Digitizing Doctor Demand: The Impact of Online Reviews on Doctor Choice

Background: Consumers, facing a physician choice, have to deal with uncertainty regarding physician quality and may rely upon imperfect signals—such recommendation from peers—to resolve such information asymmetry. With the onset of social media revolution, consumers seeking a physician recommendation, can also learn from consumer review websites, such as ZocDoc.com, that enable large-scale distribution of information from countless other consumers, and can help resolve information asymmetries among a much broader peer group than has been traditionally possible. It is, a priori, unclear whether consumer review websites will significantly affect markets for credence goods, given that a consumer is unfamiliar with the intricacies of medical knowledge, and even ex post a consumer can hardly determine the extent of the treatment required ex ante.

Methods: Our study is based on ZocDoc.com—a unique website that integrates patient reviews, and appointment scheduling for physicians on one platform. Using ZocDoc we construct a novel data set consisting of all reviews written for primary care physicians in Manhattan, New York. We then pair these reviews with data on appointments that are booked through ZocDoc, during February-May, 2013. To investigate the impact of ZocDoc, we first show that changes in a physician's rating are correlated with change in the likelihood of filling an appointment, controlling for physician characteristics, appointment characteristics, and zip level demographic information. In addition to individual reviews, ZocDoc presents the average rating for each physician, rounded to the nearest half-star. To support the claim that ZocDoc has a causal impact on appointment filling likelihood we take advantage of this feature and implement a regression discontinuity design around the rounding thresholds. We then implement the McCrary (2008) density test to rule out the possibility that gaming is biasing the results.

Results: We present empirical evidence for the impact of patient reviews on consumers’ physician choices. Our data suggest that patient reviews are becoming an important source of reputation for physicians. About 25% of New York primary care physicians are now listed on ZocDoc, and 84% of them have at least 5 reviews. We present the following key findings: (1) a half-star increase in ZocDoc rating, on a scale of 1 to 5 stars,  leads to a 10% increase in the likelihood, at the mean, that a physician will fill an appointment, (2) physicians with higher number of reviews also have a higher likelihood of filling an appointment, (3) when compared to males, female physicians have a higher likelihood of filling an appointment, (4) on the supply side, we find that physicians update the quantity of appointments supplied on ZocDoc in response to an increase in the likelihood of filling an appointment, and (5) patient ratings capture patient's visit experience, and educational differences of physicians.

Conclusion: Overall, this paper presents evidence that consumers use ZocDoc.com to learn more about physician quality. The paper shows that consumer reviews can be used to solve the information asymmetry in credence goods markets.