More on point, fee-for-service incentivizes unnecessary and expensive diagnostic tests. Let’s say a patient presents with an uncomplicated broken arm. If there’s a set payment for an outcome (successful treatment of the break, with fully restored function), and diagnostic and treatments costs come out of that fee, the incentive is to order as many x-rays as are necessary to successfully treat the injury and no more. In fee-for-service, the orthopedist has no incentive not to order superfluous imaging because it doesn’t cost her anything, and may have incentive to order extra imaging (if, say, she also owns the imaging center that bills for it).

Medicine is littered with examples where there are two tests whose diagnostic power is statistically equivalent, and yet doctors common write for the more expensive test. Fee-for-service is either a neutral or perverse incentive in these cases. Fee-for-outcome aligns the incentive with what society seems to want (lower-cost healthcare with good outcomes).