The relationship between Medicare's process of care quality measures and mortality

Using Medicare inpatient claims and Hospital Compare process of care quality data from the period 2004-2006, we estimate two model specifications to test for the presence of correlational and causal relationships between hospital process of care performance measures and risk-adjusted (RA) 30-day mortality for heart attack, heart failure, and pneumonia. Our analysis indicates that while Hospital Compare process performance measures are correlated with 30-day mortality for each diagnosis, after we account for unobserved heterogeneity, process of care performance is no longer associated with mortality for any diagnosis. This suggests that the relationship between hospital-level process of care performance and mortality is not causal. Implications for pay-for-performance are discussed.