[drak0n1c]

Sounds like an application for ML, to determine which codes frequently coincide per-patient at each provider and then assign those groupings to cross-provider "Treatment XYZ" buckets to enable apples-to-apples comparisons.

[heliodor]

I would think a basic statistical analysis should suffice.

[ejb999]

most software billed as having 'ML' capabilities is just basic statistical analysis anyway - but that doesn't make for good marketing-speak.

[mskar]

Great call, many orgs in health tech use billing/procedure code embeddings to group, just like you're suggesting.

[hugh-avherald]

Calculating a basic median for those groups would be a non-trivial (indeed, probably quite difficult) exercise at this scale.

[dekhn]

Applying ML to health care is a guaranteed path to wealth, and later, insanity.