Friday 21 September 2007

Using genetics for diagnosing diseases (and insurance purposes)

The Economist had an amazing note about genetics for diagnosing illnesses, and the privacy concerns (mainly, how it could be used by insurers to raise rates).
However, there is a systemic consequence if insurers use that information. Let's make a Gedankenexperiment (oh boy, how I love that word). If an insurer could know exactly what genetic diseases a person will get there will be three kinds of costs an insurer will have to cover:
1.an amount every person spends in health every year
2.an amount that some people might spend and others not
3.an amount for genetic diseases

The first amount is a small amount, but since every person uses that amount, it doesn't make sense to have insurance for that (since every person spends that amount, rates should include that amount and a small extra for administrative costs, so it is not convenient for the client to be insured for that, and in a competitive environment, it is not convenient for the insurer, since higher rates means clients going to other companies). I.E. large loss insurance condition is not satisfied in these costs, and that's why health insurance policies usually have deducible clauses.
The second amount is a variable amount that insurers will always cover, where there is variability among individuals. This is why insurance exist: due to the law of large numbers, risks will average among these individuals, so what could be ruinous to one person is distributed in time and among different people, making it affordable to the group.
The third amount is currently inside the second, but insurers would like to split it, so "genetically healthy" people pay less than "genetically unhealthy", giving better rates to a group and attracting more clients. However, this could be ruinous to insurers. They will insure only the small costs, not the large ones, where profits are, turning health insurance into a cheap commodity. If a disease is certain, you'd be better off saving money for the covering the costs yourself rather than giving it to insurers, since you will be saving the administrative costs. In addition, not only it would be bad for particular insurers, but for the industry. You'll have low cost insures, with thin margins, who only take "genetically healthy" clients, and all-inclusive, who theoretically will take all kinds of people, but since it is more expensive, only genetically unhealthy people will go there.

Don't you love systems theory?


Add to Technorati Favorites

No comments: