The owner of a three-physician pediatric practice forwarded us a slide from a conference. It said the median days in accounts receivable for physician practices was in the low thirties. Her own month-end report said 47. By the time she called, she had already asked the billing manager for a plan to cut the number by a third and started wondering whether to outsource. Two questions later the picture changed: her report counted credit balances in total AR and used charges from a slow month as the denominator, and the survey she was comparing against was mostly hospital-owned multispecialty groups. Recomputed the survey's way, her practice was at 38, unremarkable for a pediatric practice with a heavy Medicaid managed care mix.
Learning how to benchmark your medical practice against survey data is mostly learning what the survey measured, and then measuring the same thing. Done that way, a benchmark tells you where to look. Done the usual way, by comparing a number from your system against a number from a slide, it starts projects that were never needed and misses the ones that were.
A glossary line. A benchmark is a reference value from a group of comparable practices. The median is the middle value when all responses are sorted, and it is what most surveys report because a few large outliers drag the mean around. A percentile tells you what share of respondents fall below a value. FTE means full-time equivalent, so a physician working three days a week counts as 0.6.
Key takeaways
- A benchmark is only usable when you know its definition, its cohort and its sample size, and you compute your own number the same way.
- Six metrics cover most of what an independent practice needs: days in AR, net collection rate, first-pass denial rate, cost to collect, support staff per FTE physician and total operating cost as a share of revenue.
- Specialty, payer mix, ownership and the use of in-house ancillaries move every one of those numbers, so compare against the closest cohort you can find, not the all-practice line.
- The median is not a goal; a gap matters only when the dollars behind it are worth the project it would take to close.
- Your own trend over eight quarters tells you more than any single comparison with someone else.
Where the survey data comes from and what each source measures
The Medical Group Management Association publishes the largest cost, revenue, productivity and staffing datasets for physician practices, drawn from member surveys and reported by specialty, ownership and region. Specialty societies run their own surveys, often with better cohort matching and smaller samples. State medical societies publish regional data. Practice management and clearinghouse vendors increasingly publish aggregated metrics from their own customer bases, which are timely but reflect whoever uses that vendor. And since March 2026, payers regulated under the CMS prior authorization rule have posted their own approval, denial and turnaround metrics, a benchmark of the payer's performance rather than yours.
Each source has a methodology document. Read it before the tables. It tells you how the survey defined AR (gross or net of credit balances), what it counted as a denial (first-pass rejections, adjudicated denials or both), whether cost to collect included the physician's time, how it converted part-time staff to FTEs and, critically, how many practices answered each question in your specialty. A row built on 14 respondents is an anecdote with a decimal point; we treat anything under 30 as directional at best.
The metrics worth comparing and how each one is computed
Practices can benchmark a hundred things. Six do most of the work, and the discipline is to compute each exactly as the survey did before looking at the survey's number.
| Metric | How to compute it the standard way | What moves it | The usual trap |
|---|---|---|---|
| Days in AR | Total AR (net of credit balances) divided by average daily gross charges over the last 90 days | Payer mix, authorization-heavy specialties, charge lag | Using one month of charges; including credits; counting patient AR differently from the survey |
| Net collection rate | Payments divided by (charges minus contractual adjustments), for a matched period at least 120 days old | Write-off discipline, bad debt, payer mix | Comparing recent months whose claims have not finished adjudicating |
| First-pass denial rate | Claim lines denied on first adjudication divided by lines submitted | Front-end eligibility and authorization, coding quality | Mixing clearinghouse rejections with payer denials; counting claims instead of lines |
| Cost to collect | Total billing cost (staff, software, clearinghouse, outsourcing fees, postage) divided by total collections | Outsourcing, volume, automation | Excluding the practice manager's and physicians' time; comparing an outsourced practice with in-house cohorts |
| Support staff per FTE physician | Total non-provider FTEs divided by FTE physicians (some surveys include NPPs in the denominator) | Ancillaries, NPP use, outsourcing | Counting outsourced billing as zero staff; forgetting the denominator definition |
| Total operating cost as a share of revenue | Operating cost excluding physician compensation divided by net revenue | Rent, staffing, ancillaries, specialty | Different treatment of owner compensation and distributions |
Two productivity metrics are worth adding for physician conversations: work RVUs and encounters per FTE physician. Both are sensitive to how the survey counted NPP and shared visits; a physician whose wRVUs look low may be supervising two nurse practitioners whose work is billed under their own numbers.
A worked example, computed the survey's way
Back to the pediatric practice. Total AR on the report was $412,000. Credit balances (patient overpayments and payer overpayments awaiting refund) were $28,000, so AR net of credits was $384,000. Gross charges over the previous 90 days were $1,080,000, or $12,000 a day. Days in AR: $384,000 divided by $12,000, or 32. Her report had used $412,000 over a single month of $265,000 in charges (about $8,800 a day), which produced 47. Same practice, same day, fifteen days apart depending on the arithmetic, and only one of the two methods matched the survey.
Net collection rate for the same practice, using the six months ending 120 days ago so that nearly all claims had adjudicated: payments of $2,940,000 against charges of $4,700,000 less contractual adjustments of $1,640,000, which is $3,060,000 of collectible revenue. The rate is 96.1 percent. The 3.9 percent gap, about $119,000 over six months, is the number worth investigating, because it is what was collectible and not collected: timely filing write-offs, unpaid patient balances sent to collections, small-balance adjustments. That is a more useful finding than any comparison with a slide.
First-pass denial rate needs the clearinghouse and the practice management system to agree on definitions. This practice was counting clearinghouse rejections (claims that never reached the payer) as denials, which inflated the rate to 14 percent; payer denials alone, by line, were 7.8 percent. Surveys count this differently, and the methodology document is the only way to know which one you are comparing against.
Adjusting for specialty, payer mix and structure
Surveys report by specialty for a reason. A surgical practice carries higher days in AR because of authorizations and global periods; a primary care practice with a large Medicaid managed care panel carries a lower cost per encounter and a higher denial rate; a dermatology practice with cosmetic revenue has a patient AR profile no survey captures well. Compare against your specialty's line first, and if the sample is thin, against the closest specialty with a similar visit mix rather than the all-practice line.
Ownership matters as much as specialty. Hospital-owned practices allocate overhead differently, often carry system-level billing costs that never appear in the practice's numbers, and tolerate longer AR because cash flow is the system's problem. Independent practice cohorts are the right comparison for an independent practice, and most surveys let you filter for them. Size matters for staffing ratios, since a two-physician practice cannot employ 0.4 of a credentialing coordinator. Ancillaries matter for cost ratios, since in-house imaging or a lab raises both costs and revenue. Region matters for compensation and rent and very little for revenue cycle metrics. When we prepare a comparison, we write down these five facts about the practice and find the survey cut that matches the most of them, accepting a smaller sample to get a closer cohort.
The traps, in the order we see them
The first trap is the definition mismatch, which the worked example shows. The second is treating the median as a goal. Half of well-run practices are on each side of it by construction, and a practice at the 60th percentile on days in AR with a 98 percent net collection rate has nothing to fix; the money arrives a little slowly and almost all of it arrives. The third is the mean-versus-median confusion, usually when a vendor's marketing quotes an average that a few outliers have pulled.
The fourth is self-selection: practices that answer surveys tend to be the organized ones, so being below the survey median is not the same as being below average. The fifth is gaming your own number: a practice that writes off old AR aggressively will show beautiful days in AR and a collapsing net collection rate, which is why the two must always be read together. The sixth is comparing staffing ratios without adjusting for outsourcing; a practice that outsources billing has fewer staff and a higher cost to collect, and only the combination is meaningful. The seventh is benchmarking once. The same six metrics tracked quarterly for two years, alongside the survey values, show whether the practice is drifting, and drift is what benchmarks are actually good at catching.
Turning a gap into a project worth doing
When a gap survives the definition check and the cohort check, price it before you act on it. Days in AR of 38 against a cohort median of 33 on $12,000 a day of charges is $60,000 of cash arriving later than peers, once, not every year; that is a working capital question and may be worth a modest project. A net collection rate of 96.1 percent against a cohort at 98 percent on $6 million a year of collectible revenue is about $114,000 of revenue not collected, every year, and that is worth a real project. A support staff ratio 0.3 above the cohort in a practice with in-house imaging may not be a gap at all. Pick the gap with the most dollars behind it, run one project per quarter, and re-measure. A revenue cycle audit is, in the end, a structured version of this exercise with the definitions already settled.
Questions we hear
We cannot afford the survey subscription. What can we use?
Specialty society surveys are often included in membership, state medical societies publish regional summaries, and your practice management and clearinghouse vendors usually publish aggregate metrics for their customers at no charge. The most useful benchmark you already own is your own history: eight quarters of the six metrics, computed consistently.
Our billing company reports a 99 percent clean claim rate. Is that a benchmark?
It is a clearinghouse acceptance rate, which measures whether claims passed format edits, not whether payers paid them. Ask for the first-pass denial rate by line and the net collection rate for a period at least 120 days old. Those are the numbers a survey would compare.
How often should we re-benchmark?
Compute your own six metrics every quarter and compare against survey data once a year when the new edition is released. Surveys lag by a year and your quarterly numbers move with the season, so more frequent external comparison adds noise, not information.
What to do this week
- Write down the practice's specialty, ownership, FTE physician count, payer mix and in-house ancillaries; that is the cohort you are looking for.
- Compute days in AR net of credit balances over 90 days of charges, and net collection rate for a six-month period at least 120 days old.
- Get the methodology document for whichever survey you plan to use and match your definitions to it, line by line.
- Find the survey cut that matches the most of your five facts and note its respondent count before you read the numbers.
- Price the largest gap in dollars per year and decide whether it is worth a project this quarter.
- Set up a one-page quarterly sheet with the six metrics and the survey values, and keep it for eight quarters.
