Hospital care may not affect the risk of readmission

Background: Thirty-day readmissions have become a focal point for reducing health care spending, because they are viewed as a marker of the quality of hospital care. However, if increased time in the hospital is associated with better care, attempts to shorten length of stay (LOS) may result in increased rates of readmission. As such, we sought to explore the association of an incremental added day in LOS with the rate of readmission. Methods: We examined the rate of readmission at 30 and 120 days for 4151 patients admitted to a general internal medicine unit between July 2004 and March 2006. We used binary logistic regression to examine the relationship between an incremental added day in LOS and the probability of readmission. Results: Readmission rates were 8.7% at 30 days and 21.0% at 120 days, respectively. After controlling for demographic characteristics and severity of illness, we found that the probability of readmission varied little for an incremental added day in LOS. Conclusions: Our findings suggest that more hospital care may not affect the likelihood of readmission and thus denying payment for readmission may be unwarranted.