Why physician compensation runs break down

A physician compensation model is a policy. The monthly payout is a calculation. In most multi-location groups, the policy is clear enough, but the calculation cannot be reproduced after the fact.

That matters because a comp run is only defensible if you can re-create it. A good test is simple: take last quarter, apply a retroactive adjustment, and see whether you can re-run the period and get the same number that was actually paid. If you cannot, the problem is usually not the plan itself. It is the data flow underneath it.

This is the part many owner-operators discover only after acquisitions, new locations, or a growing mix of wRVU, collections, base-plus-production, and profit-share plans. The spreadsheet may still “work” in the sense that everyone gets paid. It stops working when a provider asks where a number came from and finance cannot trace it back cleanly.

What the calculation actually needs

To pay providers correctly, you need three layers, not one.

First, you need an attribution layer. That answers who earned the production, which often is not the same as who billed it, who supervised it, or which location booked the visit.

Second, you need a period rule. That decides whether you pay on production as it happens, on collections when cash arrives, or on a hybrid basis with a lag and a true-up.

Third, you need a restatement rule. That says what happens when a payer recoups money, a claim is corrected, or a late charge posts after the month closed.

Most compensation spreadsheets try to do all three at once. That is why they become fragile. A workbook can apply a formula, but it is a poor system for managing identity, time, and retroactive change across several source systems.

The five places the number breaks first

1. Incident-to work and APP supervision

Incident-to billing means a service provided by clinical staff can be billed under a supervising physician or other eligible professional when the billing rules are met. Advanced practice providers (APPs) such as nurse practitioners and physician assistants add another layer, because supervision, billing, and economic credit are not always the same thing.

That is where attribution goes wrong. The claim may land under one NPI (National Provider Identifier), while the group believes the production belongs to another clinician. If the plan pays on billed production, the spreadsheet will reward the billing physician. If the plan pays on earned work, the group has to define who gets credit when supervision, protocol, or team-based care is involved.

This is one reason owner-operators need more than a summary export. They need a rule that tells the system how to assign credit before the comp formula runs.

2. Hygiene, imaging, and other ancillary production

In dental and specialty groups, the treating provider often is not the only person generating revenue. Hygiene, imaging, labs, and other ancillary services may be tied to a location, a supervising doctor, or a separate clinical team.

If the comp model only looks at the provider column in the practice-management system, it misses production that should be counted. If it pushes too much into a supervising provider’s column, it can overstate that person’s contribution. Either way, the numbers may be internally consistent and still wrong.

3. Production versus collections timing

Production is earned when the work is performed or billed, depending on the plan. Collections are cash received, often weeks or months later. In many groups, there is a 60- to 120-day lag between the service date and the cash hitting the bank.

That lag matters because a comp run based on collections is always looking backward. A provider can have a strong month of treatment and a weak month of cash, or the reverse. If the group does not make that timing explicit, disputes follow quickly.

This is the core of production versus collections compensation: they are different economic measurements. A monthly workbook has to know which one is being used, at what date, and how to handle late-arriving cash.

4. Multiple systems and tax IDs after acquisition

Once a group grows by acquisition, the compensation process often has to combine several practice-management systems, several EHRs, and multiple tax IDs. That creates duplicate provider identities, mismatched location codes, and different definitions of what counts as production.

A spreadsheet can merge exports, but it cannot reliably normalize a dozen source definitions without a strong data model behind it. The more inherited plans and systems a group has, the more the monthly number becomes a manual reconciliation exercise instead of a repeatable calculation.

5. Refunds, recoupments, and retro adjustments

A payout that looked right at close can become wrong later. Payers reverse claims. Patients get refunds. Medicare or commercial plans issue recoupments. A corrected encounter can change the original month after the comp statement was already paid.

This is where defensibility shows up. If you cannot restate the prior period and explain the delta, then the comp run is not really closed. It is just frozen.

What Stark changes, and what it does not

Stark law constrains how certain physician arrangements are structured, especially around designated health services. It does not tell a group how to write a bonus formula. It does shape what the calculation has to respect.

For many groups, the practical implications are these: compensation based on personal productivity and profit sharing may be allowed in specific forms; compensation tied to personally performed services matters; and incident-to credit can follow the treating physician under the rules the group uses to operationalize the plan. Since January 1, 2022, the single-DHS-pool rule has also changed how some groups structure design calculations.

The important point is that compliance pressure pushes the work into data and attribution, not just policy language. A comp plan can be legally well drafted and still fail operationally if the spreadsheet cannot show how the number was built.

When a spreadsheet is enough, and when it is not

Some groups do not need a new system. Others do. The question is not whether spreadsheets are bad. It is whether they are still adequate for the number of moving parts.

If several rows fall in the right-hand column, the group is probably past the point where another tab will solve the issue. At that stage, the business needs a calculation layer that can absorb source-system differences, preserve history, and explain the result.

What a defensible comp run looks like

A defensible run is not just a payout total. It is a chain of evidence.

• Every dollar can be traced back to a claim line, encounter, collection event, or other source transaction.
• The attribution rule is explicit: billing NPI, supervising provider, treating provider, location, and tax ID are all defined.
• Providers can see a running statement before quarter-end, so the true-up is not a surprise.
• Prior periods can be re-run after a retro adjustment and compared with what was already paid.
• Plan changes, manual overrides, and exceptions leave an audit trail.

That last point matters more than it sounds. Once a finance lead is the only person who can explain why a number changed, the process has already become a key-person risk.

Build versus buy: where existing systems help, and where they stop

Practice-management systems, payroll systems, and analytics tools can each solve part of the problem. A PMS can hold encounter and billing data. Payroll can issue the payment. An analytics layer can help standardize reporting.

What they usually do not do by default is manage provider compensation as a full calculation system across multiple source systems and plans. They often stop at the export, the dashboard, or the payroll entry.

That is where provider compensation management software becomes relevant. If the group needs wRVU tracking, recurring restatements, provider-visible statements, or a clear calculation trail across locations, the software should do more than summarize totals. It should make the run reproducible.

For some groups, that means buying a compensation engine. For others, a reporting layer over the PMS plus a tighter payroll workflow is enough. The decision comes down to how many exceptions the business has, how often the numbers must be explained, and how much manual reconciliation the owner is willing to tolerate.

A practical test before you buy anything

Before adding a new tool, run the prior quarter again with one retroactive adjustment or late correction. Then answer four questions:

• Can finance reproduce the amount paid without using tribal knowledge?
• Can the provider see how the number was built?
• Can the group explain how a correction affects the open period and the closed period?
• Can the same logic handle another acquisition without rewriting the workbook?

If the answer is no to most of those questions, the issue is probably not the physician compensation model itself. The issue is that the calculation no longer fits in a spreadsheet owned by one person.

That is the point where a group should look at healthcare operations support and, if it needs a purpose-built workflow, a system that can handle reporting, payroll, and multi-location provider payroll together. In one example, Better built reporting, payroll, time-management, and business-operations tooling for Apex Dental Partners, a group that grew beyond 45 practices and needed to pull operating and payroll data across many locations and systems into numbers an owner could stand behind. The useful lesson is not the industry or the size. It is that the calculation layer has to keep up with the shape of the business.

FAQ

How is physician compensation determined?

At a high level, it is determined by the plan terms and the measurement base: salary, wRVUs, production, collections, profit share, or a combination. The harder part is not the formula. It is attributing the right work to the right provider and applying the right timing rule.

What is physician compensation based on wRVU?

wRVU stands for work relative value unit. It measures physician work rather than dollars collected. A wRVU-based model can be cleaner than collections-based pay when cash timing is noisy, but it still depends on correct attribution and a reliable source for the work units.

Do associate dentist compensation plans have the same problem?

Yes, often with fewer billing layers but similar attribution issues. Hygiene, imaging, supervising roles, and location-level production can all make associate dentist compensation hard to reproduce from a simple export.

When is a physician compensation calculator enough?

A physician compensation calculator is enough when the plan is simple, the source data is consistent, and there are few retroactive changes. Once a group has multiple systems, APP supervision, or frequent restatements, a calculator usually becomes a front end to a more serious calculation process.

What should I ask before buying physician compensation software?

Ask whether it can re-run a prior period after a retro adjustment, show provider-visible statements, handle multiple source systems, and preserve an audit trail for overrides. If it cannot do those things, it may improve reporting without fixing the comp run.