I run a four-location practice in Hawaii. Before I built A-Z, I went through three billing companies in five years. Each one reported collections monthly and told me everything was fine. None of them told me that $874,000 in receivables had gone unworked for six years. I found that myself when I finally audited. This framework is what I wish I had before I wrote those checks. Six reports, four questions, 90 minutes per quarter. If your vendor can't produce the answers on the spot, that tells you everything.
Most practice administrators receive a monthly report from their billing vendor. Collections went up or down. A handful of summary numbers.
The report arrives, gets glanced at, and goes into a folder.
That monthly report is the vendor's version of their own performance. They chose which numbers to include, which to exclude, and how to frame the trend. Auditing your billing vendor means pulling the numbers yourself, from your own system, and comparing them against what the vendor reports.
The gap between those two data sets is where lost revenue hides.
Monthly data fluctuates too much to draw conclusions. Seasonal patterns, payer processing delays, and single large claims can swing monthly numbers by 10-15%. Quarterly data smooths the noise and reveals actual trends.
A billing vendor whose denial rate spiked in March might have a reasonable explanation: a payer system outage, a batch of credentialing renewals, a CPT code update. The same vendor whose denial rate has climbed every quarter for three consecutive quarters has a systemic problem they haven't addressed.
Run this audit in the second week of January, April, July, and October, covering the prior quarter's data.
Pull these from your practice management system, not from your vendor's reporting dashboard. The vendor's dashboard shows you their data. Your PM system shows you yours.
Pull the aging report grouped by payer with the standard buckets: 0-30, 31-60, 61-90, 91-120, 120+ days. Calculate the percentage of total AR in each bucket. Compare to last quarter.
A healthy revenue cycle keeps 50-65% of AR in the 0-30 bucket and under 10% in the 90+ bucket. If the 90+ bucket is growing quarter over quarter, claims are aging through the system without follow-up.
Your vendor may tell you they're "working the AR." The aging report tells you whether that work is producing results.
For multi-location groups: pull this report by location. A practice-wide aging summary can mask a location where AR is deteriorating. One office with a 25% 90+ bucket hidden inside a group average of 12% is a problem nobody sees without location-level reporting.
Total denial rate is a starting point. The actionable data is denial rate by denial code and by payer. A 6% overall denial rate could mean 3% across seven payers and 18% with one payer. The one payer is the problem. The overall number hides it.
Pull the top 10 denial codes by dollar volume for the quarter. Compare to last quarter's top 10. If the same codes appear quarter after quarter, root causes aren't being fixed.
Recurring denial codes are a direct measure of whether the vendor is reducing denials over time or just resubmitting them.
Net collection rate (payments received divided by allowed amounts) is the truest measure of billing effectiveness. Pull it by payer because payer-specific collection rates reveal underpayment patterns that an overall average obscures.
A 95% net collection rate overall is healthy. A 95% average that includes 99% from Medicare and 84% from UHC means UHC is a problem.
Your vendor should be able to explain the gap and describe what they're doing to close it. Silence on payer-specific performance is itself a data point.
This is the number of days between the date of service and the date the claim was submitted. Your PM system logs both dates. Pull the average and the distribution.
Claims should be submitted within 48 hours of the encounter. Anything over 5 days indicates a workflow bottleneck. Anything over 14 days is costing you money directly through delayed cash flow and indirectly through increased exposure to timely filing denials.
For multi-location groups, check submission lag by location. A central billing office processing claims for 4 locations may prioritize one location's charges over another. The deprioritized location's claims sit longer, age faster, and hit filing deadlines sooner.
Pull every write-off for the quarter. Categorize them: contractual adjustment (expected), timely filing expiration (avoidable), patient bad debt (partially avoidable), and "unable to collect" (requires investigation). Calculate the non-contractual write-off rate as a percentage of total charges.
Non-contractual write-offs above 3% indicate that recoverable revenue is being abandoned. Timely filing write-offs above 0.5% indicate that claims are expiring in the system.
A billing vendor producing timely filing write-offs every quarter is failing at the most fundamental function of the job.
Pull the count of each E/M code (99211-99215) billed per provider for the quarter. Compare each provider's distribution to the specialty benchmarks published by CMS.
A provider billing 70% of established visits at 99213 when the specialty benchmark is 35% is undercoding. The revenue gap between 99213 and 99214 is $30-$40 per visit.
At 20 visits per day, undercoding one provider costs $57,000-$77,000 per year. A billing vendor that doesn't review coding distribution has no way to catch it.
After you've pulled your own reports, schedule a quarterly review call with your billing vendor. These four questions separate vendors who manage your revenue from vendors who process your claims.
A strong vendor answers with specific codes, specific root causes, and specific actions taken.
"CO-185 was your top denial at $14,200 because Dr. Martinez's enrollment with Aetna lapsed in February. We resubmitted the enrollment on March 3, it was effective retroactively on March 18, and we recovered $11,800 of the $14,200."
A weak vendor answers with generalities. "We had some eligibility issues we're working through." Generalities mean they aren't tracking denials at the code level, which means they aren't fixing root causes.
This question tests whether the vendor monitors their own workflow efficiency. If they can't produce the number without going to look it up, they aren't tracking it. If the lag varies significantly by location, it means the workflow treats some locations as higher priority than others.
This question tests whether the vendor audits payments or just posts them. A vendor who audits payments can produce the variance data instantly.
A vendor who just posts payments will not have this report. The absence of the report is the answer.
For multi-location groups, this is the question that reveals whether all locations are being managed equally. A vendor reporting 96% collection rate practice-wide may be collecting 99% at the flagship office and 88% at the satellite. The satellite's underperformance is subsidized by the flagship's strength in the aggregate number. Location-level reporting exposes this.
The vendor who resists location-level reporting is the vendor who benefits from hiding behind averages. If your billing vendor cannot or will not produce KPIs by location, by payer, and by provider, they are not managing your revenue cycle at the level a multi-location group requires. That limitation has a cost measured in tens of thousands of dollars per year per location.
| Finding | What It Means | Action |
|---|---|---|
| 90+ AR growing quarter over quarter | Follow-up is insufficient. Claims are aging through without being worked. | Demand a specific plan with weekly targets for reducing the 90+ bucket. Measure results at 60 days. |
| Same denial codes appearing for 3+ consecutive quarters | Root causes are not being addressed. The vendor is reacting to denials instead of preventing them. | Request a root cause analysis and a documented prevention plan for each recurring code. |
| Submission lag over 5 days at any location | Claims are sitting in a queue before being submitted. Cash flow is delayed and filing deadline exposure increases. | Set a contractual SLA: claims submitted within 48 hours of charge entry. |
| Net collection rate below 93% for any payer | Revenue is leaking with that payer through denials, underpayments, or both. | Request a payer-specific deep dive. Identify whether the gap is denial-driven or underpayment-driven. |
| Timely filing write-offs above 0.5% | Claims are expiring before being submitted or before denials are appealed. | This is the most basic billing function. Consistent timely filing losses are grounds for termination. |
| E/M distribution skewed more than 15 points from benchmark | Providers are undercoding, and the vendor isn't providing coding feedback. | Request a coding analysis and provider-specific feedback on documentation and code selection. |
Finding problems is the point. Every billing operation has room to improve. What matters is how your vendor responds when you present the data.
A vendor who takes the findings, produces a corrective action plan within two weeks, and shows measurable improvement within 60 days is doing their job. Identifying the problem is your role as the administrator. Fixing it is theirs.
A vendor who disputes your data, deflects blame to payers or providers, or promises improvement without a specific plan is telling you something important about how they'll perform going forward. The pattern of response predicts future results more reliably than any single quarter's numbers.
If two consecutive quarterly audits show the same problems with no measurable improvement, the vendor relationship has run its course. At that point, refer to our evaluation framework and begin the transition process.
The audit I wish I had done sooner. When I finally reconciled our own practice's claims against clinical records, I found $874,000 in receivables that had gone unworked across two billing systems for six years. I identified 10.5% of billed charges aging past filing deadlines, worth $382,000 annually. That reconciliation covered 47,000 claim lines. The billing vendors who managed our revenue during those years never surfaced any of it. The quarterly audit framework in this post exists because that experience taught me what happens when you trust a vendor's report without verifying it yourself.
We'll pull the six reports from your PM system and run the analysis for you. Free, confidential, no obligation.
If your vendor is performing well, we'll confirm it. If there's revenue hiding, we'll show you exactly where.
Request a Free Vendor Audit