Subscribe to DSC Newsletter

Beware: Employers Mine Your Personal Data to Predict Health

Watch out folks - there is a new breed of health analytics firms sprouting like weeds - mining "big data" about you and making broad and bold predictions about your state of health to make workplace decisions.

 

I am in my 50's and a fitness nut who preaches frequent exercise and good eating habits. I am in great health and have not seen a doctor in over 15 years. Yet I am concerned about third parties, especially employers and big government, accessing and mining folks personal data, including where you eat, shop and play - combining it with your personal health data - and using it to make decisions about your life that is likely inaccurate and unfair.

 

This type of personal data mining is ripe for abuse at many levels - and access to reasonably priced health care insurance coverage and employment hiring / promotion / firing decisions is at the top of the list.

 

There is a good chance your employer and health insurer (both partnered with big government) is currently or soon about to mine data about where and how you shop, where you eat out, whether and when you vote, and the prescription drugs you use - to predict your state of health, forecast future health problems and recommend preventive treatments.

 

While this may have some potential positive benefits like decreasing healthcare costs and nudging you to eat better foods, exercise more and use cheap preventive health care services - it is rife with diagnostic accuracy problems at this time considering the high rate of false positives, bad data, patterns that mean nothing, correlations that do not equal causation, and discounting genetic and other factors. Human physiology is a high causal density environment and assessing overall health is tricky even for trained physicians.

 

I seriously doubt the predictive accuracy from such "big data" mining at this time. Moreover, the strength of any real predictive signals is likely to be weak without more relevant smart data about your normal diet (not just dining out), exercise patterns, stress levels, happiness and genes.

 

In the future more relevant smart data from in-body sensors combined with genetic information will provide much more accurate personal health predictions. But that day has not arrived yet and using "big data" to make dubious health predictions and bold workplace decisions is - in my humble opinion - morally wrong and terrifying.

 

Make no mistake, this data may be used to fire you, drop your insurance coverage, increase insurance rates and influence job promotions. After all, who wants to hire or promote someone with health issues that could impede job performance in the future? Why take the risk? All things being equal, better to hire or promote the healthier yet less qualified chap.

 

Some purported health analytics experts assert they can more accurately predict your risk of a heart attack by where you shop and eat. Really? Do you have hard evidence to support this prediction? Does correlation here really equal causation? How does genetics, stress levels, exercise habits and diet when not eating out factor into this prediction? I am skeptical you can obtain a strong predictive signal simply from where a person eats or shops from time to time.

 

Welltok - a health analytics firm - does predictive health assessments for Colorado state employees. They claim folks who vote in midterm elections are in better health than those who don't vote in them. Are you kidding me? Please show us the hard evidence in this case that correlation amounts to causation. What is the sample size? Are Colorado state employees really representative of the population at large? Color me skeptical about this health analytical finding.

 

While health data privacy laws in most countries prevent employers from accessing personal health information, there are numerous legal ways to go around the law and get a full copy of your health records. For example, organizations hire third party health analytics and wellness firms and authorize them to collect your personal data from health care and insurance companies.

 

Those wellness and health analytics chop shops then purchase information from data brokers about your individual consumer behavior and health needs - and attempt to connect the dots. Their predictive analytical processes is opaque so we have no clue about the quality and accuracy of health predictions. I predict a majority of their health predictions are wrong. And we have no idea how big business and big government is using this highly private information about us.

 

And good luck, dear reader, getting a copy of their prediction about your future state of health. Do we really want to be heading down this road as a society? Is there no reasonable zone of privacy anymore?


George Orwell is either laughing or turning in his grave.

See: http://bit.ly/1ow9EVF

Views: 922

Tags: Data, Healthcare, Predictive, Privacy, analytics, law, mining

Comment

You need to be a member of Data Science Central to add comments!

Join Data Science Central

Videos

  • Add Videos
  • View All

© 2019   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service