ReAdmissions and Quality

Everyone wants lower hospital readmission rates, because readmissions cost money and because many are thought to be, and might be, avoidable.  However, ascertaining which ones truly were avoidable is not that simple and punishing hospitals for events out of their control is unfair and foolish.  An article in Health Affairs discusses using readmission rates as a quality metric.   (HA Article)   Using data from the Center for Medicare Services Hospital Compare program, the authors analyzed the comparison between readmissions and other quality measures.  In just looking at readmission rates from 2009 to 2011, the researchers found that lower performing hospitals in 2009 tended to improve and higher-performing ones to get worse, a classic example of regression to the mean, or just random variation in performance.  Punishing hospitals for random variation is ludicrous.  Similarly, the researchers uncovered very weak correlations, if any, between hospital readmission rates and mortality or process quality measures like delivery of certain items of care for certain conditions, like heart attacks, heart failure and pneumonia.  Readmissions are a complex phenomenon, likely due not just to hospital actions, but patient behavior and care in other settings.  While it is good to attempt to limit them, strategies to do so need to address this complexity, not just pick on hospitals.  And we have repeatedly observed that if you want to do something fair about readmissions, do a clinical review of each one and don’t pay for the one’s that were avoidable due to poor care in the original hospitalization.