At a neighborhood heat-response pilot, an outreach worker identifies a visitor who wants transport to a cooling center. The written plan treats the referral as complete when a colored card is issued. At the transport desk, however, nobody knows that the visitor is coming, what support was requested or who should confirm the pickup. Every person can truthfully say that they finished a task, yet the service still breaks between those tasks.
This is a planning problem before it is a performance verdict. A pilot can have a sound aim, committed partners and completed activities while losing the person, information or decision at a transfer point. Walking one handoff from sender to receiver gives the team evidence about the connection it actually built, not the connection described in a flowchart.
The gap can sit between two completed tasks
The World Health Organization's practice brief on continuity and coordination of care treats coordination as work across people, providers, settings and time. A local public-health pilot may involve different institutions, but the same planning question remains useful: what must continue when responsibility changes hands?
In the heat-response example, the outreach worker's task is not merely to hand over a card. The transport coordinator needs a clear signal, the authority to act on it and a way to confirm that the request was accepted. The visitor needs to know where to wait and what will happen next. Without those connections, counting issued cards measures activity at the first desk rather than completion of the intended route.
The Agency for Healthcare Research and Quality describes a clinical handoff as a transfer of information together with authority and responsibility. A community pilot is not a hospital shift change, so the clinical tool should not be copied as a script. Its distinction is still valuable: information sent is not the same as responsibility accepted.
That distinction also prevents a familiar argument. One partner may blame low uptake, another may blame transport capacity and a third may say the outreach message was unclear. Before the team debates those explanations, it can observe whether a real request crossed the boundary at all. If partners later interpret that observation differently, the site's method for making competing readings testable can be used without erasing the handoff evidence.
Follow one request until the loop closes
Choose one ordinary case that can be observed without collecting unnecessary personal details. Tell the staff that the exercise tests the route, not the individual. Start when the sender recognizes a need and end only when the receiver accepts the next action or clearly declines it. The useful unit is one complete transfer, not a morning's total volume.
Minute 0 â the sender names the next action. The outreach worker says what the visitor is being referred for, which desk should respond and what would count as acceptance. The planner records the observable signal: a card is passed, a call is made or a coordinator is approached. Private health details are not needed to test whether the route exists.
Minute 3 â the receiver shows what arrived. The transport coordinator states what was understood, who now owns the request and which condition would prevent action. A missing destination, unclear pickup point or unassigned decision is recorded exactly where it appears. The planner does not repair the process during this first pass, because an improvised rescue would hide the original gap.
Minute 8 â the visitor can describe the next step. Ask only whether the person knows where to wait, who will respond and what to do if no response arrives. This is not a satisfaction survey. It checks whether the handoff produced a usable next action for the person the pilot is meant to serve.
Minute 15 â the loop has one visible ending. The receiver confirms acceptance, declines with a stated reason or redirects the request to a named alternative. Silence is not an outcome category. If the team cannot tell which ending occurred, the first redesign should clarify confirmation before adding more outreach, staff training or promotional material.
The CDC's Program Evaluation Framework emphasizes context, collaborative engagement and credible evidence throughout evaluation. A handoff walk-through turns those principles into a bounded planning record: setting, sender, signal, receiver, confirmation and observed delay. It does not prove that the entire program works. It shows whether one necessary connection is functioning under ordinary conditions.
The practical decision should be equally bounded. Keep the handoff if the signal is understood, responsibility is accepted and the visitor can act on the confirmation. Redesign the transfer if one of those elements disappears. Then walk the same route again. By locating the break between tasks, the team can improve the connection without confusing more activity with better continuity.