NF is a potential solution to fulfilling the work hour restrictions recently mandated by the ACGME. Our study examined residents' opinions in the fundamental areas of patient care, education, medical errors, and overall resident satisfaction.
The possible lack of continuity of patient care during the NF rotation is a highly debated concern. In our study, an almost equal number of residents either agreed or disagreed that continuity of care was compromised, further supporting the complexity of the issue. Several interventions may be implemented to improve continuity of care. Since shared decision-making was thought to improve care, enhancing the communication between the two teams will likely help maintain continuity. For example, standardized sign-out practices have been shown to reduce errors made by residents during the night-time [24
]. Other strategies to improve continuity include pairing NF team members with members of the day team, all of whom would care for a specified patient population [25
], and close oversight by attending physicians. These interventions will encourage professionalism and a personal investment in the care of the patient, areas argued to be deficient in the NF system.
The lack of emphasis on education during NF is another significant concern. As institutions initiate a NF system, specific educational prescriptions are vital components for the success of the program. As noted earlier, responses to the open-ended questions in our survey suggest that the residents are concerned about the lack of feedback on their medical reasoning process and the lack of follow-up about the evolution of disease processes. Greater supervision and feedback by night-time attendings and an "evening report" (held during the beginning of each NF period) could perhaps help maintain continuity of care, provide feedback on the decisions made overnight, and clarify the development of disease processes. These interventions will also fulfill the recommendations of the Residency Review Committee (RRC), which require that "education in the context of [patient care] activities must be provided to each resident [during night float]."
In addition, a standard NF curriculum, focusing on conditions and problems that are frequently encountered during the night, might help educate residents and provide the formal conferences that were thought to be an essential but missing element of the NF experience. As Drs. Ende and Davidoff recommended, "viewing housestaff programs as enterprises for hospital-based service is increasingly unacceptable... an important step in recasting [the structure] will be the development of curricula" [26
The area of resident satisfaction revealed significant differences between the responses of interns (PGY 1) and residents (PGYs 2 and 3). There are several possible explanations for these differences. Overall, interns favored NF since it seemed to improve the quality of life in and outside of the hospital. This preference might be reinforced by the interns' schedules of having more ward months and generally being busier than residents. In contrast, the residents have fewer ward months and perhaps perceive that their limited learning opportunities are being compromised, as noted by the open-ended responses in the education domain.
There might also be a "cultural" effect. Although the interns had experienced a non-NF system for the first half of the academic year, the residents had experienced the system for one or two years. Perhaps this greater familiarity with a non-NF system influenced them to favor the traditional overnight call system. Lastly, the job functions of the interns and residents are different at our institution. Interns mainly cross-cover on patients already in the hospital, whereas residents admit new patients. This difference in the educational experience may also effect overall satisfaction.
Our study has several limitations that need to be acknowledged. Our questionnaire was designed to be the initial step towards the future development of a standardized instrument and thus did not include all the steps for ensuring reliability and validity. This process will be undertaken in future work. In addition, our survey did not collect socio-demographic information, which can potentially influence residents' perceptions. Finally, the respondents were from one institution with a specific type of NF rotation. The NF system at other institutions might be different, thereby limiting the generalizability of our results.
Compared to previous smaller studies, our detailed questionnaire (115 items) examining multiple areas and the high response rate (90%) provides a more accurate picture of the perceptions of housestaff towards the NF system. In addition, it incorporates the opinions of both interns and residents in different hospital environments, and those who had experienced both NF and non-NF systems. This is in contrast to other studies that had conducted surveys of residents who did not experience a non-NF system. Lastly, our study has been conducted at a time when the NF system and the associated working and educational environments have significantly changed since the time of many other studies. For example, there has been an increased incorporation of technology (e.g. computer-based order entry), increased ancillary support, and changes in residency educational programs, all factors which can significantly impact the residents' perceptions of the NF system.