Warning: Apple HealthKit risks driving up medical costs

How can seeing more information about your health possibly be bad? Here's how, says the Lown Institute's Right Care Weekly newsletter:

While the app and the ability to connect to medical records may be useful, it also has potential to create more worry and unnecessary medical intervention. Depending on what information the app collects and how it presents the data to users, it's easy to imagine situations where the app's findings lead to yet more overdiagnosis (for example, of harmless variation in heart rhythms). The issue isn't that people can't be trusted with their information, it's that much of the medical data that can be collected is highly uncertain, and presenting it as important may lead people to worry or seek treatment unnecessarily.

The Lown Institute, founded by Nobel Laureate Dr. Bernard Lown, works to curb medical overtreatments. And they make some convincing arguments about overtreatment in general. In short: With uncertainty in medicine, there are loads incentives in the American medical system for a doctor to "do something" or "prescribe the latest medicine/test/treatment," even if doing nothing (or promoting basic lifestyle changes instead of drugs) would be just as effective. There is financial pressure to use costly medical equipment or get insurance reimbursements for treatments (it's more lucrative to prescribe something for patients than simply talk to them), a fear of lawsuits by doctors if they don't "do something," and sometimes even pressure from patients who demand to receive what they just heard about on TV.

So the problem isn't more health data itself, but data without proper context. We can already see how this works whenever there's a study that "Substance X helps prevent Disease Y." If you were able to read the entire study, it's not always 100% clear that Substance X alone has the preventitive powers. Maybe it's actually Substance X in combination with Substance Z as it occurs naturally in some foods. Or some sort of behavior that happens to be associated with people who consume a lot of Substance X. Or actually Substance Q, which happens to be found in the same foods that contain Substance X. Perhaps the study subjects were all young men and results don't translate to other demographics. Without the full picture, jumping to conclusions can be dangerous. Yet even if a news report includes those caveats (and not all do), how many people will read one story about the benefits of Substance X and go out and buy a new supplement?

You see this also in well-intentioned data projects such as showing hospital mortality or re-admission rates. Hospitals with higher-than-average mortality or re-admission rates for a typical medical procedure must be worse than average, right? Well, it turns out some of the world's best hospitals can look pretty bad in such data because they have a higher proportion of difficult cases than your average community hospital. In fact, your average community hospital in a major U.S. metropolitan area is likely to refer the most difficult-to-treat cases to a regional medical center. Look at this data without context and you may not make the best treatment decision.

You see this also in what the Lown Institute calls "indication creep" -- where medicines, treatments or tests that are truly helpful to people with specific conditions end up "used on less- and less-sick people (for instance, when a drug manufacturer tries to expand the market for its product)." Cholesterol-lowering statin medications are one striking example -- while helpful for those with actual serious heart disease, they're also being prescribed en masse as prevention for people with high cholesterol but no known coronary disease. And that could be a mistake.

"A group led by Dr. John Abramson, a teacher of healthcare policy at Harvard Medical School, reanalyzed the largest-ever research report on statin drugs. Abramson's group found evidence that for all but the very highest-risk people, statins did not save lives and did not reduce the frequency of serious illness," writes Dr. David H. Newman at the Huffington Post. "Abramson's group found evidence that for all but the very highest-risk people, statins did not save lives and did not reduce the frequency of serious illness. . . . When low-risk people take a statin, the chance that the drug will cause diabetes is roughly equal (at best) to the chance that it will prevent a nonfatal heart attack."

The bottom line, says Dr. Newman: "The bestselling pills of all time don't save lives or reduce major illness for most who take them."

Let that sink in for a moment. Thousands upon thousands of people are paying for medication that is likely not doing anything to prevent illness, and in fact could cause harm.

It is into this environment that Apple is going to help people get significantly more medical information about themselves. Is that in itself a bad thing? No, of course not. Just like with medication, there are many good things that can come of it -- when used responsibly with proper knowledge. But also like with medication, there's the risk of unwanted side effects.

Computerworld's IT Salary Survey 2017 results
Shop Tech Products at Amazon