Closing sections

Everything After the Device

A Provider's Playbook for Point-of-Care Ultrasound Programs

Consolidated draft, August 2026 OneAnother Health, LLC

---

Why this playbook exists

Point-of-care ultrasound programs rarely fail for clinical reasons. The evidence base is settled enough that the clinical question is no longer interesting. Adding POCUS to the standard diagnostic pathway for acute dyspnea increases the rate of correct diagnoses without increasing adverse events, a finding strong enough that the American College of Physicians built a guideline around it. Systematic review of internal medicine inpatients shows POCUS changing or adding to the primary diagnosis in up to a quarter of cases and altering the management plan in roughly half. Emergency medicine has required ultrasound training for more than a decade. Critical care, anesthesia, obstetrics, and increasingly family and internal medicine have followed.

Programs fail somewhere else. They fail because images do not reach an archive, because documentation does not support a claim, because privileging was never completed for the people doing most of the scanning, because the quality process was designed for fifty studies a month and applied to five thousand, or because nobody could produce a number when the budget cycle came around. Every one of those failures is a delivery problem, and every one of them lands on the buyer rather than the vendor.

They also fail for a reason that gets almost no attention, which is that the program was specified against a device category rather than against a clinical decision. A POCUS study exists to change what happens next. If nobody wrote down what precedes it, what decision it informs, and what the finding is supposed to trigger, then the report will describe findings, the clinician will act on instinct, and the program will have no way to demonstrate that any of it mattered.

That asymmetry is the reason this document exists. Across the imaging value journey, from population risk identification through acquisition, reporting, clinical decision-making, and population health management, there are roughly fifteen foundational requirements that have to be met before imaging produces value at scale. Vendors fully own three of them. They partially address four more. The balance is unbuilt, unowned, or quietly assumed to be the customer's problem. The commercial motion that dominates this category concentrates on the acquisition step, which is the step where devices live, and treats everything on either side as implementation detail.

The result is a predictable pattern. A system evaluates devices, compares transducer configurations and AI features, negotiates a capital purchase, and then discovers that it has bought the smallest component of the thing it actually needed. What follows is eighteen to thirty-six months of internal work that nobody scoped, nobody budgeted, and nobody warned them about. Some systems do that work well. The University of Rochester Medical Center is three years into a four-year enterprise deployment, has 1,199 of a planned 2,000 probes live across more than seventy departments and divisions, and reports a 26 percent growth in hospital charges as a result. They also report, without hedging, that they have not yet been able to fully implement their image storage policy and that documentation compliance among established providers remains a challenge. That is what a well-run, well-resourced, publicly documented deployment looks like from the inside.

This playbook is written for the person accountable for that work, in a health system or in a practice. Those are different jobs and the playbook treats them as such, because the published guidance in this category almost uniformly assumes an enterprise archive, an informatics function, and a medical staff office, none of which a primary care group or specialist practice has. For those organizations the problem is not integration. It is assembly, and the assembly decision determines how long it takes to reach a billable study.

It is company-agnostic by design. No devices are named, no platforms are recommended, and no vendor case studies appear. The technology categories matter and the specific products do not, because the products change faster than the decisions do.

It will also be read by people who build and sell these systems. That readership is intended. Each phase carries a section on what to require from a vendor. For a health system, those sections are a procurement instrument. For a manufacturer, they are an itemized list of everything the current commercial model leaves on the customer's side of the line. No argument is necessary. The list makes the case.

---

Two starting points

Before the sequence, a fork. This playbook serves two organizations with very different starting conditions, and conflating them is why most published guidance is unusable for half its readers.

The health system. An enterprise archive already exists, whether PACS or VNA. There is an imaging informatics function, an enterprise IT organization, a medical staff office, a revenue cycle department, and a compliance apparatus. The program's central problem is integration. Existing infrastructure has to be extended to a new imaging modality performed by clinicians who are not radiologists, in workflows that were never designed for it. The work is largely internal, the dependencies are largely internal, and the timeline is set by internal approval cycles.

The practice. A primary care group, a specialist practice, a multi-site clinic organization, a dialysis or post-acute operator. There is no PACS. There is no VNA. There is no imaging informatics function and often no dedicated IT staff. Billing runs through a practice management system or an outsourced revenue cycle vendor. Credentialing runs through payer enrollment rather than a medical staff office. The program's central problem is not integration. It is assembly, and the assembly decision is a vendor selection decision.

The two tracks share the same logic and diverge sharply in execution. Where a phase differs materially by setting, this playbook flags it under a heading marked In the practice setting. Read your track and skim the other one, because the practice-setting sections often clarify what a health system is actually buying when it buys infrastructure it already has.

One point applies to both. Nothing in this playbook treats artificial intelligence as a category to adopt. AI-guided acquisition and AI post-processing are capabilities that answer specific constraints. If the constraint is not present, the capability is a cost. The pathway work in Phase 2 is what tells you which constraints you actually have.

---

How to use this playbook

The structure is a sequence, not a table of contents. Eight phases, each ending in a gate. The gates are the point. A phase-and-gate structure is only useful if the gates are honored, and the single most common cause of expensive POCUS program failure is running Phase 3 work after Phase 6 has already started by accident.

Phases are not parallel workstreams. Some activity within them overlaps, and the playbook flags where. But the ordering constraint is real, and it inverts the sequence most organizations follow. Clinical need comes first. Requirements derive from clinical need. Devices come late.

Each phase covers five things. What the decision is. Who owns it. What the evidence supports. What to require from a vendor. What failure looks like when the phase is skipped. Then the gate criteria that let you move forward.

Read the whole sequence before starting any of it. The gates in Phase 5 and Phase 6 constrain choices you make in Phase 3, and discovering that late is expensive.

What the industry owes and does not deliver

This section is written for the people who build and sell these systems. Health systems should read it as a procurement standard.

Every phase in this playbook contains a section on what to require from a vendor. Consolidated, those requirements describe a delivery relationship that almost no one in this category currently offers. That is not an accusation. It is an observation about how the commercial model evolved and what it optimized for.

The structural picture

Map the imaging value journey across its stages, from population risk identification through acquisition, reporting, clinical decision-making, and population health management. Then list the foundational requirements that must be met for imaging to produce value at scale. There are roughly fifteen. Vendors fully own three of them. They partially address four more. Everything else is unbuilt, unowned, or assumed to be the customer's problem.

Every column of that map except one is the health system doing the work. The vendor footprint concentrates in the middle, at acquisition and reporting, which is where devices and their immediate outputs live. The upstream question of which patients should be imaged and the downstream question of what happens to the finding are both left to the buyer.

That pattern has been stable for forty to fifty years. It is not an oversight. It is the shape of a business built around capital equipment placement, and it worked when the capital equipment was the scarce resource. In point-of-care ultrasound the capital equipment is no longer the scarce resource. Devices are portable, capable, and increasingly commoditized. The scarce resource is everything this playbook describes.

What the current model produces

Feature competition at the device level. Modality teams pursue product differentiation because that is what their P&L rewards. The consequence is that a health system buying across modalities from a single manufacturer experiences separate contracts, separate account teams, and separate roadmaps with no shared quota, pipeline, or strategy. They are buying from several companies that share a logo. What could be a platform advantage is converted into a series of isolated product transactions, each one competitively exposed at renewal.

Unstructured output. Devices produce images and reports. Health systems need coded, structured findings that can populate registries, feed risk models, support quality measures, and trigger pathways. The most valuable imaging AI produces structured inference rather than better pixels, and the architecture signals in recent research point clearly in that direction. Continuing to ship unstructured output guarantees imaging data stays outside the infrastructure that coordinates everything else in the record.

The procurement cliff. The commercial relationship intensifies through evaluation and closes at purchase order. Implementation is handed to a services organization with different incentives, and program success, which takes eighteen to thirty-six months to establish, falls outside the window anyone is measured on. The customer's hardest year begins the day the vendor's engagement ends.

A segment the model does not serve at all. The commercial motion in this category is built for buyers with imaging infrastructure. Primary care groups, specialist practices, and multi-site clinic organizations have none, and they are where a large share of the clinically appropriate volume actually sits. Serving them requires a package rather than a product, covering device, cloud archive, revenue cycle enablement, quality assurance, and competency validation, priced and contracted for an organization with no informatics staff and limited tolerance for a six-month implementation. Very few offerings in the market do this completely, and the incomplete versions concentrate on the components that resemble a product. The gap is the components that resemble a service, which is precisely where the practice cannot substitute its own capability.

What a platform participant would do differently

The alternative is not for imaging companies to claim they orchestrate care pathways. They do not, and the claim is not credible to anyone who has run a health system. Imaging contributes decisive information at specific decision nodes. The electronic health record coordinates the continuum around those nodes. Claiming otherwise invites a comparison the imaging vendor loses.

The more defensible position is platform participation, and it requires three shifts.

Shift	What it means
From device features to platform coherence	Consistent data standards and one implementation relationship across modalities. Organizational more than technical, and hard because it cuts against modality P&Ls.
From reports to structured findings	Coded output with a published specification, built to be consumed by systems the vendor does not control. Less short-term lock-in, more long-term indispensability.
From hardware placement to program enablement	Contract for the delivery requirements, own more of the fifteen than three, and be measured on program outcomes at eighteen months rather than installed units at ninety days.

The market access point

For AI-enabled capability specifically, regulatory clearance is necessary and not sufficient. A cleared device that no state permits the intended operator to use, that no code describes, and that no cost-effectiveness analysis supports is a product with a clinical story and no commercial pathway.

The three barriers are state licensure, coding, and health economics evidence, and none of them resolve on their own timeline. Companies that treat market access as a track running in parallel with product development will reach scaled deployment. Companies that treat it as a downstream problem to solve after clearance will spend years explaining strong clinical data to buyers who cannot act on it.

Health systems can accelerate this. Ask for the health economics evidence during evaluation. Ask which operator populations are within cleared intended use. Ask what code the vendor expects you to bill. The questions themselves change what gets built, because a question asked consistently across a market becomes a requirement.

Descope and exit

Every playbook in this category ends with expansion. This one ends with the harder section.

Programs should be capable of getting smaller. Building that capability in advance costs nothing and preserves the option.

Signals worth acting on

Signal	The read, and the response
Capture rate stays below 50 percent after two stable quarters, and the cause is workflow rather than training	A Phase 3 problem that Phase 5 effort will not fix. Return to the architecture.
Participation concentrated in one or two departments 18 months after launch	The operating question was written for one department and applied to all. Narrow the program formally to where it works.
Quality findings persist through two feedback cycles in an application or operator population	The competency model for that combination is wrong. Suspend that application rather than adding a third cycle.
Devices in a department that has not scanned in 90 days	Stranded capital. Redeploy, and track monthly so a quarter of the fleet is not found unused at refresh.
An expanded operator program that cannot clear all three gates	Convert to a documented cost center with an explicit clinical rationale, or discontinue.

How to descope well

Narrow by department or by application rather than by reducing standards. A smaller program running at full compliance is defensible. A large program running at partial compliance is a finding waiting to be documented by someone else.

Say what changed and why. Programs that quietly shrink teach the organization that the governance body does not mean what it says, which costs more than the descope.

Preserve the infrastructure. Archive integration, documentation templates, privileging pathways, and quality processes survive a scope reduction and are the expensive part to rebuild. Devices can be redeployed. Architecture should not be dismantled.

A note on sequence

The eight phases in this playbook describe roughly two to three years of work for an enterprise deployment, considerably less for a single department, and something closer to three to six months for a practice that selects a complete package and runs the pathway work honestly.

The sequence is the argument. Clinical need first. Requirements derived from clinical need. Infrastructure built to those requirements. Devices selected to fit. Capability added only where a documented constraint requires it.

Organizations that run it in this order spend more time before the first device arrives and less time afterward explaining why the program has no record of what it did. Organizations that start with device selection, which is the sequence the market is structured to encourage, arrive at the same requirements eventually and pay more for them.

The gray space in this category is the eighteen to thirty-six months between procurement and program, where nobody's incentives align and most of the value is either created or lost. The white space is the seven delivery requirements that nobody currently owns. Both are addressable. Neither is addressed by a better device.

Companion research

The analysis underlying this playbook is developed at greater length in four related pieces at erikabel.org.

• The Imaging Value Journey and Delivery Matrix. The fifteen foundational requirements, ownership mapped across the care continuum, presented as an interactive framework.
• The Pathway Illusion in the Imaging Industry. Why device-level feature competition fragments the enterprise relationship, and what platform participation would require structurally.
• The Imaging Gap Nobody Built. The three-tier imaging pathway, the interoperability problem, and the preventive imaging market the incumbent industry never served.
• AI-Guided POCUS Has a Market Access Problem. The three structural barriers to scaled AI-guided deployment, with a full obstacle matrix mapping each to reimbursement and compliance exposure.

Companion tools

Eleven worksheets accompany this playbook as separate downloads.

1. Current-state inventory worksheet
2. Pathway map and requirements derivation worksheet
3. Delivery requirement ownership map
4. Approval and dependency checklist by function
5. Documentation elements checklist
6. Privileging pathway templates by operator type
7. Expanded operator three-gate screen
8. Quality assurance sampling design worksheet
9. Financial model architecture
10. Vendor requirement summary
11. Practice-setting integrated package evaluation

A note on sources

Two sources shape this playbook and warrant explanation.

The financial figure discussed in Phase 5, describing a $280,000 investment generating four to five million dollars over five years, originates in a multi-state health system's internal POCUS program guideline published in 2024. It is an internal guideline document rather than peer-reviewed literature, and that provenance is precisely the point of the discussion. The same document is the source of the two to ten percent emergency department utilization band, the roughly three-quarters facility share of the charge opportunity, and the three percent with minimum five per provider quality sampling convention. All four figures circulate throughout this category with their origin unstated.

The University of Rochester Medical Center deployment is named because it is published, peer-reviewed, and attributed by its own authors, and because their willingness to report what has not worked is more useful than most of what has been published on what has.

Sources

Program implementation and governance

Bottemiller A, Ferre RM, Folio LR, et al. Considerations for Pre-deployment Planning in Point-of-Care Ultrasound Program Implementation: A HIMSS-SIIM Enterprise Imaging Community Whitepaper in Collaboration with AIUM. Journal of Imaging Informatics in Medicine. 2026. doi:10.1007/s10278-026-02021-y

Ma IWY, Francavilla ML, Nomura JT, et al. Governance Considerations for Point-of-Care Ultrasound: A HIMSS-SIIM Enterprise Imaging Community Whitepaper in Collaboration with AIUM. Journal of Imaging Informatics in Medicine. 2025;38(5):2585-2599.

Waldman D, Doughton J, Pino C. Roadmap to success: Blueprint for enterprise-wide deployment of a point-of-care ultrasound platform, inclusive of governance, policy, education, credentialing, and quality assurance (Part 2). Journal of Clinical Imaging Science. 2025;15(28). doi:10.25259/JCIS_76_2025

Shen L, Lobo VE, Cordova D, Larson DB, Kamaya A. An Institutional Approach for Developing a Point-of-Care Ultrasound Program Infrastructure. Journal of the American College of Radiology. 2024;21(8):1269-1275.

Lee JY, Conlon TW, Fraga MV, et al. Identifying Commonalities in Definition and Governance of Point-of-Care Ultrasound Within Statements From Medical Organizations in the United States. Journal of Clinical Ultrasound. 2023;51(9):1622-1630.

Clinical evidence

Díaz-Gómez JL, Mayo PH, Koenig SJ. Point-of-Care Ultrasonography. New England Journal of Medicine. 2021;385(17):1593-1602.

Qaseem A, Etxeandia-Ikobaltzeta I, Mustafa RA, et al. Appropriate Use of Point-of-Care Ultrasonography in Patients With Acute Dyspnea in Emergency Department or Inpatient Settings: A Clinical Guideline From the American College of Physicians. Annals of Internal Medicine. 2021;174(7):985-993.

Gartlehner G, Wagner G, Affengruber L, et al. Point-of-Care Ultrasonography in Patients With Acute Dyspnea: An Evidence Report for a Clinical Practice Guideline by the American College of Physicians. Annals of Internal Medicine. 2021;174(7):967-976.

Cid-Serra X, Hoang W, El-Ansary D, et al. Clinical Impact of Point-of-Care Ultrasound in Internal Medicine Inpatients: A Systematic Review. Ultrasound in Medicine and Biology. 2022;48(2):170-179.

Díaz-Gómez JL, Sharif S, Ablordeppey E, et al. Society of Critical Care Medicine Guidelines on Adult Critical Care Ultrasonography: Focused Update 2024. Critical Care Medicine. 2025;53(2):e447-e458.

Billing, coding, and economics

Hughes D, Corrado MM, Mynatt I, Prats M, Royall NA, Boulger C, Bahner DP. Billing I-AIM: a novel framework for ultrasound billing. The Ultrasound Journal. 2020;12:8. doi:10.1186/s13089-020-0157-0

Zeidan A, Liu EL. Practical Aspects of Point-of-Care Ultrasound: From Billing and Coding to Documentation and Image Archiving. Advances in Chronic Kidney Disease. 2021;28(3):270-277.

Dhamija A, Perry LA, OConnor TJ, et al. Development and Implementation of a Semi-Automated Workflow for Point-of-Care Ultrasound Billing and Documentation Within an Electronic Health Record. Journal of Digital Imaging. 2023;36(2):395-400.

Barton MF, Brower CH, Barton BL, et al. POCUS-first for Nephrolithiasis: A Monte Carlo Simulation Illustrating Cost Savings, LOS Reduction, and Preventable Radiation. American Journal of Emergency Medicine. 2023;74:41-48.

Barton MF, Barton KM, Goldsmith AJ, et al. POCUS-first in Acute Diverticulitis: Quantifying Cost Savings, Length-of-Stay Reduction, and Radiation Risk Mitigation in the ED. American Journal of Emergency Medicine. 2025;88:204-212.

Brower CH, Baugh CW, Shokoohi H, et al. Point-of-Care Ultrasound-First for the Evaluation of Small Bowel Obstruction. Academic Emergency Medicine. 2022;29(7):824-834.

Van Schaik GWW, Van Schaik KD, Murphy MC. Point-of-Care Ultrasonography in a Community Emergency Department: An Analysis of Decision Making and Cost Savings Associated With POCUS. Journal of Ultrasound in Medicine. 2019;38(8):2133-2140.

AI-guided acquisition and expanded operators

Kim J, Maranna S, Watson C, Parange N. A Scoping Review on the Integration of Artificial Intelligence in Point-of-Care Ultrasound: Current Clinical Applications. American Journal of Emergency Medicine. 2025;92:172-181.

East SA, Wang Y, Yanamala N, Maganti K, Sengupta PP. Artificial Intelligence-Enabled Point-of-Care Echocardiography: Bringing Precision Imaging to the Bedside. Current Atherosclerosis Reports. 2025;27(1):70.

Chiu IM, Lin CR, Yau FF, et al. Use of a Deep-Learning Algorithm to Guide Novices in Performing Focused Assessment With Sonography in Trauma. JAMA Network Open. 2023;6(3):e235102.

Gallant C, Bernard L, Kwok C, et al. AI-Augmented Point of Care Ultrasound in Intensive Care Unit Patients. Journal of Clinical Medicine. 2025;14(9):2899.

Kayarian F, Patel D, O'Brien JR, Schraft EK, Gottlieb M. Artificial Intelligence and Point-of-Care Ultrasound: Benefits, Limitations, and Implications for the Future. American Journal of Emergency Medicine. 2024;80:119-122.

Regulation, evaluation, and equity

Warraich HJ, Tazbaz T, Califf RM. FDA Perspective on the Regulation of Artificial Intelligence in Health Care and Biomedicine. JAMA. 2025;333(3):241-247.

Jacob C, Brasier N, Laurenzi E, et al. AI for IMPACTS Framework for Evaluating the Long-Term Real-World Impacts of AI-Powered Clinician Tools. Journal of Medical Internet Research. 2025;27:e67485.

Chin MH, Afsar-Manesh N, Bierman AS, et al. Guiding Principles to Address the Impact of Algorithm Bias on Racial and Ethnic Disparities in Health and Health Care. JAMA Network Open. 2023;6(12):e2345050.

Ferryman K, Mackintosh M, Ghassemi M. Considering Biased Data as Informative Artifacts in AI-Assisted Health Care. New England Journal of Medicine. 2023;389(9):833-838.

Barriers and implementation research

Theophanous RG, Tupetz A, Ragsdale L, et al. A Qualitative Study of Perceived Barriers and Facilitators to Point-of-Care Ultrasound Use Among Veterans Affairs Emergency Department Providers. PLoS One. 2024;19(11):e0310404.

Hofmann R, Blaskova LJ, Jones N. A Theory-Informed Approach to Identify Barriers to Utilising Point-of-Care Ultrasound in Practice. Advances in Health Sciences Education. 2025.

Arnold AC, Fleet R, Lim D. Barriers and Facilitators to Point-of-Care Ultrasound Use in Rural Australia. International Journal of Environmental Research and Public Health. 2023;20(10):5821.

---

OneAnother Health, LLC. Pittsburgh, Pennsylvania. oneanother.health