Associations between patterns of online engagement with an online health community and changes in patient activation and health care utilisation: longitudinal web-based survey

Background: Online health communities (OHC) are widely used in the UK. They allow patients to share experiences of living with disease and self-management, but it is unknown whether use improves health outcomes. Objectives: The aims of this study were: 1) to determine change in activation over three months in new users of an OHC, 2) to describe patterns of engagement with an OHC, 3) to examine whether patient characteristics at baseline were associated with subsequent patterns of engagement, and 4) to determine if patterns of engagement during the three months were associated with changes in patient activation, healthcare utilisation and health status. Methods: Active new OHC users on HealthUnlocked (HU) were surveyed to measure demographics, levels of “patient activation” (describing a person’s confidence in managing their own health; scale 0-100 with four categories), healthcare utilisation and health status using a web-based survey at baseline and three months. Patient activation at baseline and three months was compared (Aim 1). Alongside, for a sample of HU users and survey responders, daily OHC website usage data was automatically captured indicating whether or not a user interacted with HU. This was used to identify clusters of engagement with HU (Aim 2). For survey responders, baseline characteristics, patient activation, healthcare utilisation and health status was compared at baseline and three months, overall and between engagement clusters using t-tests and chi-squared tests (Aims 3 & 4). Results: In 329 people who completed both the baseline and follow-up surveys, baseline activation was most frequently ‘level 3’, described as “taking action but still lacking confidence”. At follow-up a change of 2.6 points was seen, with the greatest change seen in those at lowest baseline activation levels. Four clusters of engagement were identified - low, medium, high and very high - who were active on HU for a mean of 4, 12, 29 and 59 days, respectively. Survey responders were more commonly high or very high engagers. Baseline activation was highest in low and very high engagers. Overall activation increased over time in all engagement groups. Very high engagers had the greatest improvement in activation (5 points), although the average change was not above what is considered clinically meaningful for any group. Accident and emergency visits at follow-up were lower with higher engagement, though this trend was not seen for other healthcare utilisation measures. There was no change in health status at three months. Conclusions: This observational study provides some insight into how patterns of engagement with OHCs are associated with changes in patient activation, healthcare utilisation and health status. Over three months overall the change in activation was not clinically significant, there were some indications that OHCs may be of benefit to particular groups. However, the study limitations prevent firm conclusions about causal relationships.