Medicine in the World of the Smart Phone

Some nine years ago as Obamacare was being started and electronic health records were being promulgated I suggested that the patient "own" their health record not the physician and that an entity like then then benevolent Google could enable it. Google eventually dropped the idea and became less benevolent. But the idea still as merit. That is a patient owned and provider accessible data set not owned or operated by the Government. Like the Internet could have been until the FCC came along. It seems the Government besides being incompetent, remember Obamacare web sites, is also invasive, think NSA and FBI.

Now the NY Times reports Apple and others using smart phones for clinical data and trials. They note:

People often learn about new research studies through in-person conversations with their doctors. But not only did this study, run by Stanford University, use a smartphone to recruit consumers, it was financed by Apple. And it involved using an app on the Apple Watch to try to identify irregular heart rhythms. Intrigued, Mr. ..., who already owned an Apple Watch, registered for the heart study right away. Then he took to Twitter to encourage others to do likewise — suggesting that it was part of a breakthrough in health care. “It’s not inconceivable, by the time I graduate from medical school,” Mr. ... said, “that the entire practice of medicine can be revolutionized by technology.”

Unfortunately using this approach can introduce massive biases in any clinical data. Clinical trials have evolved to test hypotheses and must be highly regulated. The FDA does extensive audits to ensure compliance and accuracy, and just casual consumer testing is rant with risks of distorted data. There are no standards for procedures for such trials and no acceptance of the results. 

There is potential....but. Take an EKG, or ECG if you will. I admit my age. To determine a cardiac problem one needs all twelve leads taken under controlled conditions. A single lead may tell you something but hardly enough. In fact biases from a single lead could bias a whole study.

Medical data is not like other data. It is patient dependent and patient history dependent. We still need a trustworthy, secure, capable patient centered health record.