Skip to content
ExplainerEconomics

Revenue per provider: the benchmark that hides the problem

It is the most quoted number in practice management and the least diagnostic. What to look at instead when the figure looks wrong.

6 min readGrowUrology editorial

An old desktop calculator with a blank paper roll curling across a wooden desk in raking light.

In short

  • Revenue per provider blends volume, coding accuracy, payer mix, collections performance and case mix into one number — so a low value tells you nothing about which one is broken.
  • Decompose it before acting: encounters, work per encounter, payment per unit of work, and the share of billed work actually collected.
  • Compare against your own trend before comparing against an external benchmark.

Revenue per provider is a summary statistic that four different problems can produce. Acting on it directly is guessing.

Decompose before you diagnose

The figure is the product of roughly four things. Take them one at a time:

  • Encounters per session — a scheduling and template question.
  • Work per encounter — case mix, procedure share and E/M level distribution.
  • Payment per unit of work — payer mix, contracted rates and site of service.
  • Collected share of billed work — the revenue cycle: denials, underpayments, write-offs and patient balances.

Each has a different owner and a different remedy. A practice with strong volume and weak collections needs nothing that a practice with weak volume and clean collections needs.

On external benchmarks

Benchmark data is genuinely useful and genuinely easy to misuse. Definitions differ between sources, sample composition differs by region and practice size, and a percentile is not a target. Use benchmarks to generate hypotheses about where you differ, then test the hypotheses against your own decomposition.

We do not publish benchmark figures we have not licensed and verified. Where you see benchmark structures on this site, they show the metrics and definitions we intend to publish, with sourcing, once validated data is in place.