Decreasing Inter-resident Conflict by Using Computer-generated On-call Schedules

Abstract

Although significant attention has been paid to the number of hours worked by residents, little consideration has been given to how the hours are assigned. This project describes an alternative to having Chief Residents manually create on-call schedules. In order to enhance objectivity and transparency, reduce perceived inequities in the process, and reduce inter-resident conflict, Harvard South Shore Psychiatry Residency Training Program experimented with a computer-generated on-call schedule. A locally written MATLAB script generated an on-call schedule for academic year (AY) 2012-2013. Measurements to assess the manual scheduling method (from AY 2011-2012) and the computer-generated method included the balance in the total number of hours assigned to individual residents; the number of call switches over two six-month periods; and survey of the residents' perception of fairness of the two scheduling methods and preferences. A retrospective analysis of the AY 2011-2012 Chief Resident-generated call schedule found a range of differences of up to 25.8% between total hours assigned to individual residents in a given year. In the AY 2012-2013 computer-generated schedule, the differences in total hours assigned were reduced to a maximum of 6.1%. There were 63% fewer call switches resulting from the computer-generated as compared to the Chief Resident-generated method. Resident survey response rate was 76%. Seventy-seven percent of resident respondents (N = 22) perceived the computer-generated method to be fairer, and 90.9% of residents preferred having a summary table of hours of call per resident. Residents perceived the computer-generated method as resulting in less inter-resident conflict. Methods for assigning duty hour schedules that are transparent, equitable, and require less Chief involvement may result in perceptions of greater fairness and less inter-resident conflict.