Physician quality report cards are everywhere: The federal government, consumer groups, professional organizations, US News and World Report, and ProPublica all produce reports on physicians’ quality of care. However, accurately assessing physician performance is difficult. The measures must be valid and reliable. This assessment requires getting enough appropriate data that are risk adjusted. If the data are good, presenting the results in a patient-friendly way can still be challenging.

The ProPublica Surgeon Scorecard illustrates just how difficult creating a good report card is. ProPublica used inpatient Medicare data for eight surgeries to rate surgeons. It measured complications as a composite of inpatient mortality and readmission rates. Although this method sounds reasonable, the strategy excluded 82 percent of the cases, missed 84 percent of complications, and was only weakly correlated with generally accepted surgical outcomes, such as the overall death rate, infections, or morbidity (Ban et al. 2016).

Researchers have found little evidence that physician report cards have a significant impact (Shi et al. 2017). In principle, report cards may push out low-quality firms, induce entry by high-quality firms, or encourage existing firms to improve quality, but the evidence is far from compelling. Most studies have found that report cards have modest impacts on referrals and market share, probably because referring physicians already steer patients to higher-quality providers.

Discussion Questions

• Do patients actually use report cards?

• What evidence can you find that report cards have improved quality?

• How could report cards improve reported outcomes?

• Does the scarcity of scientific evidence on the effectiveness of report cards matter?

• Could publication of performance data be advantageous to physicians?

• How do report cards address information asymmetries?

• Would reducing information asymmetries guarantee better markets?

• Does it matter whether report cards are produced by governments or private organizations?

• Why are a few patient switches enough to influence market outcomes?