By Alice Lam
In the NYTimes article, “When a Health Plan Knows How You
Shop” predictive analytics seem to be the new wave of helping companies
determine consumer behavior and habits.
However, predictive analytics come with costs. University of Pittsburgh Medical Center is
now one of the leading institutions running predictive analytics to provide
better, more efficient healthcare. Although
it sounds great, the fact that UPMC has an insurance division does incentivize
the company to cut costs where it can. Targetting the more high risk patients may seem
like a noble deed, we run the risk of allowing the insurance companies know so
much about the health of the customer, the allowance of patients who in need of
care may be harder (since a higher risk patient would cost the insurance more)
and cherry picking the best patients who are healthy so the insurance can be
more profitable. Since UPMC has all this data showing correlation among claims,
prescriptions and demographics and probability of urgent care, UPMC predict if
the patient might end up in the emergency room.
Knowing so much feels like an invasion of privacy. Many argue that companies buy this kind of
consumer behavioral data with other goods and health data is just another
extension of that. However, healthcare
data seems to be a tricky case because there are HIPAA guidelines and the
Hippocratic oath that seems to be a conflict of interest. When a doctor can’t even reveal to the wife the nature of her husband’s
disease, it hardly seems right for a company to know all the patient. It doesn’t seem right for the insurance to
encroach on the patient’s privacy so the insurance can nudge the patient into
taking better care of himself. Although
saving lives is very honorable, we would never take a person’s life so we can
donate the organs to many. Predictive
analytics may be effective but also extremely utilitarian. It is really hard to judge the intention of
the hospital/insurances (better care or profit), how much marketing companies
nudges the patient and how well it actually works. Even if the intention is good and the patient
does do more screening, health outcomes may not be better. Like with all other forms of digital marketing,
only time will tell.
No comments:
Post a Comment