Charges are not revenue and revenue is not cash. A forecast built on expected reimbursement per payer — contracted rate, historical payment lag, expected denial rate — beats one built on billed charges every time, because billed charges bear no relationship to what arrives.
Practices forecast on charges because charges are the number they have. It is also the number least connected to what will land in the bank.
Three different numbers
Charges are what you billed. Largely a function of your fee schedule, which you set, and mostly disconnected from what anyone will pay.
Expected reimbursement is what your contracts say those services should pay. This is the number worth forecasting from.
Cash is expected reimbursement minus denials, minus what patients do not pay, shifted forward by however long each payer takes.
Forecast from charges and you will be wrong by the size of your contractual adjustments — which is to say, badly.
Building one that works
- Start with expected reimbursement per payer, from your fee schedules. If you cannot produce it, that is itself a finding.
- Apply each payer's actual lag, from your own history rather than their stated turnaround.
- Subtract a realistic denial and write-off rate, again from your own numbers.
- Model patient responsibility separately. It collects at a different rate and over a longer period.
What the exercise reveals
Usually two things. First, patient responsibility is a larger share than anyone assumed and collects far more slowly. Second, one payer is materially slower than the rest and is quietly setting the practice's cash position.
Neither is visible in a charge-based forecast, which is why the forecast is worth building even in a month you do not need it.
Build it from claims already submitted
The next ninety days of cash is mostly determined already — it sits in claims submitted and not yet paid. Ageing those by payer with each payer’s typical lag gives a forecast grounded in fact rather than assumption.
Only beyond that horizon do you need volume assumptions, and those are the part everyone argues about while ignoring the part that is already knowable.
Discount honestly
Apply your real denial rate and your real write-off rate, not the ones you would like. A forecast that assumes everything pays is not a forecast; it is the charge master with dates on it.
What the exercise usually surfaces
Concentration. Most practices discover a larger share of cash depends on one or two payers than they assumed, which reframes a contract negotiation and a slow-pay period as strategic questions rather than billing annoyances.
It also surfaces how much of the forecast now depends on patients paying, which behaves nothing like a payer remittance.
Common questions
- How do I forecast cash flow for a medical practice?
- Start from expected reimbursement per payer rather than billed charges, apply each payer’s historical payment lag, and discount for the denial and write-off rate you actually experience.
- Why can I not forecast from billed charges?
- Charges are a list price nobody pays. The contracted allowable, the patient portion and the denial rate all sit between the charge and the deposit.
- What payment lag should I assume?
- Your own, measured per payer. Lags vary widely between commercial plans, Medicare and Medicaid managed care, and an average across all of them forecasts none of them.
- How far ahead can a practice forecast?
- Ninety days with reasonable confidence, because most of that cash is already in claims you have submitted. Beyond that it depends on volume assumptions.
- What does the forecast usually reveal?
- That the practice is more exposed to one or two payers than anyone realized, and that a single payer slowing down would be a cash event rather than an inconvenience.
Denials Piling Up?
We handle the revenue cycle end to end — coding by certified coders, claim submission, denial management and appeals, and A/R follow-up, with six reported numbers every month.