Documentation, coding, and capture
Test claims have been adjudicated and the denial rate is measured on real volume.
A program that produces clinical value and no financial evidence loses its budget line to something that reports a number.
This phase is where most of the honest work of program sustainability happens, and it is the phase most often delegated to someone who joins the effort late.
Reimbursement rules are workflow requirements
Most of what looks like billing detail is a workflow specification written in another language. A payer does not pay for a scan. It pays for a documented, medically necessary study with a stored image, an interpretation, and a signature, performed and coded correctly for the setting. Every one of those conditions is a requirement on a system built in an earlier phase, which is why reimbursement belongs in the design conversation rather than at the end of it.
| Reimbursement rule | The workflow requirement it creates | Built in |
|---|---|---|
| A claim requires a permanently stored, retrievable image | Archive and image-retention workflow, not an optional nicety | Phase 3 |
| A claim requires a documented interpretation and a signature | A structured report template carrying the required elements | Phase 5 |
| Focused POCUS is usually coded as a limited or focused exam, distinct from a complete study | Defined acquisition protocols per application, so the views scanned match the code billed | Phase 2 and Phase 4 |
| Professional and technical components split by site of care | A financial model that separates the two, and clarity on who bills what in facility versus office | Phase 5 |
| Payment is conditioned on medical necessity and diagnosis linkage | Documentation that captures the indication and the ICD-10 code, drawn from the pathway map | Phase 2 and Phase 5 |
| Bundling and same-day edits limit separate payment | Ordering and scheduling rules that head off predictable denials | Phase 5 |
| Payment is conditioned on the performing operator's credentials | The three-gate screen resolved before an expanded operator scans | Phase 4 |
| Outpatient studies frequently require prior authorization | An authorization workflow owned by patient access before the scan | Phase 5 |
Read the table one way and it is a coding primer. Read it the other way and it is the requirements list for Phases 2 through 5. They are the same list, which is the point.
The code-level specifics sit a level below this and they change every year, so treat them as a lookup rather than as something to memorize. Four decide most cases. The line between a limited and a complete study, the professional and technical component split with its modifiers, the diagnosis codes that establish medical necessity, and the bundling edits that govern same-day studies most often determine whether a defensible study is also a paid one. The Society of Point of Care Ultrasound maintains a practical reimbursement resource for program leaders, and the AMA CPT code set is the authority for the codes themselves. Confirm the current year's detail there rather than carrying last year's assumptions into a new payer contract.
Documentation as the capture instrument
The Billing I-AIM framework, developed through an institutional collaborative in emergency medicine, is the most useful available structure for this. It adapts the I-AIM teaching model of indication, acquisition, interpretation, and medical decision-making by replacing the final step with money, producing a four-part sequence of indication, acquisition, interpretation, and money that maps the specific checkpoints determining whether a study is billable. It was built to work across variable POCUS environments rather than to codify one department's practice.
The framework matters because POCUS billing differs structurally from comprehensive radiology ultrasound. In radiology, a sonographer acquires and a radiologist interprets, and the two roles generate separate documentation. In POCUS, the same clinician acquires and interprets at the bedside, which collapses the documentation into a single act performed under time pressure by someone whose training emphasized the clinical question rather than the billing requirement. Denial rates reported by providers reflect that mismatch.
The documentation elements are consistent across payers even where the thresholds are not. Indication supporting medical necessity. Description of structures examined. Findings. Interpretation. Provider signature. Reference to permanently stored, retrievable images. Any of these missing renders the study clinically useful and financially invisible.
Build these into structured reporting templates in the EHR rather than relying on free text, and design the template so that completing it correctly is faster than completing it incorrectly. That is a workflow design problem, not a compliance communication problem, and treating it as the latter is why compliance campaigns produce short-lived improvement.
Professional and facility components
POCUS generates both a professional fee for performance and interpretation and a facility fee for space and equipment, and in hospital settings the facility component is typically the larger of the two by a wide margin. In the most frequently cited emergency department model, facility charges represented roughly three quarters of the total capture opportunity.
This has an organizational consequence that programs consistently miss. The clinical department bears the workflow burden and the documentation burden while the majority of the financial benefit accrues to the facility. If departmental budgets are managed independently, the department doing the work sees cost and the enterprise sees revenue. Programs that do not address this misalignment explicitly should expect exactly the participation their incentive structure produces.
The financial model, built honestly
A figure circulates widely in this category, describing a $280,000 investment in four emergency department ultrasound units generating four to five million dollars over five years. It is worth understanding what that number is before repeating it.
It originates in one health system's internal program guideline as an estimate of what would have been captured had every POCUS study performed in a single emergency department during one year been billed. The majority of that estimate sits in facility charges. It is a charge-capture estimate rather than a realized-collections figure. It is extrapolated across an assumed five-year equipment life, which is shorter than the seven to nine year life cycle described in the current informatics literature. And it is net of nothing, including the director FTE, the workflow manager licensing, the storage infrastructure, the integration labor, the training time, and the quality assurance effort that make capture possible at all.
The figure is not dishonest. It answers a specific question, which is how much billing the organization was leaving on the table. It simply does not answer the question a finance committee will ask.
Build the model with the following architecture instead.
Revenue side. Volume by application and by setting. Charge capture rate, which is documented and billed studies divided by studies performed. Realized collection rate net of denials, downcoding, and contractual adjustment. Professional and facility components separated. Payer mix, with the outpatient book modeled separately because prior authorization and site-of-service policy change the economics materially.
Cost side. Devices, transducers, and disposables. Workflow manager licensing and the integration labor to implement it. Storage and infrastructure. Network and security upgrades. Director FTE with support staff and space. Training development, delivery, and clinician time. Quality assurance effort in reviewer hours. Ongoing maintenance and the replacement reserve against a seven to nine year life cycle. Support and help desk load.
Avoided cost, stated separately and labeled as modeled. POCUS-first diagnostic strategies have been modeled to produce meaningful national savings in specific conditions, with published simulation estimates ranging from roughly $16 million annually for nephrolithiasis to $30 million for small bowel obstruction and $95 million for acute diverticulitis. These are real and they are simulations. Present them as directional rather than as line items, and never blend modeled avoided cost into a revenue projection.
The anchor, with its limitation stated. The strongest available real-world evidence of financial effect at enterprise scale is URMC's reported 26 percent growth in hospital charges across a deployment of nearly 1,200 probes in more than seventy departments. It is measured, published, attributed, and accompanied by an honest account of what remained unfinished, which makes it a better citation than any modeled projection.
It is also charges, and the authors say so directly. They report that return on investment has been difficult to assess, that demonstrating financial viability is particularly hard on the inpatient side where they operate within a diagnosis-related group framework, and that isolating POCUS-attributable revenue in a complex medical ecosystem proved difficult enough that they focused on charges rather than more precise revenue tracking.
That admission is not a weakness in the citation. It is the most important finding in it. A leading academic medical center, three years into a well-resourced enterprise deployment, could not cleanly isolate the revenue effect. Any business case that promises your organization will do better should explain how, and the answer usually involves the DRG problem, which is that inpatient limited exams frequently do not generate separate global charges and are absorbed into the episode payment. Model the inpatient contribution conservatively and put the weight of the case on the outpatient book, on avoided cost, and on the capacity and throughput effects the program was actually built to produce.
The outpatient reality
An inpatient business case does not transfer to ambulatory deployment without rework, and the difference is prior authorization.
Access expansion programs are built to reach patients in clinics, dialysis centers, post-acute facilities, and homes. Those are precisely the settings where utilization management applies most aggressively and where site-of-service policy steers imaging toward designated facilities. A program designed around expanding access will encounter authorization requirements that the acute care program never sees, and the incremental administrative cost per study can consume the margin the model assumed.
Model the outpatient book separately. Document current authorization requirements for every code in scope, by payer, before deployment rather than after. Where the program depends on expanded operators, remember that payer requirements condition payment on the credentials of the performing professional, which links this directly back to gate three of the Phase 4 screen.
Patient access owns this work and should be assigned it explicitly. Eligibility verification, authorization submission, tracking, and appeal are operational processes with staffing implications, and an authorization requirement discovered after go-live typically lands on a scheduler who was never told the service existed. Where the pathway map records what preceded the POCUS study, that record is frequently the clinical criteria the authorization requires, which is one of the practical reasons Phase 2 pays for itself here.
In the practice setting
The practice version of this phase is simpler in structure and harder in execution, because the capability being described above as a department is here a vendor relationship or a single person.
Test the revenue cycle before deploying, not after. Run a small number of real studies, document them completely, submit real claims, and watch what happens. Do this during Phase 3 rather than treating it as a Phase 5 activity. The first claim is where you discover that your billing service has never submitted this code for this practice, that the payer requires a modifier nobody mentioned, or that the ordering and performing provider fields resolve differently than assumed. Discovering that on twelve studies is a fixable problem. Discovering it on twelve hundred is a write-off.
Confirm coverage in an office place of service. Coverage for an application in a hospital setting does not imply coverage in an office. Check place-of-service policy for every code and payer in the practice's mix, and treat any payer without a clear answer as an unknown to resolve rather than an assumption to carry.
Assign denial work to a named person. Not to a vendor description of denial management, and not to the practice manager as an implied duty. Denials on a new service line arrive in a cluster in the first quarter and they are informative. Nobody working them means the information is lost and the revenue with it.
Decide who owns the model. In a practice there is often no finance function separate from the owner or administrator. That is workable. What is not workable is a business case built from a vendor's projection with no internal owner, because there is then nobody positioned to notice when the projection and the deposits diverge.
What to require from your vendor
Documentation templates that map to actual payer requirements rather than to the device's report format. Charge capture reporting you can audit independently, with access to the underlying study data rather than to a summary dashboard.
Direct clarity on which claimed financial benefit is charge capture and which is realized collection. A vendor unable or unwilling to make that distinction in a business case is telling you the number is charge capture.
Where the vendor offers reimbursement support, ask what that support consists of, whether it is included, and whether it extends past go-live. Ask specifically what health economics evidence exists for the deployment model you are pursuing, and treat the absence of a cost-effectiveness or budget impact analysis as a finding rather than as a gap someone else will fill.
Failure mode
The program runs well clinically for two years and produces no defensible financial record. At the next capital cycle it is compared against a service line that reports revenue, and it loses. Nobody involved believes the program failed, and the budget outcome is the same as if it had.
Gate criteria
This workstream is done when test claims have been submitted and adjudicated, denial and downcode rates are measured on real volume rather than estimated, the financial model is owned by finance and built on collections rather than charges, and outpatient authorization requirements are documented for every code and payer in scope.