Medical Resident Turnover and Its Association with Inpatient Mortality in Patient Discharges with a Primary Diagnosis in the Heart Disease, Cancer, or Stroke Diagnostic Groups at U.S. Teaching Hospitals, 2002
Objective: In the medical community, there is an assumption that there is a higher probability of inpatient mortality during July at U.S. teaching hospitals. This is what is known as the ‘July effect’. Our descriptive, cross-sectional study tests this assumption by examining the association of the medical resident turnover period in July with inpatient mortality in those patient discharges with a primary diagnosis in the heart disease, cancer, or stroke diagnostic groups. These three discharge diagnostic groups are heterogeneous in nature and are the leading causes of death in the United States. Methods: Our analysis was based on 4,881,434 patient discharges in the Healthcare Cost and Utilization Project (HCUP) 2002 Nationwide Inpatient Sample (NIS). Frequency distributions and proportions for inpatient mortality were calculated on weighted sample data from the NIS to describe the patient discharges. We used multiple logistic regression analyses to examine the association of medical resident turnover with month of hospital admission and inpatient mortality. These regression models were stratified by teaching status and location of the hospitals sampled, and the diagnostic groups. Comorbidities were included in the adjusted regression models and were defined as 0, 1, 2, 3 or more, where each number category denoted the number of comorbidities that a patient discharge had. Results: Overall, patient discharges from urban, non-teaching hospital had a higher proportion of death (2.67%). Patient discharges from urban, teaching hospitals with cancer as a primary diagnosis had the highest rate of inpatient mortality (6.69) among all the patients. Overall, July did not show statistically significant (p< 0.05) greater odds of death. The adjusted odds of dying in the hospital for all patient discharges from a U.S. hospital, including the adjustment for comorbidities were highest during the winter months (Nov-Mar). Of the winter months, February had the highest adjusted odds ratio estimate at 1.168 and a confidence interval of (1.134, 1.203). Conclusion: Our findings indicate that concerns about the “July effect” may be unfounded. Instead, quality improvement efforts should be directed to reduce mortality rates in the winter months. Our research study is significant because it provides new insight to the literature on the ‘July effect’, the literature on seasonal trends in inpatient mortality, and the literature on the types of patient discharges that die in the hospital.
School:Case Western Reserve University
School Location:USA - Ohio
Source Type:Master's Thesis
Keywords:medical resident turnover health services research inpatient mortality in hospital death patient safety july effect heart disease cancer stroke teaching nonteaching epidemiology
Date of Publication:05/13/2009