Marshall Allen is an award winning journalist whose current ProPublica/NPR series Health Insurance Hustle is a must-read for doctors, nurses, medical students, policymakers, healthcare business professionals and patients (basically everyone). His deep dives touch on the rising costs of healthcare as well as how “big data” and wearables are impacting patients and driving decision making at insurance companies.

Allen’s article on “health scores” was one of the main influences on my starting this blog and if you read only one health care story this year, make it that one.

Allen was gracious to take time from his busy reporting schedule for a Q&A with Data_Chi.

How did you come to learn about “health scores” and what interested you in writing about the topic?

Marshall Allen: I had a data analyst from an insurance company give me a tip about this issue. Her name is Erin Kaufman and I interviewed her for the story. When she told me about it I looked online and saw the LexisNexis health risk scores, which became featured in the story.

Companies such as 23andMe are also collecting data, genetic data specifically that has much larger implications for privacy and predictive analysis. How do you see the evolution of this data collection effecting these health scores?

I don’t know for sure because it’s hard to see what happens inside these companies. But I do know that the insurance industry is gathering as much personal information about us as it can to try and figure out how much we might cost to cover. The industry has used this type of information to discriminate against sick patients to pass along higher costs to them, which can reduce access to care for the sick. Genetic data would open doors to continue this type of behavior.

One of my objectives is advocating for the open publishing of these health scores similar to FICO scores, hopefully making them more fair and balanced. From your research, what are you thoughts on that and did any of the companies you spoke to or researched have any interest in doing this?

This seems like a sensible way forward. We have not had these type of conversations in policy circles in any meaningful way. The laws have been lax and the companies have been able to operate without a lot of limitations. So now we are playing catch-up. But it seems fair that if our health and other information is going to be used to rate us, we should at least be able to make sure it’s accurate. And there should also be accountability for the algorithms that spit out these scores. Computers may produce a lot of numbers and make a lot of inferences that are inaccurate and then that could have costly consequences for people.

This image is from a LexisNexis brochure on their socioeconomic health scores

The California Consumer privacy act and the EU’s General Data Protection regulation seem to the inevitable legislation for some sort of standardized regulation of these scores. Currently, it seems to be the wild, wild west in terms of the data these companies are collecting. Where do you see further domestic regulations heading?

I don’t know, but there has been very little in terms of public policy debate about these subjects, especially when you consider how far we have already gone down this road.

People are well aware of the Facebook privacy scandals, but it seems the health care data issue lurks in the shadows – why do you think this isn’t being paid enough attention?

It’s complicated so people don’t understand it. I’ve also found that some people care about privacy and some don’t. We also can’t be sure what type of harm has already occurred, so some people just don’t care. This also would be more negative for sick people, and so people in good health may not feel as worried. 

Ultimately I see Amazon and/or Google making a much bigger footprint in health care and as a result the privacy and data implications are huge, what are your thoughts on how things look in the coming years?

I obviously don’t know what they will do. But I have found that companies are striving to make money for themselves and their shareholders. They are not operating in the public interest, so we should not assume they are somehow going to make things better for the public. Everything they do should be scrutinized, as is true for anyone who is promising to solve these problems. Hopefully they will find a way to benefit the public while they are pursuing their profit. But we should not assume they will do so.

Follow Marshall Allen on Twitter

Bill Fair and Earl Issac started the Fair, Issac and Company (FICO) after meeting at Stanford University in 1956. Using their combined skill sets, Fair an engineer and Issac a mathematician, the two developed an algorithm to predict credit risks. FICO built their predictive model using consumer behavior  and demonstrated that, contrary to many old-fashioned banker beliefs, race played no part in a borrower’s ability to repay loans.

As a consumer, you are entitled to one free credit report annually and there are now services such as Credit Karma that allow you to essentially pull your credit score without any impacts whenever you wish. If you notice errors or discrepancies in your report, you can take actions to resolve them.

This same model is needed in the consumer healthcare industry. Companies that collate data on patients and calculate these scores also need to make them publicly available and there should be an agreed upon standard set for who can collect such data, what they can collect, what data points should be allowed in factoring these scores and how long can this data can be stored.