Your Record Precedes You

Via Ashish Jha, the fascinating story of what happens when Big Data meets electronic medical records meets American health care:

… a growing number of hospitals [are] using their patients' health and financial records to help pitch their most lucrative services, such as cancer, heart and orthopedic care. As part of these direct mail campaigns, they are also buying detailed information about local residents compiled by consumer marketing firms — everything from age, income and marital status to shopping habits and whether they have children or pets at home.

I did a piece a while back on privacy and electronic medical records, but I confess I didn’t think about this particular permutation of privacy breaches. I do feel pretty dumb for not anticipating this particular consequence, because in retrospect it’s obvious.

The first worry I have is about privacy—I’d prefer, however anonymized, that my data not fall in the hands of these tracking companies (who I’m guessing are combining medical data with credit card data and internet tracking data and so on and so forth). The second big worry here is that it may be a way of promoting provider-induced demand, i.e. care that isn’t really needed.

Austin Frakt pointed out that it’s unclear whether providers do induce demand in patients, but there’s a fairly decent story we can tell as to why it might be true—we do generally trust our doctors to recommend the right care for us and therefore it would seem to be possible for our doctors to get us to take care we might not otherwise want or think to get. An intense version of this story might have doctors recommending care that’s known to be unnecessary.

One of the anecdotes in the story might confirm this particular interpretation—one hospital runs a campaign promoting mammograms for women over 40 and finds it to be quite successful. If you recall the recent mammogram controversy, you’ll remember that the guidelines were changed—the new ones suggest that women between 40 and 49 decide for themselves and then recommend mammograms after 50. As you might expect, the evidence on either side of the debate is muddled; consequently, it does seem a bit off to be conducting an advertising campaign aimed at drumming up business.

What I think this goes to show is that electronic medical records are wonderful tools, but they work best within a broader context: a health system concerned primarily with promoting health (rather than providing services for fees) might look askance at such promotions.