Long COVID patients' reconstruction of medical gaslighting discourse in online epistemic communities.

Objective: To analyse long COVID patients' reconstruction of medical gaslighting (MG) discourse in online epistemic communities in order to identify the MG types experienced and unfold root causes.

Method: The mixed-methods of corpus‑based critical discourse analysis was applied to an ad hoc corpus of 18 threads (n = 1176 posts) from two sub-communities on the social media site Reddit. Comparative concordance analyses were performed on the two sub-corpora to semantically code concordance lines as MG types. Critical discourse analysis was conducted to uncover power/ideological drivers of MG discourse.

Results: Keyness log‑likelihood statistical measures displayed higher values for the pronouns I, me, you, it, they and the noun doctor(s). KWIC (Key Word in Context) output for these items exhibited seven main MG types: scapegoating, downplay, treatment refusal, psychosomatic disorder, disbelief, denial and dismissal. Concordance lines revealed three major sites of contention - epistemic knowledge, epistemic authority, long COVID indeterminacy - where MG tactics operate through epistemic injustice, ignorance, discrimination and lack of epistemic humility.

Conclusion: MG discourse is triggered by deep-seated ideologies rather than by mere doctor-patient interaction as a face-saving strategy which faults patients as epistemic agents.

Practical Implications: Patients' social media mobilisation helps unmask MG tactics, raising providers' awareness of the need for perspective-taking to build collaborative doctor-patient relationships.