Establish what you already have

Most organizations do not have a POCUS program. They have an installed base.

Ultrasound adoption is organic before it is organized. Residency-trained physicians arrive with the skill and expect the tool. Departments buy handheld units out of operating budgets because the capital process is slow. Individual clinicians purchase personally owned devices. Nursing units run bladder scanners that nobody classifies as imaging. By the time an enterprise program is chartered, there is almost always a decade of accumulated practice that the charter does not account for.

Starting from a greenfield assumption is the fastest way to lose credibility with the clinicians whose participation the program requires. Phase 0 is inventory and honesty, and it usually takes six to ten weeks.

The decisions

Device census, including what you are not supposed to have. Count every ultrasound device capable of producing a stored image, regardless of how it was purchased or who owns it. Biomedical engineering holds part of this list. Departmental managers hold another part. A meaningful fraction exists only in individual clinicians' possession. Assess age, condition, network capability, and DICOM export capability, not just presence. Older devices frequently lack functioning network capability or the ability to export DICOM files at all. Newer devices sometimes depend on third-party software to reach long-term storage. Both facts change the cost of everything downstream. Plan against an equipment life cycle of roughly seven to nine years depending on utilization, which is longer than the five-year assumption embedded in the return calculation most frequently quoted in this category.

Volume, three ways. Scans performed, scans documented, scans billed. These three numbers are never equal and the gaps between them are the program's opening business case. Published utilization patterns give a rough calibration for emergency departments, where POCUS touches somewhere between two and ten percent of patients depending on program maturity. Residency programs run at the top of that range, established non-residency programs in the middle, and new programs at the bottom. If your documented volume implies a rate far below what your clinicians describe, the gap is not clinical behavior. It is capture.

Where images go now. For each device, trace the actual path an image takes after the study ends. In most organizations the honest answer for a large share of devices is that the image stays on the device until it is overwritten. That answer is a finding, not an embarrassment, and it needs to be written down.

Existing privileging language. Some departments have it. Most have language that covers the physicians who arrived with residency training and covers nobody else. Collect what exists before drafting anything new, because medical staff offices react badly to parallel processes.

Utilization management exposure. For any ambulatory or outpatient site in scope, document current prior authorization requirements and site-of-service policy for the ultrasound codes you expect to use. The inpatient business case does not transfer to the clinic, and finding that out in Phase 5 wastes a year.

The amnesty question

The installed base has to be resolved explicitly, and there are only three defensible answers. Absorb the existing devices into the program with a compliance deadline. Sunset them on a defined schedule tied to equipment life. Or grandfather them out of the program entirely with a written acknowledgment that studies performed on them are not part of the medical record and not billable.

Choosing none of the three is the common path, and it produces the worst outcome, which is a program that formally requires archived images while a third of actual scanning volume happens on devices that cannot archive. The policy is then honored selectively, which teaches everyone that program policy is optional.

What to require from your vendor

At this stage the useful vendor conversation is diagnostic rather than commercial. Ask any incumbent supplier for a complete inventory of what they have sold your organization, by department, with install dates, current software versions, network capability, and DICOM export capability. Ask what a compatibility assessment against your enterprise archive would cost and how long it would take. Ask which of the devices in your building they will still support in three years.

A supplier who cannot produce an accurate inventory of their own installed base at your organization is telling you something useful about what the implementation relationship will look like.

Failure mode

Program design proceeds on greenfield assumptions. Six months in, three departments that bought their own devices decline to participate, a nursing unit reveals it has been running a substantial unmanaged scanning operation, and the program's first published policy has to be rewritten to accommodate reality. The rewrite costs less than the credibility.

Gate criteria

This workstream is done when you have a named executive sponsor with budget authority, a written current-state inventory covering devices, volume, image destination, and existing privileging language, and a documented decision on amnesty terms for the installed base. It runs alongside governance rather than strictly before it.