“Jerk” or “Judgemental”? Patient Perceptions of Male versus Female Physicians in Online Reviews

We analyze patient reviews (posted online) of male and female physicians with respect to numerical ratings and language use. We find that females tend to receive less favorable numerical ratings overall, and that there seems to be higher variance in the sentiment of words used in reviews of female vs. those of male physicians. Introduction and Motivation Individuals are increasingly turning to the web to gather information relevant to their healthcare. An important example of this includes online reviews of physicians: a relatively recent survey (Fox and Duggan 2013) found that 72% of internet users have looked online for health information in the past year. And one in five of these users have looked for reviews of either particular treatments or doctors. Patient-generated reviews are especially interesting as a data source because they provide a direct, unmediated window into the patient experience. Further, these reviews may influence other individuals’ opinions of (potential) physicians (Grabner-Kräuter and Waiguny 2015), in turn affecting patient care. Indeed, Li et al. (2015) concluded via a randomized trial that exposure to negative reviews “led to a reduced willingness to use the physician’s services.” Much of the previous work on examining online reviews of physicians has been qualitative in nature (López et al. 2012; Gao et al. 2012; Kilaru et al. 2016). Following our prior work (Wallace et al. 2014; Paul, Wallace, and Dredze 2013), we adopt a more data-drive approach in this study. Our focus in this work concerns investigating differences in online reviews of male versus female physicians, both with respect to patient satisfaction (ratings) and language use. We aim to complement the relatively robust body of evidence that strongly suggests that female physicians “don’t get the credit they deserve” (Roter and Hall 2015). For instance, Hall et al. (2011) performed a meta-analysis of studies looking at patient satisfaction and concluded that female physicians “are not evaluated as highly by their patients, relative to male physicians, as one would expect based on their practice style and patients’ values.” Does the same hold in online reviews? Copyright c © 2016, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved. In the remainder of this paper, we (i) describe the dataset we use, (ii) present results from regression analyses of numerical ratings and (iii) results from a lexical analysis of review texts. We then (iv) conclude with a discussion of our results and limitations. As far as we are aware, this is the first work to explicitly explore the relationship between physician gender and attributes of online reviews regarding the care they provide.