Insights Into the Impact of Online Physician Reviews on Patients’ Decision Making: Randomized Experiment

BACKGROUND Physician-rating websites combine public reporting with social networking and offer an attractive means by which users can provide feedback on their physician and obtain information about other patients' satisfaction and experiences. However, research on how users evaluate information on these portals is still scarce and only little knowledge is available about the potential influence of physician reviews on a patient's choice. OBJECTIVE Starting from the perspective of prospective patients, this paper sets out to explore how certain characteristics of physician reviews affect the evaluation of the review and users' attitudes toward the rated physician. We propose a model that relates review style and review number to constructs of review acceptance and check it with a Web-based experiment. METHODS We employed a randomized 2x2 between-subject, factorial experiment manipulating the style of a physician review (factual vs emotional) and the number of reviews for a certain physician (low vs high) to test our hypotheses. A total of 168 participants were presented with a Web-based questionnaire containing a short description of a dentist search scenario and the manipulated reviews for a fictitious dental physician. To investigate the proposed hypotheses, we carried out moderated regression analyses and a moderated mediation analysis using the PROCESS macro 2.11 for SPSS version 22. RESULTS Our analyses indicated that a higher number of reviews resulted in a more positive attitude toward the rated physician. The results of the regression model for attitude toward the physician suggest a positive main effect of the number of reviews (mean [low] 3.73, standard error [SE] 0.13, mean [high] 4.15, SE 0.13). We also observed an interaction effect with the style of the review—if the physician received only a few reviews, fact-oriented reviews (mean 4.09, SE 0.19) induced a more favorable attitude toward the physician compared to emotional reviews (mean 3.44, SE 0.19), but there was no such effect when the physician received many reviews. Furthermore, we found that review style also affected the perceived expertise of the reviewer. Fact-oriented reviews (mean 3.90, SE 0.13) lead to a higher perception of reviewer expertise compared to emotional reviews (mean 3.19, SE 0.13). However, this did not transfer to the attitude toward the physician. A similar effect of review style and number on the perceived credibility of the review was observed. While no differences between emotional and factual style were found if the physician received many reviews, a low number of reviews received lead to a significant difference in the perceived credibility, indicating that emotional reviews were rated less positively (mean 3.52, SE 0.18) compared to fact-oriented reviews (mean 4.15, SE 0.17). Our analyses also showed that perceived credibility of the review fully mediated the observed interaction effect on attitude toward the physician. CONCLUSIONS Physician-rating websites are an interesting new source of information about the quality of health care from the patient's perspective. This paper makes a unique contribution to an understudied area of research by providing some insights into how people evaluate online reviews of individual doctors. Information attributes, such as review style and review number, have an impact on the evaluation of the review and on the patient's attitude toward the rated doctor. Further research is necessary to improve our understanding of the influence of such rating sites on the patient's choice of a physician.