A physician owner showed us her billing company's monthly report. It listed a 98 percent clean claim rate, 27 days in AR, and a 99.2 percent net collection rate. She wanted to know why, with numbers that good, the practice had drawn on its line of credit twice that spring. The answer took an afternoon to find, and the numbers were not lies. They were just measured in ways that made them look good. The clean claim rate counted only claims that reached the payer. Days in AR excluded anything over 120 days because those balances had been "reserved." The net collection rate used the same month's charges and payments, so it moved with the calendar rather than with performance.
This is why we think every practice owner should understand revenue cycle KPIs well enough to ask one hard question about each. Not to run the billing office. To know when a report is telling the truth.
A glossary line to start: a KPI (key performance indicator) is a number tracked over time to judge whether a process is working. In revenue cycle work the useful ones are ratios with a defined numerator, a defined denominator and a defined time window. If any of the three is missing from the report, the number is decoration.
Key takeaways
- Every revenue cycle KPI needs a stated numerator, denominator and time window; ask for all three before you trust the number.
- Days in AR and net collection rate are the two numbers that describe the whole cycle; clean claim rate and denial rate describe the front end.
- Benchmarks are useful for direction, not for judgment; a 32-day AR is good for orthopedics and mediocre for a practice with mostly Medicare and copays collected at the desk.
- The most common way a KPI is gamed is by shrinking the denominator, so ask what was excluded.
- Trend beats level: a metric moving the wrong way for three months tells you more than one month against a benchmark.
The ten KPIs, with formulas
| KPI | Formula | Commonly quoted target | How it gets gamed |
|---|---|---|---|
| Clean claim rate | Claims accepted by the payer with no edits or rejections on first submission divided by all claims submitted | 95 percent or higher | Counting only claims that passed the clearinghouse, so internal scrubber rejections vanish |
| First pass resolution rate | Claims paid (or correctly adjudicated) on first submission divided by all claims submitted | 90 percent or higher | Counting partial payments as resolved |
| Denial rate | Claim lines denied divided by claim lines submitted, by count or by dollars | Under 5 to 10 percent, depending on specialty | Excluding "soft" denials such as requests for records |
| Days in AR | Total outstanding AR divided by average daily charges (last 90 days of charges divided by 90) | Under 35 to 40 days for most office practices | Excluding old balances, credit balances netted against debits, or using gross charges when AR is stated net |
| AR over 90 days | AR older than 90 days from date of service divided by total AR | Under 15 to 20 percent | Aging from last billed date instead of date of service |
| Net collection rate | Payments divided by (charges minus contractual adjustments), using a window at least 90 days old so claims have adjudicated | 95 to 99 percent | Same-month charges and payments; counting bad-debt write-offs as contractual adjustments |
| Gross collection rate | Payments divided by gross charges | No universal target; depends entirely on the fee schedule | Not gamed, just misread; a low number often means high chargemaster prices, not poor collections |
| Charge lag | Days from date of service to charge entry, median and 90th percentile | Under 2 to 3 days | Reporting the average, which hides the one provider at 14 days |
| Point-of-service collection rate | Patient payments collected at the visit divided by patient responsibility identifiable at the visit (copays, known deductibles) | Over 90 percent of copays; deductibles vary | Using only copays in the denominator |
| Cost to collect | Total billing cost (staff, software, clearinghouse, vendor fees, postage) divided by payments collected | Often quoted at 3 to 6 percent for physician practices | Leaving out the front desk time spent on eligibility and collections |
Two numbers describe the whole cycle
If you only look at two, look at days in AR and net collection rate. Days in AR tells you how fast money moves. Net collection rate tells you how much of the money you were entitled to actually arrived. A practice can have excellent speed and poor yield (fast write-offs), or excellent yield and poor speed (everything gets paid eventually, after three appeals). Neither is good, and each is invisible if you only watch the other.
Days in AR is straightforward if the inputs are honest. Take total AR as of the last day of the month. Take charges for the prior 90 days and divide by 90 for average daily charges. Divide the first by the second. A practice with $412,000 in AR and $1.26 million in charges over the last quarter has average daily charges of $14,000 and 29.4 days in AR. Now ask: does the $412,000 include credit balances (which reduce it), does it include patient balances (it should), and does it include claims over 180 days (it should, until they are written off with a documented reason)?
Net collection rate takes more care. Use a window old enough that the claims in it are finished. We use dates of service from four to seven months ago. For those dates, sum payments received to date and sum charges less contractual adjustments. Bad debt, timely filing write-offs and "unable to appeal" adjustments are not contractual; they are money you were owed and did not get, and they belong in the shortfall. A practice that adjusts denied claims off with a contractual code will show a 99 percent net collection rate while losing 6 percent of its revenue.
A worked example
Take a four-provider family medicine practice. In the first quarter it posted $1.26 million in charges, about 7,600 encounters. Here is what an honest set of KPIs looked like, and what each one said.
- Clean claim rate 91 percent. Roughly 680 claims needed a touch before the payer accepted them. Most were eligibility rejections, which pointed at the front desk, not the billers.
- Denial rate 8.4 percent by line count, 6.1 percent by dollars. The gap between count and dollars meant the denials were mostly small lines: 36415 venipuncture bundling, 96372 with an E/M lacking modifier 25. Small and fixable.
- Days in AR 34. Fine on its face. But AR over 90 days was 24 percent, which meant a slice of old claims nobody was working, propped up by fast Medicare payments on the rest.
- Net collection rate 94.2 percent on dates of service from the prior autumn. The 5.8 percent shortfall was $58,000 for the quarter, of which $31,000 was denials written off without appeal.
- Charge lag median 1 day, 90th percentile 9 days. One provider was signing notes in weekend batches. Nine days of lag on a fifth of the practice's charges is a week of cash every month.
None of these numbers was alarming alone. Together they described a practice that submitted claims quickly, gave up on denials too easily, and had one provider and one front desk process costing it about $10,000 a month. That is the point of KPIs: not the dashboard, the story.
Benchmarks and why to distrust them
Trade groups such as MGMA and HFMA publish benchmarks, and the ranges in the table above are the ones most commonly quoted. Use them to know which direction is good. Do not use them to decide whether your billing team is doing well, because your payer mix, specialty and front desk policies move the numbers more than billing skill does. A practice that collects copays and known deductibles at check-in will have a lower days in AR than one that statements everything, with identical billing staff. A surgical practice with heavy prior authorization will have a higher denial rate than a dermatology practice, with identical coders.
What you can compare fairly is your own practice against itself. Three months of a metric moving the wrong direction is a finding. One month against a benchmark is a conversation.
Questions to ask about any report
When a report lands, we suggest four questions, in this order. What is excluded from the denominator? What date is the aging based on, date of service or last billed? For collection rates, how old is the window? And who produced the number, the system directly or a person in a spreadsheet? None of these are accusations. They are how a report earns trust. If the answers are not on the report, ask that they be added, and if a vendor cannot answer them, that is itself a finding. Our RCM audit starts by recalculating these ten numbers from raw data for exactly this reason.
Questions we hear
Which single KPI should a physician look at every month?
Net collection rate on a lagged window, with the shortfall shown in dollars and split into contractual, denials written off, timely filing and bad debt. It is the closest thing to a single truth about the billing operation. Days in AR is a close second.
Our billing company reports a denial rate of 3 percent. Is that plausible?
It is plausible if the denominator is claims rather than claim lines and the numerator excludes anything that was corrected and resubmitted before the report ran. Ask for the definition. A true initial denial rate below 5 percent in an office practice is unusual and worth verifying, not celebrating.
How often should these be reviewed?
Monthly for the full set, with a twelve-month trend line on each. Weekly for charge lag and unbilled encounters, because those are operational and can be fixed the same week. Quarterly for cost to collect, because staffing and vendor costs move slowly.
What to do this week
- Take your most recent billing report and, for each KPI, write down the numerator, denominator and time window. Mark any you cannot find.
- Recalculate days in AR yourself from the AR aging report and the last 90 days of charges, and compare it with the reported figure.
- Pull the adjustment codes used in the last quarter and separate contractual adjustments from denials, timely filing and bad debt.
- Ask for charge lag as a median and a 90th percentile, by provider.
- Set a twelve-month trend chart for net collection rate and days in AR and put both on the monthly meeting agenda.
