Purpose: The impact of image is widely investigated in various research fields. However, its effect in online health communities is rarely studied. In this research, the authors develop a theoretical model to assess the impact of physicians' image on patients' choices in online health communities.Design/methodology/approach: The authors developed a web crawler based on R language program to collect more than 40,000 physicians' images and other related information from their homepages in Haodf.com–a leading online health community in China. The features of physicians' images are computed by Face++ Application Programming Interface (API) through the following variables: beauty, smile and skin status.Findings: The empirical results derive the following findings: (1) physician's beauty or physical attractiveness has no significant effect on patients’ choice; (2) Smile has a positive effect on patients’ choices; (3) Physician's skin status also positively affects patients' choices; (4) Physician's professional capital strengthens the effect of beauty, smile and skin status on patients' choices; (5) Beauty and skin status are the substitutes for each other, and smile and skin status are the substitutes for each other too.Research limitations/implications: Also, this study provides implications for both physicians and online health community platform managers.Originality/value: This study provides new evidence in understanding the impact of physician's online image and contributes to the literature on signaling theory, impression management theory and patients' choices.

Purpose: Online medical teams (OMTs) are gaining popularity as a new form of online health service to provide patients with prompt and guaranteed treatment. While the effective development of an OMT depends on physicians’ active participation, there is insufficient research on how a doctor gains from the OMT, especially from the multilevel and cross-level perspectives. In attempting to narrow this knowledge gap, the authors hypothesize multilevel and cross-level professional capital determinants of physicians’ performance in online health-care communities (OHCs) through the lens of social exchange theory. Design/methodology/approach: This study develops a cross-level model to explain the effects of individual and team professional capital on physicians’ performance. To test the research model and hypotheses, the authors leverage data of 10,398 physicians engaged in 2,611 popular OMTs in China in conjunction with the hierarchical linear model approach. Findings: The results indicated that physicians’ status capital (SC) and decisional capital (DC) are positively related to their performance. The SC and DC of an OMT not only increase physicians’ performance but also indirectly strengthen the positive effect of physicians’ SC on their performance. In contrast, OMTs’ SC and DC lessen the importance of physicians’ DC in promoting their performance. Originality/value: By studying the mechanism between professional capital and physicians’ performance, this study provides several contributions to theory and practice. Specifically, this study contributes to the extant professional capital research by uncovering the influencing pathways of professional capital on physicians’ performance from a cross-level perspective. These findings suggest physicians pay close attention to the strength and mechanism of OMTs’ professional capital in improving their online performance.

Purpose: The purpose of this paper is to explore the influence of a physician’s prosocial behavior on a patient's choices in the online health community (OHC) context. Moreover, the authors explore how such effects differ across different online word-of-mouth (WOM) and professional titles.Design/methodology/approach: Guided by the motivation, opportunity and ability (MOA) framework, this paper develops hypotheses and an econometric model. Then this paper used spline regression to test hypotheses on 6,204 physicians at The Good Doctor (www.Haodf.com), which is one of the largest Chinese OHCs. The authors conducted the propensity score matching and difference-in-difference method (PSM-DID) to address the concern about the bias caused by possible endogeneity concerns.Findings: The authors’ results show that a physician’s prosocial behavior improves a patient's choice only when the strength of a physician’s prosocial behavior is below the tipping point. In addition, the influence of a physician’s prosocial behavior is heterogeneous for physicians with different online WOM and professional titles. For physicians with higher online WOM, the effect of a physician's prosocial behaviors on a patient's choice is positive, while for physicians with lower online WOM, a physician’s prosocial behavior has no impact on a patient’s choice. For physicians with higher professional titles, the quantity of a physician’s prosocial behavior has a positive impact on a patient’s choice, while for physicians with lower professional titles the quality of a physician’s prosocial behavior has a positive impact on a patient’s choice.Originality/value: This study contributes new knowledge and provides new perspectives to study a patient's choice by addressing the importance of physician's prosocial behavior. With the effort of explicitly explaining the complex mechanisms, this study encourages physicians' engagement in a physician’s prosocial behavior and gives some implications on how to perform the behaviors strategically.