mHealth: Do health apps really work?

Anindya Ghosh writes:

In partnership with a major mHealth app platform in Asia, we designed and implemented a large-scale randomized field experiment based on detailed patient lifestyle activities (e.g., steps walked, exercise time and calories spent, sleeping patterns, and food quality and quantity) and blood glucose values from chronic diabetes patients over a 15-month time frame. The randomization involved some patients having access to the mHealth app, some having access to web-based version of the app, and the rest (the control group) not having access to any of these apps or devices.

The adoption of the mHealth app led to an improvement in both short-term metrics (such as a reduction in patients’ blood glucose and glycated hemoglobin levels) and longer-term metrics (such as a reduction in hospital visits and medical expenses). Patients who adopted the mHealth app undertook higher levels of exercise, consumed healthier food with lower calories, walked more steps, and slept longer on a daily basis.

This is a fascinating study – one because it is a randomised control study with the proper identified control group, and they have demonstrated an improvement in the metrics. I don’t have the access to the full paper, but the takeaway from the HBR linked article is fine.

Here’s an introguing takeaway:

That said, our randomized experiments demonstrated that compared to generic messages, personalized messages were more effective in reducing in-person doctor visits and replacing them with telehealth services. Post-experimental surveys of the experimental subjects revealed that the accuracy of these personalized messages, in fact, made patients comfortable with adopting telehealth services deployed by the platform. Thus, they were substituting their offline physician interactions with online ones, reducing their overall medical expenses. This was a silver lining of personalization.

It is also a validation of Tele-health done right. I think the idea of telemedicine is to supplement the routine visits to one-off interaction through the continuous monitoring of variables. I am not sure if Telehealth is being pushed to the sidelines due to “zoom fatigue”. It is a useless construct.

The more troubling suggestion is:

Third, mHealth devices and apps could provide health insurance companies with an opportunity to personalize premiums. They could allow them to reward consumers who make the effort to exercise more often, eat healthier, and sleep longer with lower insurance premiums. 

I am sure this will raise significant questions and although implied, users don’t like to held accountable for their actions or lifestyles. They make choices out of peer pressure, and possibly out of conformity, which will eventually have consequences leading them to seek healthcare. Should their premiums be decided on this? Although data federation may answer some potential issues but I don’t see users opting in any time soon