The role of trust in eHealth websites: the enduring power of impartiality

Background: Within a year period, 72% of internet users seek health information online (Pew Research, 2014). The internet has provided a valuable alternative to physical contact with health practitioners by increasing the availability and access of information. However, with this increased access come concerns about the quality of the information source that consumers choose to trust and act upon. Understanding what causes users to trust health websites is important to understand how it leads to behavioural changed. We have studied eHealth trust during early internet adoption (Briggs et al. 2000) and from the introduction of web 2.0 (Harris et al., 2011), however more attention on the role of advice from patient experiences (PEx) (Sillence et al., 2013) and privacy concerns (Kim, 2014) on health website trust formation is needed. Objective: The current study aimed to (1) update our existing trust questionnaire to incorporate greater consideration of PEx and the role of privacy concerns, (2) assess the factorial structure of the trust questionnaire and (2) model how these factors predicted trust, information corroboration, users’ threat perception, coping assessment and readiness to act on the advice contained within health-related websites. Methods: Participants were asked to recall a site they had used previously to search for health information. They were then asked to complete the trust questionnaire in relation to the website and answered a number of questions assessing their threat and coping appraisal, information corroboration and checking. 1123 participants were recruited from an online panel with 56% from the USA and 44% of the UK. An online panel was chosen to recruit a demographic similar to that recruited in our existing studies as our previous recruitment platform no longer existed. The factorial structure of the eHealth web trust scale was assessed using Exploratory Factor Analysis (EFA) and then modelled how those factors predicted the outcome measures using Structural Equation Modelling (SEM). Results: EFA revealed four factors: PEx, credibility and impartiality, privacy and familiarity. In the SEM, of these four factors, only credibility and impartiality had a direct effect on trust. The role of PEx had an indirect effect through information corroboration suggesting that whilst users may not trust PEx content, corroborating such information through other sources heightens users trust. Privacy and familiarity only had indirect effects on trust. The final model accounted for the trust markers, trust in health information, information corroboration, coping and intention to act. This model achieved a good fit (GFI (.95), AGFI (.93), RMSEA (.50) and CFI (.97)) and accounted for 65% of the variance in trust, 27% of the variance in coping and 41% of the variance in intention to act. Conclusion: In line with our existing studies (Harris et al., 2011; Sillence et al., 2013), impartiality is the only trust marker to remain a stable predictor of trust. Previous PEx work (Sillence et al., 2013) has found the content to negatively influence trust but our findings suggest it can have a positive influence on trust provided that users corroborate the information through additional sources.