After many failed attempts, the US House of Representatives succeeded in “repealing” the Affordable Care Act, a.k.a. Obamacare. Few people seem to be sure of what’s in the new bill, except that any remaining protections can be overridden by individual states, if they get permission from the Feds to do so
Much commentary and no small number of protests have followed, but one comment by an MD caught my eye. This author regurgitates the oft-repeated argument that personal financial incentives will improve health and lower health-care costs. He cites the case of a financially-strapped hypochondriac who was apparently wasting the author’s time with repeated visits. By making those visits successively more expensive he could “guarantee” that he would see this patient less often.
Indeed. By extension, it would appear that by allowing physicians to focus on rich hypochondriacs while ignoring the poor ones, the problems of the health care industry would be solved.
I think not. Before going any further, this might be a good time to pause and reflect on a universal truth, which apparently has not reached the rarified realm of Planet MD: the vast majority of us really don’t want to spend much time with you guys. It’s not personal, it simply reflects the simple idea that if we’re going to see the doctor a lot, there is something wrong!
That said, let’s look at the idea of financial incentives as a means of reducing self-injurious behavior. It’s a dubious connection: smoking has become increasingly expensive, but smokers are disproportionately of lower income. There are gambling addicts poor and rich alike. Economic motivators are rational, but self-injurious behavior is not – which do you think will prevail?
We can make all the arguments we like about offering people a financial stake in their health outcomes, but many people are beyond monetary motivation. Ironically part of the remedy in such cases would entail counseling, i.e. more health care rather than less. Should individuals with addictions be punished? Well, let those of us with no addiction to work, to power, to money, to sex, to gambling, to drugs, to smoking, to overeating, to eating poorly, to a sedentary lifestyle, to multitasking, to depression, or to alcohol cast the first stone. Economic incentives of this kind will probably be proposed as long as there are Republicans to propose them, but that doesn’t mean there is an objective basis for their standalone utility.
Of course our behavior is only, as analysts put it, a partial predictor of our health. Our environment and genetic makeup will play a substantial role in our longevity and health. Why does one person contract cancer, while another person with a similar lifestyle does not? Why does one person have an accident and break a leg, when another does not? We can’t tell – we don’t have the information. And because we don’t have the information, we cannot predict the outcome, and when we cannot predict the outcome we have no chance to fully control it. Regardless of the financial penalties our government might imposed on us for unexpectedly becoming seriously ill or injured, some of us will become so anyway.
Individual responsbility is laudable, but has its limits. If we plan to apply a financial lever to promote responsibility, but cannot predict the role of responsibility on all outcomes, precisely how does that work? Well, it doesn’t work.
What about the idea that eliminating a class of unworthy patients will free up resources in the healthcare system and improve healthcare? It turns out this notion also assumes a great deal too, in particular that by freeing up MD time that time won’t be wasted in some other way, for example by rich patients who are not really sick, or who have time on their hands, or need a day in bed to remedy the effects of a virus the MD can’t deal with anyway. One effect is often connected to other related effects, with the result that we don’t get the rsult we hoped for.
All of this has analytics precedent. It applies both to the realm of predictive modeling, where we often find that models for individuals require a knowledge about their environment that is not forthcoming. At the same time the inputs we hope are relevant, such as economic incentive, may prove to be marginally effective.
Beyond that, once we seek to optimize a system based on our model, we encounter other issues. First, what do we optimize in this case? Cost per patient? Patient outcomes? There isn’t a simple and straightforward “best” answer. Second, it’s very tempting – but very often wrong – to assume that if we alter one variable the overall system will respond to that variable alone. Will reducing visits by one kind of un-sick patient will allow doctors to spend time with only the truly sick? Maybe, but we might just end up with patients just slightly less sick, or with patients equally healthy but a little richer, or in a provider system with enough slack, with little to do. Until we understand all of the constraints on this particular optimization problem, and what we’re really trying to optimize, we cannot make a sound decision. It’s just about as hard as it sounds – the posing of the problem is as big a challenge as solving it, once posed.
As Trump has admitted, while continuing to promote dubious solutions, health insurance is a very involved matter, probably made more complicated by the current House bill. Insurance should be about managing group risk in the face of random factors. The complexity – and let’s be honest, the duplicity – of the House bill obscures that fact.
To top everything off, health insurance is really only the beginning of any healthcare discussion – is a very imperfect predictor of health outcomes. To have the possibility of health insurance is no guarantee of affordable insurance, or good coverage, or good care, or a good outcome. The health insurance discussion is about the potential of good care, not its reality. In a country where 20% of the GDP is related to health care and we are facing the return of unaffordable health insurance for many Americans, we really face issues everywhere along the financial health care chain, from insurance to provider costs.