A patient communication pilot is a time-boxed test of whether adding plain-language explanations to your report delivery changes anything measurable. Ninety days is usually the right length: long enough to gather a defensible baseline and see behaviour change, short enough that nobody has to defend an open-ended commitment. The most common failure is not technical. It is starting without a baseline, which makes the result unarguable in either direction.
Pilots exist to reduce risk, but a badly designed one adds it. It consumes staff attention, creates a change nobody asked for, and ends with a meeting where everyone has an opinion and nobody has data. What follows is a structure that avoids that, written to be used whether or not the vendor is us.
Before day one: define what would make you say no
Write down, in advance, the result that would cause you to stop. Pilots without a defined failure condition almost never fail, because success gets redefined at the end to match whatever happened. Agreeing the threshold beforehand is the single highest-value thirty minutes in the whole exercise.
Decide too who owns the pilot. Not the executive sponsor, the person who will actually check the numbers weekly. Pilots die of ambiguity more often than of poor results.
Days 1 to 30: measure, change nothing
The first month is baseline only. Deploy nothing. Tag report comprehension contacts separately from scheduling and billing, record handling time, and note how often a call escalates to a technologist or radiologist. Most centres have never held this number, and it is frequently two to four times what people estimate.
Resist the temptation to skip this month. Without it you cannot distinguish an improvement from a quiet week, and you will spend the final review arguing about seasonality instead of results.
Days 31 to 60: deploy narrow
Pick one modality or one location. Narrow scope produces a cleaner signal and a smaller blast radius if something needs correcting. Ultrasound and routine blood panels are usually good starting points: high volume, high query rate, lower average anxiety than cross-sectional imaging.
Brief the front desk properly. They are the people who will notice the change first, and if they do not understand what patients are now receiving, they will contradict it. Give them the actual patient-facing output to read before it goes live.
Days 61 to 90: compare, and ask the referrers
Run the same measurements as month one against the piloted segment. Then do the thing most pilots skip: ask three referring clinicians whether patients from that segment are arriving differently. It is qualitative and it is the most persuasive evidence you will collect, because referrer perception is what actually drives volume.
| Measure | How to capture it | Why it matters |
|---|---|---|
| Report comprehension contacts per 100 reports | Front-desk tag | The primary operational number, and the easiest to defend |
| Escalation rate to clinical staff | Call log | Shows whether expensive time is being protected |
| Average handling time | Phone system or manual sample | Volume can hold steady while calls get shorter |
| Patient feedback mentioning clarity | Existing feedback channel | Slow-moving but the one leadership remembers |
| Referrer perception | Three short conversations | Qualitative, and the most commercially relevant |
| Staff sentiment | Ask the front desk directly | If they hate it, it will not survive whatever the data says |
Four ways pilots fail
- No baseline. The most common failure by a wide margin. Everything afterwards becomes anecdote.
- Scope creep. Adding a second location in week six because it is going well destroys comparability.
- No named owner. Somebody must look at the numbers weekly. Without that, month three becomes a scramble.
- Excluding the front desk. They experience the change most directly and are usually consulted last.
What a pilot cannot tell you
Ninety days will show you operational effects: call volume, handling time, staff and referrer perception. It will not show you retention, referral growth or reputation change, which move over quarters, not weeks. Be honest about that boundary when presenting results, because overclaiming in the first review is how good pilots lose credibility in the second.
