Last week’s New Yorker has another great article from Atul Gawande, this time about Medical Malpractice (print only, Q&A with author here).

It’s a musing on what the right remedy for medical malpractice is. The example that encapsulated the article best for me was towards the end of the article, when he compared being a doctor with a baseballer. He said that a good third baseman will make a stupid mistake that costs a home run, and sometimes a game, maybe 2% of the time. That doesn’t sound like much. But if you’re a surgeon, and 2% of the time you make errors that injure or disable someone, and every now and again they cost someone’s life, then what should happen?

His suggested solution is the NZ system, which is an accident compensation system for any kind of accident. If you are the victim of an accident that causes personal injury (there are a bunch of exclusions, but its basically anything other than illness) then you are entitled to compensation which is limited to a maximum lump sum amount if you are catastrophically injured.

I used to think that a no fault system like the NZ was the best system for any kind of injury compensation, until I realised that there is potential for an adversarial legal system to create change; i.e. if you know you are likely to be sued for something that is your fault, you might do something about it. I realised this when I watched the NSW education system completely change their system for coaching rugby union in schools not when a 15 year old boy became a quadriplegic from a collapsing scrum, but when the courts awarded him a $2m payout.

The trouble is, the legal system is a very blunt instrument. The system described in Atul Gawenda’s article, which seems pretty similar to what I know about the NSW system, is one in which the adversarial nature makes it almost impossible for doctors to learn the real lessons that arise when something goes wrong. A big payout like the rugby scrum payout might change behavior because it is new and different, but a smaller payout like a moderately routine case that reduces someone’s mobility might not be noticed by anyone other than the doctor involved.

Some errors might be completely unavoidable, and it doesn’t matter how you compensate people. But some errors may be about processes that are flawed, that could be changed if the system of compensation made learning lessons possible. A system like NZ’s, by having all errors centrally administered and compensated, does also allow for people to learn from their mistakes, if the systems are set up to do so.

I’m currently thinking that on balance, the collective system (a la NZ) is best. But I don’t believe it is cut and dried, and I do wish that policymakers thought more about how any compensation system might ensure an overall lower error rate when they are thinking about how a system should work.