Seeking Support, Posting Your Sorrows, and Getting Misinformed in Return A codebook for a quantitative content analysis on mental health online support groups (Preprint)

Abstract

The potential of the Internet to help chronic patients to cope with their condition was present from the times when it had become clear what types of services the new device would provide. Expected changes were beneficial, many thought, for communication between patients and physicians, patients and health care institutions. Some reserve was discernible when Web 2.0 came, and increased communication from patient to patient with it.

Keeping this development in mind, the projected study is intended to find out how and why such online support groups for mental health can have negative outcomes for the people who turn there for help. It aspires to reach beyond the simple equations that communication between patients and physicians is good, while patient-to-patient communication is dangerous.

A codebook for the content analysis of Facebook online support groups has been developed and a content analysis will be conducted on bundles of posts. Three consecutive periods of one year will be studied. The sample will consist of utterances in two groups, one moderated the other unmoderated. The major analysis will bring together indications of health care shortcomings, medical errors, and holding wrong health beliefs on the one side and conditions and perceptions on the other. Aside from a few analyses comparing different bundles, individual utterances will be the unit of analysis in most cases. A bundle will be selected (in a time-stratified sample) if a group member asks a declarative or procedural knowledge question, seeks help in making a decision, or wants to be emotionally supported. The design will allow three more minor perspectives: comparing moderated and unmoderated support groups, describing the effects of the COVID-19 pandemic by separating the study period into before Covid (year 2019), Year 1 (2020) and Year 2 (2021) of the COVID-19 Pandemic.

We demonstrated the usability of the proposed systematic framework with 11 threads (61 utterances) coded independently by two coders: using Krippendorff’s alpha, the coders met intercoder reliability (α = .88, range: .73- 1).

The codebook is a rigorous and standardized method for the analysis of discussions in online support groups. For discussion and interpretation, we expect unhealthy present relations between health literacy and patient empowerment or a development towards such a state to decisively explain an output by the health care system that falls short of an optimum.