Here is how the denial conversation usually goes in a practice without a template. The administrator notices that deposits are down. Someone pulls a denial report, which is 40 pages long and sorted by patient name. The billing lead says most of it is eligibility. The physician says the payers are denying everything. Everyone agrees denials are a problem, and nothing specific gets assigned. Six weeks later deposits recover and the report is forgotten.
The fix is not a better report. It is a fixed set of numbers pulled the same way every month, and a meeting with an agenda short enough to finish. We have used the template below in practices of every size, and the value is almost entirely in the consistency. The first month is a baseline. The third month is where the pattern appears.
Key takeaways
- Pull the same eight numbers the same way every month from the 835 remit data, not from the practice management system's denial flag.
- Group denials into six categories by CARC, each with a named owner: eligibility, authorization, coding, medical necessity, bundling, and timeliness.
- The meeting is forty minutes with five people and ends with at most three assigned fixes; a review that ends without fixes was a status update.
- Under 5 percent of lines denied on first pass is where well-run office practices sit; the dollar rate can hide a few expensive denials the count rate misses.
- Rejections are not denials, corrected-and-paid is not free, and quarterly reviews mean every fix takes a quarter to test.
The eight numbers
| Measure | How to calculate | Why it matters |
|---|---|---|
| 1. Initial denial rate (count) | Claim lines denied on first adjudication divided by claim lines submitted, for the month | The headline. Under 5 percent is where well-run practices sit; above 10 percent means process failure |
| 2. Initial denial rate (dollars) | Billed dollars on denied lines divided by billed dollars submitted | A low count rate can hide a few expensive denials |
| 3. Denials by category | Denied lines grouped into six categories using the CARC (see below) | Tells you which department owns the problem |
| 4. Denials by payer | Denial rate for each of the top eight payers | One payer at twice the average rate is a policy or setup issue, not a coding issue |
| 5. Top ten CARC and RARC combinations | Count and dollars for each reason code pair | The specific fixes live here |
| 6. Appeal overturn rate | Appeals decided in favor divided by appeals decided, for the month | Low overturn means you are appealing the wrong things or writing weak appeals |
| 7. Denial aging | Denied lines not yet worked, in buckets of 0 to 30, 31 to 60 and over 60 days | Anything over 60 days is approaching a timely filing or appeal deadline |
| 8. Denial write-offs | Dollars adjusted off with a denial-related code, with the reason recorded | The final cost of the month's denials, and the number that should fall over time |
The six categories
Every denial fits one of six buckets, and each bucket has an owner. Group by the claim adjustment reason code on the remit rather than by whatever free text the billing system offers.
- Eligibility and coverage (CO-27 coverage terminated, CO-22 other payer primary, CO-31 patient cannot be identified, CO-109 not covered by this payer). Owner: front desk. Fix: verification at scheduling and again before the visit.
- Authorization and referral (CO-197 precertification absent, CO-15 authorization number invalid, CO-198 precertification exceeded). Owner: authorization coordinator. Fix: authorization tracking log and a check that the authorized code matches the billed code.
- Coding and claim data (CO-4 modifier inconsistent, CO-11 diagnosis inconsistent with procedure, CO-16 missing information, CO-181 invalid procedure code). Owner: coding. Fix: scrubber edits and coder education on the specific pair.
- Medical necessity and policy (CO-50 not deemed medically necessary, CO-96 non-covered charge, CO-167 diagnosis not covered). Owner: coding and the physician. Fix: documentation templates aligned to payer policy, and appeal with records.
- Bundling and frequency (CO-97 included in another service, CO-236 procedure not compatible, CO-119 benefit maximum reached, CO-151 frequency exceeded). Owner: coding. Fix: NCCI edit review and modifier use where appropriate.
- Timeliness and duplicates (CO-29 timely filing expired, CO-18 duplicate claim). Owner: billing. Fix: work queues by age, and stop resubmitting claims that are pending.
Some codes are ambiguous. CO-16 with a remark code about a missing modifier is coding; CO-16 with a remark about a missing referring provider is front desk. Read the RARC before assigning.
Here is what a month looks like once the numbers are pulled. A four-physician practice submits 4,200 claim lines in January and 310 deny on first pass, a 7.4 percent count rate. The billed dollars on those lines are $61,000 against $690,000 submitted, an 8.8 percent dollar rate, so the denials are slightly skewed toward expensive lines. Grouped by category, 140 lines are eligibility (mostly CO-27 and CO-31, which in January means plan changes), 55 are coding, 40 are authorization, 35 are bundling, 25 are medical necessity and 15 are timeliness or duplicates. By payer, one Medicaid managed care plan is denying 14 percent of lines while the others sit between 4 and 8. The top CARC pair by dollars is CO-197 on a single imaging code at one payer, $9,800 across 12 lines. That is the whole agenda for the meeting: the front desk owns January eligibility, the authorization coordinator owns the imaging code at that payer, and somebody calls the Medicaid plan's provider representative. Three fixes, three owners, and next month's numbers will say whether they worked.
The forty-minute agenda
Attendees: practice manager, billing lead, front desk lead, one physician or the medical director, and the coder. Not the whole staff.
- Five minutes: the eight numbers compared to last month and to the same month last year. No discussion yet.
- Ten minutes: the top three CARC combinations by dollars. For each, the owner explains the cause and proposes one fix.
- Ten minutes: the payer with the highest denial rate. Is it a policy change, a setup error, or a pattern that belongs in the next contract conversation?
- Five minutes: appeals. Overturn rate, anything over 60 days, and whether anything should be escalated to a peer-to-peer or an external review.
- Five minutes: last month's assigned fixes. Done, not done, and did the number move?
- Five minutes: assign this month's fixes with an owner and a date. Three at most.
The last item is the whole point. A denial review that ends without three named fixes was a status update.
Pulling the data
You need the 835 remittance data, not the practice management system's denial flag. The remit carries the CARC and RARC; the system flag is whatever a poster chose. Most systems have a denial report that reads from the posted remit data and can group by CARC; if yours cannot, export the remit detail and group in a spreadsheet. Define "denied" once and write it down: we use any claim line with a zero payment and a CO or PR adjustment other than deductible, coinsurance, copay or contractual obligation. Use the same definition every month or the trend is meaningless.
For the first three months, pull the numbers for the previous twelve months too. The seasonal pattern (eligibility denials in January, timely filing in the quarter after a staff change) is only visible with a year of data.
Mistakes we see
Counting rejections as denials. A clearinghouse rejection never reached the payer and belongs in a different report. Counting a denial that was corrected and paid as a success without noting the cost of the rework. Letting the meeting become a payer complaint session; payers deny predictably, and predictable is fixable. And reviewing quarterly instead of monthly, which means every fix takes a quarter to test.
This template is the backbone of the monthly report our denial management service sends to practices, and the categories match the root cause approach we described in denial root cause analysis versus denial follow-up. A practice that wants a baseline built from its own remit data can start with an RCM audit.
Questions we hear
What denial rate should we aim for?
Under 5 percent by count is realistic for most office-based practices with a working eligibility process. The dollar rate depends on specialty; a surgical practice with a few high-value authorization denials will have a higher dollar rate than count rate. The trend matters more than the target in the first six months.
Our system cannot report by CARC. What do we do?
Export the posted remittance detail to a spreadsheet each month; most systems can produce it even if they cannot report on it. If yours truly cannot, that is a real limitation worth raising with the vendor, because denial management without reason codes is guessing.
Who should run the meeting, the billing lead or the practice manager?
The practice manager. The billing lead presents the numbers, but most of the fixes belong to other departments: the front desk, the authorization coordinator, the physicians. A meeting run by the billing lead tends to turn into the billing lead asking colleagues for favors. A meeting run by the practice manager assigns work. If the billing is outsourced, the vendor presents and the practice manager still runs it, because the vendor cannot fix the front desk either.
What to do this month
- Write down your definition of a denied line and the source report, and do not change either for a year.
- Pull the eight numbers for last month and, if you can, for the previous twelve months as a baseline.
- Map your top twenty CARC and RARC combinations into the six categories and name the owner of each category.
- Schedule the forty-minute review for the same day each month with the five attendees, and send the one-page numbers the day before.
- Leave the first meeting with three fixes, each with an owner and a date, and start the next meeting by checking them.
