Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

This is not necessarily easy. For example, one of the metrics that patients in my field often advocate for is a stochastic, noisy signal at the end of a long causal chain. It's an important metric, but it's one that's very hard to actually use.

But it's one that feels straightforward: Death.



> For example, one of the metrics that patients in my field often advocate for is a stochastic, noisy signal at the end of a long causal chain. But it's one that feels straightforward: Death.

Which is why this is a dumb metric.

You don't need to measure time to death. You need to measure time to response and time that the patient got with the doctor.

And then, build systems that minimize mttr and maximize time with patient.

To minimize mttr, hire more doctors who can be available.

To maximize time with each patient, hire more doctors.

Ultimately you need to have the right intent to begin with. And the right intent is to not measure the time to death. It is to measure the time to help and to make sure doctors get all the help in minimizing the time to help.


Feel free to start with HEDIS and then move along to whichever healthcare quality metric group strikes your fancy next!

https://www.ncqa.org/hedis/measures/


What happens when doctor attention is negatively correlated with quality of life?


A very inspiring series on data driven approach to management discusses this: https://commoncog.com/goodharts-law-not-useful/




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: