Changes in Care or Changes in Coding: Unintended Consequences of the Federal Pay-For-Performance Programs?
Changes in Care or Changes in Coding: Unintended Consequences of the Federal Pay-For-Performance Programs?
Monday, June 13, 2016: 8:50 AM
G65 (Huntsman Hall)
Previous work has shown that mandatory public reporting of 30-day mortality and readmission rate of pneumonia led some hospitals to recode pneumonia to clinically similar diagnoses. In the advent of new financial incentives that penalize hospitals for poor performance on readmissions and mortality for pneumonia, there is significant concern that this re-coding may be amplified. However, the degree to which these national pay-for-performance programs have changed hospital coding practices for pneumonia is currently unknown. Therefore, we examined whether the introduction of the Hospital Readmission Reduction Program (HRRP) and the Hospital Value-Based Purchasing (VBP) program shifted documentation of patients with pneumonia to potentially avoid being eligible for penalty. Using 100% inpatient Medicare claims data from 2006 to 2013, we identified all patients admitted to U.S. acute care hospitals with primary diagnoses of pneumonia or secondary diagnoses of pneumonia if accompanied by a principal diagnosis of sepsis or respiratory failure. We first evaluated trends of primary pneumonia admissions over time and explored the possibility of substitution occurring between related diagnoses after the introduction of the penalty programs. We then used a difference-in-trends estimator, using critical access hospitals (who are not subject to financial incentives) as a control, to test whether the introduction of pay-for-performance had an influence in diagnostic coding for pneumonia admissions. Our results suggest that the introduction of HRRP and VBP was associated with a decrease in patients admitted with primary diagnoses of pneumonia relative to patients admitted to critical access hospitals. In the hospitals participating in these programs, we also find evidence of increased admissions of patients with a principal diagnosis of sepsis or respiratory failure and secondary diagnosis of pneumonia. These findings raise concern about the possibility of gaming by U.S. hospitals to avoid incurring federal penalties.