Negotiating trust on health-related forums
Online health forums are where many patients with chronic illness — particularly contested conditions like ME/CFS, POTS, and dysautonomia — do much of their information gathering. They turn to communities of people with similar experiences because the medical system has often failed to provide adequate information, explanation, or validation. The information they find there shapes their understanding of their conditions, their expectations about what symptoms mean, and their decisions about management. What determines which information they trust? A 2016 qualitative study by Brady, Segar, and Sanders in Health Expectations examined this question directly, and the answer is not reassuring: trust on health forums is primarily social, not scientific. The signals people use to decide what to believe have more to do with social resemblance and community membership than with the quality of evidence behind the claims.
The Study: Qualitative Analysis of Forum Trust Dynamics
Brady and colleagues conducted a qualitative analysis of online health forum participation, examining how members evaluated the credibility of information from other users. The research used thematic analysis of forum interactions and user interviews to identify the trust signals that actually determined whether forum members accepted or rejected health information. The methodology is appropriate for this question: trust dynamics in online communities are not straightforwardly measurable through quantitative methods but reveal themselves in patterns of engagement, response, and belief that qualitative analysis is well-suited to identify.
The patient population studied had chronic conditions requiring ongoing management and active information seeking — the profile shared by most people using health forums for substantive medical information, including those with ME/CFS, dysautonomia, and overlapping conditions. The forum environment studied reflected the typical dynamics of health communities: a mix of personal experience sharing, management advice, research discussion, and emotional support.
Social Similarity as the Primary Trust Signal
The core finding was that identity similarity — whether the poster seemed to share the reader's condition, illness trajectory, demographic situation, and experiential context — was the primary driver of credibility judgments. Posters who seemed relatable were trusted. Their information was accepted, acted on, and shared. Posters who seemed different — whether through describing a different illness course, a different response to treatment, a different demographic background, or a different level of severity — were treated with skepticism, regardless of the quality of their content.
The mechanism identified by Brady and colleagues is dual-process in character. Information credibility is evaluated through two routes: the cognitive route, which involves assessing the quality of evidence and argument, and the social route, which involves assessing whether the source belongs to the same in-group and shares the same experiential context. In complex, contested conditions where the cognitive route requires significant technical knowledge to navigate effectively, the social route tends to dominate. Evaluating whether a claim about dysautonomia physiology is consistent with the research literature requires substantial domain knowledge that most forum participants do not have. Evaluating whether the person making the claim seems like someone with a similar experience is much easier, and it determines trust in the absence of the technical tools needed for cognitive evaluation.
How Social Trust Selection Creates Information Distortion
The consequence of social-route trust selection is systematic distortion of the information environment. Information that confirms existing community beliefs circulates widely because it comes from voices the community has already decided to trust — insiders with similar experiences telling the community what it already believes. Information that challenges existing beliefs, introduces contradictory evidence, or comes from sources that seem demographically or experientially distant gets filtered out through the same mechanism. The credibility judgment operates before the content is evaluated, and the credibility judgment is made on social rather than epistemic grounds.
This produces a feedback loop. The community's existing shared model of the illness determines what kinds of voices are seen as credible. Credible voices confirm and extend the existing model. The model strengthens. Challenges to it are filtered out as coming from non-credible sources — people who clearly don't understand because their experience is different. Over time, the information environment becomes progressively more self-confirming, and the gap between community consensus and research literature can widen without community members being able to detect it, because the detection mechanism itself is compromised by the same social filtering process.
The Specific Vulnerability of Contested Illness Communities
The Brady findings are particularly relevant for ME/CFS and dysautonomia communities because the conditions are contested. Groenevelt and de Boer's 2023 analysis of ME/CFS communities specifically documented how the experience of systematic medical disbelief shapes how patients present their illness narratives, which in turn shapes what gets amplified by community validation processes. A community that has been disbelieved by medicine has strong motivations to reject information from sources associated with that disbelief — including clinicians, researchers, or fellow patients who describe improvement — regardless of the quality of what those sources are saying.
The social filtering mechanism documented by Brady compounds this. When a clinician or researcher challenges a community belief, they fail the identity-similarity test immediately: they are not a patient, their experience is different, they cannot truly understand. Their credibility is low on social grounds before any evaluation of their argument's quality occurs. Meanwhile, community members who confirm existing beliefs — particularly those who describe severe and persistent illness, who frame improvement as suspicious, who validate the shared narrative — have high social credibility by virtue of their membership and resonance. The result is that the community's epistemic environment is structurally tilted toward confirming its existing models, regardless of whether those models are accurate.
Using Online Communities Without Being Misled by Them
The Brady findings do not argue that online health communities are useless or that patients should avoid them. For people with poorly understood conditions who face medical systems that are inadequate for their needs, community provides real value: emotional support, practical management experience, and the experience of being understood by others who share the condition. These are not trivial benefits, and no research on information quality challenges their importance.
What the findings argue for is epistemic awareness when using forums for medical information specifically. The trust signals that feel reliable on a forum — this person seems like me, their experience resonates, they are an established community member — are not reliable guides to whether the medical information they are providing is accurate. The two things are independent. Research on POTS content quality on social media platforms confirms this independence: creator credibility to an audience is uncorrelated with information quality as assessed by clinical standards. The person who feels most credible by social-route assessment may be providing the least accurate information by cognitive-route assessment. The mechanism that makes someone feel like a trustworthy insider is independent of whether what they are saying is clinically correct.