Posted by: AllinHIT
PHR, PHRs and patient engagement, self diagnosis
Some years ago, while I was a consultant for Quest Diagnostics, I remember talking to various physician groups about utilizing Google Health’s PHR.
The ROI argument centered around the idea that the program would reduce patient phone calls and staff call backs, increase patient engagement, and could increase business (if you are an OBGYN, reporting normal paps via the PHR is a selling point). Quest had the ability to automatically send normal lab results to the PHR, or a physician could control the release of results.
Regardless of any selling point, it was completely overshadowed by one concern: a patient self-diagnosing results. Fast forward to now, where patients frequently Google any and all symptoms, conditions, and lab result analytics, and it seems the physician fears were warranted. It has officially been proven that medical apps can be more of a danger in self-diagnosing!
The University of Pittsburgh Medical Center recently evaluated four medical smartphone apps developed to evaluate whether skin moles were cancerous. The study, published in JAMA Dermatology a few weeks ago, had dismal results. The best performing app was 98% accurate, however the other three were found to perform far worse than imagined by the participating dermatologists. Why did the best performing app do so well, while the other three didn’t? That was my immediate question. Then after reading more of the study, my answer came. It was because the best app involved doctors forwarding the skin mole images to their peers for review!
This study is a perfect example of how using ONLY apps for self diagnosis is a danger to patients. Patients must realize these tools are not for determining conditions, nor any routine of care. However, the onus should not be only on the patient. The company developing the app should always include a disclaimer to be acknowledged by the user prior to use and again, when reporting the results! This is being done on most apps already. However, I think the message should be stronger and re-enforced through its use.
Lastly — I hate to say it –but the app world needs some monitoring for accuracy, such as only becoming available on the market if they meet a minimum gauge (e.g. 80% accuracy). Without this accuracy model, apps diagnosing something as serious as cancer will continue to scare physicians, and worse, negatively effect patient care!