Public Reporting for Percutaneous Coronary Interventions in New York State

Background:

Prior studies of public reporting on risk‐adjusted mortality for coronary artery bypass grafting (CABG) have found that public reports have strong predictive validity and likely encourage poor‐quality surgeons to leave practice. However, except for CABG surgery, there are few data on the predictive accuracy of public reports or their impact on providers’ practice. We used data from New York State on percutaneous coronary interventions (PCIs) to address 3 questions. (1) What is the predictive accuracy of public reports for PCIs? (2) What is their impact on market share? (3) Is report performance associated with decisions to leave practice?

Methods:

We examined quality performance by hospitals (and cardiologists), as measured by publicly reported risk‐adjusted mortality rates (RAMRs) for nonemergent PCIs performed in New York State between 1998 and 2007. For hospitals (and cardiologists) in each performance quartile, we estimated: (1) the average risk‐adjusted mortality rate after report publication, (2) the change in market share from prerelease to postrelease year, and (3) the proportion of physicians leaving practice in the postrelease year.

Results:

Between 1998 and 2007, the New York State public reports included data on 351 cardiologists who performed nonemergent PCIs at 49 hospitals. Patients who picked a hospital in the highest quartile of performance in the most recent report did not have a lower chance of dying than those who picked a hospital in the lower 3 quartiles (RAMRs of 0.61, 0.59, 0.58, 0.71; P > 0.05). Results were similar for cardiologists (RAMRs from top to bottom performance quartile: 0.57, 0.59, 0.57, 0.74; P > 0.05). Performance ranking was not associated with a change in market share for hospitals or for physicians (all P > 0.05). There was no association between report performance, and decisions to stop practicing in New York after report publication (4% in top and bottom quartiles; P > 0.05).

Conclusions:

Public reporting on percutaneous interventions in New York State provides information that, even if used, would not significantly help patients decrease their risk of dying after a nonemergent percutaneous coronary intervention. Public reporting appears to have had no effect on hospital or physician market share, or physicians’ decisions to leave practice. The utility of public reporting may differ substantially for different procedures.

Disclosures:

L. Chen ‐ none; A. Epstein ‐ none