Fred Trotter: Data, damn data, and statistics

Why does this blog use the word “damn” so often? A search produces a whopping 38 hits, such as:

These and other “damning” articles deal with statistics, so no wonder our normally polite bloggers would resort to swearing. How our health care data is used and by whom is incredibly important to e-patients, and the fact that much of this is out of our hands rouses a lot of emotion.

This post by guest blogger Fred Trotter originated in an SPM members’ listserv thread that debated which is scarier: the theft or loss of data or the intentional use of health care data by third parties?

There is a lot of danger in “legal but unethical” secondary data use. It is a far greater threat than hackers who steal health care data, IMHO.

I do want to point out, however, that there may be some benefits to a well-managed and ethical “health credit score” concept.

Consider how fundamentally unfair and inaccurate the current credit score system is (the normal one). But that system has created a backbone where normal people in the United States have real-time access to credit decisions. In practical life, this means that if you have decent credit but no money, and you need a sofa, you can get a sofa now. You can use that data liquidity to buy a boat or you can use it to buy a laptop for your new business.

In other countries, personal credit is unworkable because they do not have an effective system for ensuring that people will pay back debt (i.e., not paying it back will hurt their credit scores and limit future purchasing options).

From a health care perspective, that is where we are in the US; we are paralyzed. Without a health credit score, people who are disciplined about their health (i.e., e-patients) end up paying for people who are less disciplined. This creates a crazy incentive where healthy people have strong financial motivation to not get health insurance at all. Which of course makes it more expensive for everyone.

The individual mandate is an alternative approach to a “health credit score,” but it may not survive politically. If it does not, a health credit score might be an important enabler for ensuring that healthy people participate in insurance pools, which is a critical issue.

The real problem here is that the distance between an unethical system which penalizes the sick for just being sick (which is what we have now in any case) and an innovative system where being an e-patient pays off is just a hair’s breadth … and may in fact require the use of precisely the same data.

I do not want to pretend that I have a deep understanding of these issues. I certainly do not. But as a community, e-patients really need to begin grappling with these delicate data-use-balance issues.

The e-patient community is really one of the few that has a tractable notion of what a “high performing patient” might look like. But insurance companies, economists, policy makers, doctors, congresspeople, and lobbyists are all using other models to make assumptions about what patients might be capable of. I am convinced that many of these assumptions are untenable even for high-functioning patients, much less patients of average education and motivation.

Given the scientific approach that I and other “N=1/Quantified Self” e-patients have taken in trying to improve our personal health or wellbeing, I submit that if none of us can vouch for an assumption being made about patients, it is probably a bad assumption. How do we communicate this sort of correction to policy makers? How do we continue to discharge the “patient scientist” portion of the e-patient philosophy in a way that helps policy makers make good decisions? How do we ensure that patient data is used fairly?

Hell if I know.

Print

Posted in: general | hc's problem list | medical records | policy issues | reforming hc | understanding statistics

 

 

Comments

4 Responses to “Fred Trotter: Data, damn data, and statistics”

  1. Kelly Young says:

    I’m sorry. I’m confused by this: “The e-patient community is really one of the few that has a tractable notion of what a ‘high performing patient’ might look like.”

    I fancy myself a “patient scientist” of sorts, but I do not know whether I can be a “high performing patient.” I had not connected “e-Patient” with necessarily being “more healthy” or “less health care use.” I had thought that it was more a function of how we respond to illness. Some sicknesses are behavior-related, but many are not.

    Rheumatoid disease is systemic, progressive, and destructive. There is no known way to prevent or cure it. Remissions are rare. The treatments for Rheumatoid disease are some of the most expensive medications in the world, but they work well on a minority of patients. I do know that when we do get the monitoring or treatment that is needed, it costs even more. A majority of us become disabled. Where do e-patients with Rheumatoid disease fit in this scenerio regarding “high performing patients”?

    Before this disease became full-blown in 2006, I would have had an excellent “health credit score” to go with my perfect financial credit score that was entirely behavioral based. Being jsut as “disciplined” about my health did not make a difference and we cannot assume that it will.

    Again, I admit it’s confusing to me and I’m sorry if I just don’t get it.

  2. Fred Trotter says:

    Kelly,
    I think, in fact, you -do- get it.

    If a “health credit score” is going to work, it has to successfully differentiate between a persons “health” and a persons “health behaviors”.

    So if the conceps were going to be viable, then you should have had a good score before -and- a good score after your disease became full-blown in 2006. On the other hand, if a patient is told by their doctor “you are going to develop Diabetes unless you eat better”, and then still eats like crap for ten years, and then gets diabetes… that is a different story.

    That person should have a bad health credit score, both before and after they get diabetes.

    For the “health credit score” to work, it seems critical that be entirely seperate from the issue of health and instead be linked to “healthy behaviors”.

    The normal credit score actually has a similar disconnection with personal wealth. You can be poor with good credit, poor with bad credit, rich with good credit, and rich with bad credit. A health credit score would have to be similarly disconnected to health. (in fact even more so, since rich people can “hack” their credit scores)

    I want to reaffirm, that I am not advocating for a health credit score. I am just thought-gaming how it might work. I am not at all sure that this is a good idea.

    -FT

    • Fred, I couldn’t agree more that IF there were a health “credit score,” it would have to be tied to behaviors. IMO it’s infinitely unethical to rate a *person* on something they can’t control.

      But I say the entire rating industry should be damned, or at least regulated 1000x more than it is today, because the totally unlevel playing field of that industry today is disgusting. I’m still working my way out of the financial consequences of my cancer year (2007), (which was greatly compounded by the real estate collapse that year, another long story).

      And try as I might, to this day I CANNOT GET AN ANSWER on what steps I must take to put my credit score back where it was.

      I can get (vague) REASONS why it’s not better, but not a word about what specifically would give me a better score. And I can’t get a soul anywhere to give me an explanation of how that’s fair, by any stretch of the imagination: to have someone trapped in a situation they did everything possible to avoid, and not let them know specifically what will get them out.

      So to the idea of a health credit rating, I say: To hell with them all. To hell with them.

      • Kelly Young says:

        Thank you for your reply, Fred.
        I really am surprised to read that because people in my condition cannot usually buy life insurance. We are a very poor risk so I assumed that we would have a low score.

        Smoking, drug abuse, and obesity are the only easily measurable risk behaviors I can think of and there could be problems scoring even those. So far, I like Dave’s “to hell with them” best of all. Who would know or judge what is under my control or not? Most people do not understand Rheumatoid disease including most insurance companies – we are constantly told to exercise to improve our condition and that only works in mild cases – as a glaring example. For many of us, it can contribute to damage and disability.

        As for finances, I’m with ya Dave! The same year I got sick, I bought a small house at the peak of the bubble right next to NASA, which closed down over the last 5 yrs. My house is worth less than 50% of what I paid for it and the only house that sold here in the last several months was a foreclosure that was vacant for years. One day I’ll have to bail just like the folks on 60 Minutes the other night because my kids will need to find work somewhere & I cannot live alone – WHEN the day comes that I let go of this house, my credit score will be the same as my health: shot to hell no matter what I did.

Leave a Reply