The major trend in the past fifteen years in health insurance has been an embrace of high deductible health plans (HDHPs) under the guise of “consumerism”. The theory is that if people are on the hook for the first $1,000, $1,500, $2,500 or $5,000 of medical expenses, they’ll aggressively shop for care as well as take the necessary preventative steps. The preventative steps are no longer just moral nagging (dropping 40 pounds might help your knee) but financial cost avoidance (dropping 40 pounds will help avoid/delay a knee replacement and a $5,000 bill you don’t want to pay).
This is an analysis and a solution set that comes out of the assumption that most people consume too much care, and furthermore most people are fully rational consumers of everything. It is very much an Econ 101 problem where the assumption of low information costs, and low confusion costs can lead to optimization behaviors.
Two interesting pieces on decision making popped up at the Incidental Economist today that highlights the basic fact that there are significant decision, search and processing costs as well as strong social pressures that move people out of the rational man realm.
Quantitative intuition is a tough skill to build, but one that is required to make relative risk assessments and value judgements. Most peope suck at it (that is why the scam of actively managed mutual funds are still an ongoing concern instead of seeing the bottom 98% of the income distribution dump their retirement savings into either a Target 20XX fund or an index fund depending on risk tolerance):
These researchers got together more than 1800 adults, about half of whom had diabetes. Then they showed them lab results of A1c values randomized to be either slightly or moderately high. They mixed these in with other random lab values. Then they asked them if their A1c level was outside the normal range, what this said about their glucose control, and if they would call their doctor out of concern.
Just over half of participants correctly realized that the A1c was out of the normal range. Those that HAD DIABETES only realized this 56% of the time.
The people who have been exposed to A1C (the diabetic population) still sucked at recognizing what the numbers meant.
Under a rational regime of treatment and screening, priorization should take place. If I enter a trauma center with a sucking chest wound, a sword cut to my abdomen and coincidentally a blown ACL, the prioritization of care is simple — keep me breathing, keep my heart going, and then deal with the big issues that could kill me in the next twenty minutes and work the way down to referring me to an orthopedic surgeon to take care of my knee sometime in the next six months.
That makes sense. However, on non-acute situations, we screen people for diseases that may kill them after they’ve already been dead for a decade:
if you have a limited life expectancy, or a short time to live, then screening for some diseases really doesn’t make much sense. If cancers have a very high 10-year survival rate, then it doesn’t really pay to do much screening if you have less than 10 years to live….
Patients with a very high 9-year mortality risk were screened for cancer quite often. Women who had had a hysterectomy were screened with PAP smears between 34% and 56% of the time within the last 3 years. Men were screened for prostate cancer at a rate of 55%.
A patient who is told that they won’t last long enough for a cancer screening to matter will scream bloody murder and get on the front page of the major local papers and on at least two 6 o’clock news hours and be placed before the weather and sports report. We don’t have the discussion that an 87 year who is already fairly frail probably won’t make it to 100 so we should not care about things that will kill them when they turn 100.
Those are flaws to rational man; they are not particular flaws to satisficing/most rational with known human heuristics, schemas and decision process weaknesses, but consumerism as a means of saving money is designed for rational man.