JMIR Publications

ABSTRACT

Patients are increasingly using online consumer ratings to help select high performing physician. While ratings websites provide consumer ratings across a number of physician-based (e.g. trust in provider, how well he/she explains condition, spending sufficient time) and office-based metrics, it is unclear if consumers are able to distinguish between different dimensions of perceived quality, since these metrics have not been psychometrically validated. In this study of consumer reviews of 830,000 physicians in the Healthgrades website, we show that ratings of physician-based subdimensions of care are extremely highly correlated (r=0.96–0.98), demonstrating that patients rarely differentiate between the various rated characteristics of a physician. This is in contrast to office-based subdimensions of care, which show less correlation with each other and with physician-based metrics. These findings call into question the utility of subdimensions of physician ratings; either the vast majority of physicians are all good or all bad, or consumers are unable to differentiate between the measured characteristics of physicians using Likert scale metrics.