Social Media: Does It Really Change Patient Behavior?

Thanks to the technologic allure of iPhones replacing stethoscopes, apps substituting for doctors and electronic information substituting for having to actually talk to patients, the thoroughly modern Disease Management Care Blog is all about medical-social media. 

Think Facebook for the flu.  Twitter for tinnitus. Egads, listen to the typical consultant, pundit or futurist and it's easy to believe that we're on the verge of a silicon-based health care revolution

But then reality intrudes and some skeptic somewhere always asks about the bang for the buck, the juice for the squeeze, the return for the investment. It's a good question.

For something of an answer, consider the results appearing in a recently published randomized clinical trial.  The actual condition in question is going unmentioned for now, so that the DMCB can better focus on the issue of patient engagement

Over a 4 month period, "at risk persons" were recruited for a clinical research trial with on-line ads (Facebook banners, Craigslist, for example) as well as announcements in community settings and venues.  Once persons met the usual inclusion criteria and had a unique Facebook account, they were randomly assigned to one of two treatment arms.

One treatment arm used a closed Facebook group to coach persons about their at risk condition.  The other treatment arm similarly used Facebook to coach persons about general health improvement.  Lay "Peer Leaders," who were given a three hour training session on "epidemiology of the condition or general health subjects and ways of using Facebook to discuss health and stigmatizing topics," were assigned to lead the groups.

Peer Leaders attempted to reach out to their assigned group persons with messaging, chats and wall posts.  Once the link was established, the relationship in the intervention group included communication about prevention and treatment of the condition. At the end of 1, 2 and three months of the study, participants completed a variety of surveys.

Results?

57 individuals were in the control general health group and 55 were in the condition coaching group.  According to the surveys, intervention patients were ultimately statistically significantly more likely to agree to condition testing (44%) than the control patients (20%).  Because there were few participants, the modest decrease in actual tests or risk behaviors were not statistically meaningful.

The DMCB's take:

While this was a small study, this is the first time that the DMCB has seen reasonable proof that social media by itself can move the behavior needle.  On the other hand, this did not result in a patient engagement stampede toward better care or hard clinical outcomes.  A majority of participants (56%) did not appear to benefit.  Nonetheless, the results do support the inclusion of Facebook-style closed group social media in the suite of population health management services.

That being said, the condition at risk was HIV and study population was men who have sex with men ("MSMs").  It doesn't necessarily follow that what would work in this community of persons would necessarily be transferrable to other conditions, such as diabetes.  The DMCB doesn't think that's really true and finds it credible that 112 persons with diabetes or hypertension would probably achieve the same kind of results (A1c testing or home blood pressure monitoring) in a similarly tailored Facebook closed group.

Here's the study.

Let the research continue!