Stream: implementers
Topic: resource for non-probabalistic risk scores
Joanna Gaskill (Jun 07 2017 at 21:32):
I have a need to be able to represent a variety of risk scores that don't correlate at all to a % likliehood, for example, I need to represent a care management risk stratification that assigns all patients to a numeric value from 0 (no benefit to complex case management) to 5 (very high risk individual with high likelihood could benefit from care management). We also do similar concepts for numeric stratification for provider outreach and other initiatives, as well as representing numeric risk scores that represent cost-adjustments based on a patient's overall risk (such as the Medicare HCC risk score or Verisk scores). I initially thought RiskAssessment would be the correct resource to use but I read some info suggestion Observation might be preferable. I think the data would fit better in Observation but was also curious how others are solving this problem so we don't do something too novel.
Lloyd McKenzie (Jun 07 2017 at 21:35):
Either is permissible. RiskAssessment isn't terribly mature yet. If all you need is a simple score, Observation is probably more appropriate.
Last updated: Apr 12 2022 at 19:14 UTC
