It is a quality measure, but to use it to compare quality you need a lot more data than that.
First of all you would need the same data for other countries. Then you would have to make sure that their definitions of errors, and their reporting of errors were the same or very similar, and if not you would have to figure out a way to adjust for those differences. Then you would have to turn the total numbers in to rates (and also pick which rate you find most important, or perhaps publish more than one, rate per person, rate per patient, rate per incident, rate per incident adjusted by severity, etc. Then you might have to adjust for other factors.
Say country A has more hip replacements then country B (both in total and per capita), and more deaths during the hip replacements (in total, and per capita, but not per incident).
Do you mark A down for having higher deaths per capita from this procedure? Maybe? It makes sense if it has more deaths from having more transplants in borderline or even sub-borderline cases where the benefit of the procedure is marginal or even negative (even assuming it is done right). If its more deaths because of dubious over treatment then that would be a black mark against it.
OTOH maybe its more deaths from treatments that help the majority of those who receive them, even the majority of marginal cases. Maybe there are very few questionable or unnecessary procedures in country A, and Country A has more cases then B, because B fails to provide this procedure to many who could use it. That lack of provision would be a strike against B, but measuring the death rate (per capita at least if not per incident) would show A as worse. |
