Unscheduled absences in a cohort of nurse anesthetists during a 3-year period: Statistical implications for the identification of outlier personnel.

Abstract

STUDY OBJECTIVE\nTo estimate the prevalence of unscheduled absences in a cohort of certified registered nurse anesthetists (CRNAs) over a 3-year period, for purposes of critiquing statistical review of individual providers relative to potential identification of patterns of such absences.\n\n\nDESIGN\nRetrospective, observational study.\n\n\nSETTING\nUniversity hospital.\n\n\nSUBJECTS\n99 CRNAs performing clinical assignments in the operating rooms.\n\n\nINTERVENTIONS\nNone.\n\n\nMEASUREMENTS\nCRNA daily clinical assignments and unscheduled absences were retrieved from the department s staff assignment software package. Data were extracted and analyzed to estimate the prevalence of unscheduled absences by CRNAs by day of the week, and whether each absence occurred on the workday before or after either a holiday or a personal vacation. A statistical power analysis was performed to determine the number of workdays of data required to identify outlier personnel above the 95th percentile among all CRNAs while controlling for a family-wise error rate of 5%.\n\n\nMAIN RESULTS\nThe overall incidence of unscheduled absences pooled by days was 1.7%, with small differences among days of the week, and before or after vacations. A year of data would be required to detect outliers for unscheduled absences exceeding the 95% upper confidence limit among all CRNAs. Attempting to identify patterns of absences being on specific days of the week or as related to holidays and vacations would require multiple years of data.\n\n\nCONCLUSIONS\nOR managers can detect CRNAs with excessive numbers of unscheduled absences, but at least a year of data is required. Detecting apparent patterns of absences would require multiple years of data and is thus impractical.