical floor team at PVAMC; (2) the medical/cardiac intensive care unit at PVAMC, which recently switched to a 12 hour shift system primarily using moonlighters in place of residents; (3) a general surgery floor team. These will provide useful contrasts of how different types of services have coped with the ACGME duty hour restrictions.
Methodology: The observation of residents will be led by Professor Charles Bosk, a sociologist with extensive experience with hospital observational studies and site visits examining resident training, including seminal work on the training of surgeons. Dr. Bosk will direct a team of 4 upper-level graduate students in sociology and anthropology to observe and interview residents. Site visits will consist of (a) direct observation of resident involvement in provision of hospital care; and (b) interviews with residents and attending physicians. At the six-week period of the eight weeks of observations, the observers, under Dr. Volpp's and Dr. Bosk's supervision, will develop an interview protocol for discussing observations with residents. Dr. Volpp will work with Dr. Bosk to develop an interview protocol to be used with attending faculty. The interview protocols will then be piloted during the last week in July and the first week in August. All interviews with providers will be audio-taped and transcribed.
We were able to develop protocols for observing internal medicine and general surgery services. The response to duty hour limits was variable by specialty: residents in internal medicine viewed the limits favorably while residents in general surgery worried that the regulations were likely to diminish their operative experience during training.
We are still formally developing refined codes for analyzing our observational and interview data. However, we are confident that a few broad general impressions from the observational and interview data will hold up to the most refined analysis.
First, organizational responses to duty hours affect the quality of patient care. We observed three different responses to duty hour restrictions in the observational setting. On general medicine floors, resident had an every third night call; general surgery residents were on call once a week and coverage was provided by moonlighting surgical residents during their laboratory years; and in the MICU, coverage was provided by a combination of physicians on staff or moonlighters. Residents were most comfortable signing out to residents that they knew and uncomfortable signing out to 'strangers'.
In neither specialty did residents feel that duty hour limits were likely to any but the most minimal impact on patient safety. Residents in internal medicine felt that duty hour restrictions were not sufficient to relieve chronic fatigue. Surgical residents felt that not only were duty hour restrictions insufficient to reduce fatigue but they also added to confusion about who was responsible for patient coverage.
All residents had devised for themselves standardized ways of signing out patients to covering residents and standardized ways of making sure that they were brought up to speed when coming on duty. These efforts to create personal checklists for sign-outs were made necessary by the absence of any organizational protocols.
Had a standardized procedure for sign-outs existed, it is unclear residents would have complied with its demands. A number of factors create variability from one hand-off to the next, despite the efforts of residents to follow some standardized way of giving and receiving critical information. Among these factors are the following: (1) the number of 'unstable' patients being signed out; (2) the level of fatigue for the resident going of shift; (3) the level of familiarity with patients being covered possessed by resident coming on shift; (4) the level of trust between the resident coming off and the resident coming on duty; and (5) the amount of confidence that both residents have in nursing coverage.
There are two significant impacts from the summer's work. First we demonstrated the feasibility of developing observational and interview protocols to investigate the impact of duty hour restrictions on inpatient care, patient safety, and resident education. Second, we found that the impact of duty hour restrictions varies with the strategies organizations use to adapt to these restrictions.
The work done in this proposal was helpful in securing funding from NIH for a larger-scale R01 study in which we will do site visits at 8 different residency training programs and use insights from these visits to conduct nationwide surveys of residents and residency program directors to quantify the behavioral responses to duty hour reform. The approach we have developed to site visits will be instrumental in this project.
External Links for this Project
None at this time.