> 1. Healthy patients cost less in the long term. Solving for long term issues (type 2 diabetes!) vs. the immediate issue and getting someone out the door is profit seeking.
This is an extremely common misconception. Luckily, you've picked a really easy example to illustrate.
"Solving" type 2 diabetes is an incredibly tricky problem. It's not something that can even be measured in the short-term, let alone solved. It's also quite individualized, more so than probably any other common diagnosis except cancer. There's no magic bullet intervention, and it takes a lot of active work to build the provider-patient relationship to a point where it can identify the appropriate interventions for that patient, identify problems with those interventions, and help the patient execute on them successfully.
Because it's so individualized, and because it's so long-term, any system designed to "streamline" that is ultimately going to fail, because the curse of dimensionality will inevitably collapse away the requisite information, and the pursuit of short-term profit will inevitably cause the system to prioritize short-term interventions over long-term ones which will actually address the issue.
"Healthy patients cost less in the long term" is technically true, but there's no way of turning unhealthy patients into healthy ones that doesn't, in aggregate, cost more than the theoretical savings.
> Why would Amazon not aggressively track outcomes, measure every dimension of provider care, and ruthlessly fire / downgrade low performers to maintain the above goals?
Again, this is an incredibly technocratic approach to a field (care delivery) that requires a different approach. It works if you assume that negative outcomes are the fault of individual providers ("low performers") and that individual providers have both the agency and resources to solve arbitrary cases. In practice, that's almost never the case: there are way too many confounding factors that impact clinical outcomes, and providers are quite limited in their power compared to other players: the insurance companies, the administration of the system they work within, and most importantly, the patients themselves.
Honestly, the only way you could meaningfully "fire low performers" is to identify patients who aren't performing well and then drop them. Which is, in a way, what the health insurance system does... but we tend to consider that an antipattern.
This is an extremely common misconception. Luckily, you've picked a really easy example to illustrate.
"Solving" type 2 diabetes is an incredibly tricky problem. It's not something that can even be measured in the short-term, let alone solved. It's also quite individualized, more so than probably any other common diagnosis except cancer. There's no magic bullet intervention, and it takes a lot of active work to build the provider-patient relationship to a point where it can identify the appropriate interventions for that patient, identify problems with those interventions, and help the patient execute on them successfully.
Because it's so individualized, and because it's so long-term, any system designed to "streamline" that is ultimately going to fail, because the curse of dimensionality will inevitably collapse away the requisite information, and the pursuit of short-term profit will inevitably cause the system to prioritize short-term interventions over long-term ones which will actually address the issue.
"Healthy patients cost less in the long term" is technically true, but there's no way of turning unhealthy patients into healthy ones that doesn't, in aggregate, cost more than the theoretical savings.
> Why would Amazon not aggressively track outcomes, measure every dimension of provider care, and ruthlessly fire / downgrade low performers to maintain the above goals?
Again, this is an incredibly technocratic approach to a field (care delivery) that requires a different approach. It works if you assume that negative outcomes are the fault of individual providers ("low performers") and that individual providers have both the agency and resources to solve arbitrary cases. In practice, that's almost never the case: there are way too many confounding factors that impact clinical outcomes, and providers are quite limited in their power compared to other players: the insurance companies, the administration of the system they work within, and most importantly, the patients themselves.
Honestly, the only way you could meaningfully "fire low performers" is to identify patients who aren't performing well and then drop them. Which is, in a way, what the health insurance system does... but we tend to consider that an antipattern.