Concerns over patient safety in hospitals have focussed attention on the causes and epidemiology of adverse events and complications. Prior research suggests that hospital working conditions may impact the rate of complications. One situation where the discrepancy between patient needs and care provided may be accentuated is during weekends and holidays, when reduced hospital resources may lead to reduced capacity for patient care. Another situation, in teaching hospitals, occurs in the month of July, when new house officers assume the care of patients, and may lack the experience for recognizing and preventing complications. These hypotheses have not been tested before due to the absence of a reliable tool to look at rare adverse events with sufficient power of discrimination. We will use AHRQ's Patient Safety Indicators to look at three years of hospital discharges contained in HCUP's State Inpatient Databases from several states distributed over several geographical regions. We will use expert consultants to assist and advise on which indicators to use and how to best use them. We will model and test the hypotheses in two different ways. First, we will compare rates of some complications for which the date of the complication and the date of risk exposure are easily identifiable. Sub-analyses will look at the rates in specific groups of patients, such as patients admitted through the emergency room, or in patients with specific procedures. Second, we will perform patient-level logistic regressions for individually selected PSIs to assess the impact of day of the week on the likelihood of a PSI, controlling for patient and hospital characteristics. A similar analysis will be performed to test for the July phenomenon. This study has implications for how hospitals organize their workforce on weekends, and will inform the debate over whether hospitals should become complete 7-day organizations. If a July effect is found, it will have implications for how teaching hospitals organize their work when new residents begin their training.
