How to Benchmark Your POCUS Charge Capture

The denominator is missing. Billing systems report on claims submitted and have no visibility into scans that never became claims, which in POCUS is the majority of them. A practice with a healthy denial rate can believe it runs a healthy programme while losing most of its ultrasound revenue upstream of the billing office.

The method below takes about a day.

The four ratios

1. Documentation rate = scans with a filed interpretation ÷ scans performed
2. Archive rate = scans archived ÷ scans documented
3. Billing capture rate = scans billed ÷ scans archived
4. Collect-on-billed rate = scans generating payment ÷ scans billed

Define them conditionally, as above, and they multiply. A programme running at 85% on each of the first three converts 61% of performed scans into claims, before any denial. Multiply by the fourth and you have the proportion of clinical work that produces revenue.

Alerhand’s department, the only one to publish all four stages, converted 42% of performed scans into claims and collected on 55% of those, so 23% of performed scans generated payment.

For what these ratios look like in published departments, see phantom scans. In summary, 42% billing capture is a realistic starting point and 75% a realistic target, with collection on billed studies a separate question your own remittance will answer.

The five-step audit

Step 1: get the denominator

Pull exam counts from the ultrasound devices themselves rather than from the billing system. Handheld and cart-based systems both log studies. Cover a full month and include every device, including personally owned handhelds if clinicians use them.

If your devices do not expose a count, a two-week tick-sheet tally by the scanning clinicians will get you close enough. Do not ask them to estimate retrospectively, because retrospective estimates run low.

Step 2: get the archive count

Query your PACS, middleware or archive for POCUS studies over the same period. If you cannot query by study type, filter by performing department and modality.

If the answer is that you have no archive, that is your finding, and it is a common one. Two published baselines report archive rates at or near zero before intervention.

Step 3: get the billed count

Pull every POCUS CPT code submitted for the same period. The common ones are 76604, 76705, 76775, 76815, 76817, 76856, 76857, 76881, 76882, 76937, 76942, 93308 and 93971. One trap: if radiology bills some of the same codes, filter by rendering provider or department, or their volume will inflate your figure.

Step 4: get the collected amount

Take this from remittance rather than from charges. Charges tell you what you asked for. Divide collected dollars by billed studies to get your blended collected amount per POCUS study. Expect somewhere between $25 and $70 depending on setting and payer mix. Hospital-based professional-only billing sits at the low end, since the professional components of the common limited codes run roughly $23 to $32 under Medicare. Office-based global billing sits considerably higher.

Commercial payers typically pay above Medicare, though Medicaid managed care and some commercial products pay at or below it. RAND’s hospital price transparency work puts professional services at 184% of Medicare and all services at 254%, based on 2020 to 2022 claims. Those figures describe prices paid to hospitals and hospital-based professionals, so treat them as an upper reference point rather than a direct multiplier for an office-based practice.

Step 5: calculate leakage

(Scans performed − scans billed) × net collected per billed study = annual leakage, once you annualise the month.

A department performing 4,000 POCUS studies a year at a 42% billing capture rate and $45 net collected per billed study is leaving roughly $104,000 unbilled annually. Moving capture to 75% recovers about $59,000 of it. The clinical work is identical in both scenarios.

Run steps 1 and 3 first

If you do nothing else, get the device count and the billed count and put them side by side. The gap between them is almost always larger than your entire denial rate, and it appears in no report your billing system produces. Denial reporting can only describe claims that were made, so it is structurally blind to scans that never entered the process.

Two things that will distort your numbers

Educational scans. Some proportion of POCUS is performed for teaching or practice, on patients who did not need it or who had already been imaged. These should not count in your denominator, and if you cannot distinguish them your collection rate will look worse than it is. Capturing the clinical versus educational designation at the point of scan is worth building in for this reason alone, and it matters for credentialing too.

Multi-code studies. A single encounter can legitimately generate more than one code, for example a lung scan and a limited echo. Decide up front whether your denominator counts scans or codes, and stay consistent.

What to do with the number

Take it to whoever signs off capital and software spend, alongside your device count. The argument that lands with a finance director is that a fixed asset you have already paid for is returning a fraction of what it could, for reasons nobody in the department chose. Framing it as a documentation discipline problem invites a memo and changes very little.