Complication rates are affected by hospital procedures and policies, staffing levels, training, and presumably the pressures that are or aren't put on doctors.

Complication rates are affected by hospital procedures and policies, staffing levels, training, and presumably the pressures that are or aren't put on doctors. All of those are under the control of the administrators who look out for the bottom line. Whether complications make money or cost money surely affects administrators' eagerness to revamp and/or enforce handwashing policies, update training for nurses, keep staffing levels high enough that inexperienced nurses are properly supervised, and put pressure on doctors with high complication rates. To make it more concrete, there was an article on HN a few years ago suggesting that post-surgical complication rates could be reduced by creating surgery checklists where very basic surgical steps would be checked off as the surgery proceeded: patient identity verified, everything properly sanitized, materials accounted for afterwards, things like that. You would expect a hospital administrator to make a cost-benefit analysis taking into account the cost of researching this idea, creating a trial program, and monitoring compliance and effectiveness. The cost-benefit analysis looks like this: If the program is ineffective, we lose money. If it's effective, we lose even more money. If it's really, really effective, it could screw up the bottom line so much that I lose my bonus or even my job.