Wednesday, June 5, 2019

quantifying risk

Use of probabilties p and surprisals k ln[1/p] in communicating and monitoring risks to medical patients could make patient decisions about actions with a small chance of dire outcomes as informed as possible. This could reduce the costs of medical malpractice in the long run by empowering patients with tools to make informed and responsible choices, making the need for legal redress less frequent. Thus the media could play a key role in reducing the costs of defensive medicine.

The same may be true of defensive practices e.g. in food and chemical distribution. Instead of all bakery goods saying "manufactured in a facility that processes peanuts" so that folks with a peanut allergy (not to mention teachers wanting to bring a treat to class) are thrown under the bus, instead teach the population to quantify risk (using surprisals or probabilities) and then develop guidelines for putting the risk into context based on real data. In this way, folks who use a product take ownership of the risk that it confers at the outset, taking personal injury-suit lawyers out of the loop except to the extent that they can contest the data used to report the risk to begin with.

The bad news is that every thing you do brings with it a finite risk of almost any dire outcome. The good news is that this risk is likely to be much smaller than you might guess, from the non-quantitative hype that the media brings to every problem that it decides to fan flames on today.