Net Collection Rate vs Gross Collection Rate

Gross Collection Rate (GCR) = Total Payments / Total Gross Charges. Gross charges are the practice's full fee schedule charges — what the practice would collect if every payer paid its full sticker price (which no payer does). Net Collection Rate (NCR) = Total Payments / (Total Gross Charges - Total Contractual Adjustments). Contractual adjustments are the dollar differences between billed charges and the contracted allowed amount — the agreed-upon write-off baked into every payer contract. NCR therefore measures payments as a percentage of what was actually collectable.

Read every published figure against the definition it was published under, and against its date. The AAFP calls this metric the adjusted collection rate and states the calculation as payments (net of credits) divided by charges (net of approved contractual agreements), multiplied by 100 and calculated over a 12-month time frame (AAFP, Managing practice finances, read 17 September 2026). That is the same arithmetic as NCR above — the label differs, which is why the same target gets quoted under two names. HFMA's MAP Keys publish metric definitions and equations only, with no target values attached, so a number presented as "the HFMA net collection rate benchmark" has nothing behind it. MGMA's collection-rate medians sit inside licensed DataDive products and are not free to quote, which is why this page states the blank instead of relaying a number.

GCR depends almost entirely on two factors a billing team does not control: the practice's fee schedule and the payer mix. That is arithmetic, not a survey finding. If contracted payers pay close to the Medicare allowable and the practice bills at 200% of that allowable, payments divided by charges lands near 50% no matter how well the billing runs (100 / 200). Bill the same work at 150% and the same collected dollars produce a GCR near 67% (100 / 150); at 125%, near 80% (100 / 125). Neither number reflects billing performance — they reflect fee schedule design. Likewise, a practice with a heavy Medicaid payer mix will have a structurally lower GCR than a practice with mostly commercial payers, regardless of billing operations. Reporting GCR to a board or to the practice owner without that context produces incorrect inferences about billing health. GCR is still worth watching for what it does measure — whether the fee schedule sits above every contracted allowable, so no charge is accidentally billed below what a payer would have paid.

NCR's denominator subtracts contractual adjustments before computing the rate, which neutralizes both the fee schedule effect and most of the payer mix effect. The remaining denominator reflects only what the payer agreed to pay plus what the patient owes — the universe of dollars the practice could legitimately collect. NCR isolates one question: did the practice collect the dollars it was contractually owed, or did it leave revenue on the table through denials, write-offs, bad debt and unworked A/R? The AAFP describes the same metric as showing "how much revenue is lost due to factors in the revenue cycle such as uncollectible bad debt, untimely filing and other noncontractual adjustments" (AAFP, Managing practice finances, read 17 September 2026).

What NCR does not tell you: how fast the money arrived (days in A/R), how much rework it took to arrive (denial rate, first-pass resolution), or whether the fee schedule is set correctly (GCR). A practice can post a 96% NCR and still be financing a 70-day A/R. Report it beside those metrics, not instead of them.

The naming is not standardized across the industry, and that is the single biggest source of mismatched benchmarks. Three things get called a collection rate:

• Gross collection rate — payments / gross charges. Denominator is sticker price.
• Net collection rate — payments / (gross charges - contractual adjustments). Denominator is what the contracts allowed. This is the metric the AAFP publishes under the name adjusted collection rate: payments net of credits over charges net of approved contractual agreements.
• A bad-debt-adjusted variant — some organizations subtract bad debt and charity write-offs from the denominator as well. It necessarily reports higher than NCR, because it removes from the denominator the dollars the practice has already given up on.

Before comparing your number to anyone's target, confirm which of the three the target was published for, which adjustment categories its denominator removes, and over what period. The AAFP figure is published for the second definition and calculated over 12 months. The third has no published target that we could source, and it is the one to be most careful with internally: a practice that aggressively writes off difficult patient balances improves it while masking a self-pay collection failure. Keeping bad debt inside the denominator is what keeps the metric honest about the patient-responsibility dollars.

Four rules for an NCR you can defend in a board meeting.

1. Use a 12-month window, not a single month. Single-month NCR swings on payment timing alone. The AAFP recommends a 12-month time frame for this calculation (AAFP, Managing practice finances, read 17 September 2026); a rolling 90-day view is useful for spotting a trend, not for benchmarking.
2. Align the numerator and denominator on the same basis. Match cash posted in the period against contractual adjustments posted in the same period. Mixing accrual-basis charges (when the service was rendered) with cash-basis payments (when payment posted) distorts the result; the size of the distortion depends on your own lag, so measure it rather than assuming a rule of thumb.
3. Exclude non-revenue cash from the numerator — refund reversals, credit-balance recoupments and capitation payments unrelated to fee-for-service charges.
4. Categorize your adjustments. The calculation is only as good as the split between contractual adjustments (which leave the denominator) and noncontractual write-offs such as untimely filing, missed authorizations and bad debt (which stay in it). Miscoding the second as the first is the most common way a reported NCR flatters the practice.

Worked example — illustrative figures, arithmetic reproduces. Twelve months: gross charges $4,000,000; contractual adjustments posted $1,600,000; payments posted $2,160,000, of which $60,000 is a credit-balance recoupment that does not belong in the numerator.

• Collectable base: $4,000,000 - $1,600,000 = $2,400,000
• Numerator: $2,160,000 - $60,000 = $2,100,000
• NCR: $2,100,000 / $2,400,000 = 87.5%
• GCR on the same data: $2,100,000 / $4,000,000 = 52.5%

The gap between 87.5% and 52.5% is entirely fee schedule design. The gap between 87.5% and the AAFP's 95% floor is the part that is worth a work plan: $2,400,000 x 0.95 = $2,280,000 expected, against $2,100,000 collected — $180,000 sitting in denials, write-offs and aged A/R. Run the same three lines on your own export before accepting any vendor's version of your collection rate.

Structurally, NCR should compress across specialties in a way GCR does not: subtracting contractual adjustments removes the fee-schedule and payer-mix differences that make GCR incomparable between two practices. That is a property of the formula, and it is why NCR travels better as a comparison metric.

What we cannot give you is a sourced table of NCR medians by specialty. No freely public source publishes one — MGMA's collection-rate medians are inside licensed DataDive products, the AAFP publishes a single practice-management target with no specialty cut and no sample behind it, and HFMA's MAP Keys publish definitions rather than values. Any page presenting specialty-by-specialty NCR medians without naming a published dataset is inventing them.

What is defensible without a benchmark: a surgical practice carrying more global-period and bundling disputes, or a behavioral health practice carrying more medical-necessity denials, will see more of its collectable dollars contested — and contested dollars are where NCR leaks. Compare your practice to its own prior 12 months first; that comparison needs no external dataset.

When NCR sits below the AAFP's 95% floor, four sources are usually responsible, and they should be investigated in order. First, denial rework backlog — claims denied 90+ days ago that have not been worked or appealed are simply not collected, and they show up as the gap between the contractually allowed amount and actual payments. Second, missed timely filing on initial submissions or appeals. Filing limits are set by each payer's contract and have to be read per payer; the one that is fixed in regulation is Medicare's, where a claim must be filed no later than one calendar year after the date of service (42 CFR 424.44, read 17 September 2026). Untimely filing is a noncontractual write-off: it stays in the NCR denominator permanently. Third, self-pay balances aging past the point where in-house collection effort works — track your own recovery curve by aging bucket rather than assuming a published rate, because that curve is practice-specific. Fourth, posted contractual adjustments that exceed the actual contract terms — usually because of fee schedule load errors or rate updates that did not flow through to the practice management system, which understates the denominator and flatters the result.