How patients adopt physician services in online medical communities: a polynomial regression with response surface analysis

Purpose Information cues in online medical communities (OMCs) critically shape patients’ physician service adoption, yet their joined effects remain underexplored. Based on the elaboration likelihood model (ELM) and the service quality model, this study develops an extended adoption framework to examine (1) the relative effects of central-route cues (information quality and interaction quality) vs peripheral-route cues (electronic word-of-mouth, eWOM) and (2) their joint effects under balanced and imbalanced conditions. Design/methodology/approach Using archival data from Haodf.com, this study applies polynomial regression with response surface analysis (RSA) to capture the complex relationships among adoption cues. Findings Results show that eWOM exerts a significantly stronger positive effect on service adoption than interaction quality. When service quality and eWOM are balanced, adoption is higher when both are high than when both are low. Under imbalance, adoption is higher in the low service quality – high eWOM condition than in the high service quality – low eWOM condition. Originality/value This study advances understanding of patients’ service adoption in OMCs by revealing the asymmetric and joint effects of central- and peripheral-route cues, offering actionable implications for platform governance and physician service strategies.