Everything at once, with elapsed time - which is what turns individual calls into a picture of the ward.

A corridor display shows one call, or cycles through several, in characters large enough to read at distance. It cannot show how long each has been waiting, how many there are, or which are already being attended, because there is no room for that information at that character size.

A panel is read standing in front of it, so it can show a list or a plan: every active call, its type, its timer, and its status.

That supports the triage decision. Three calls waiting, one of them a staff-assist from a bay where a colleague is already working and one routine call that has been waiting eleven minutes, is a different situation from three calls raised in the last thirty seconds - and only the panel makes that visible.

It is also usually where calls are acknowledged, which is how the rest of the system learns that someone is responding.

Lamp matrices for absolute simplicity and reliability; graphical panels for information and integration - and many buildings run both.

A lamp matrix is one indicator per room on a fixed layout. It is instantly readable with no interpretation, works regardless of software, and fails in an obvious way. What it cannot do is show elapsed time, call history, or anything about status beyond on and off - and it is fixed, so a change in room numbering means new legends.

A graphical panel shows call type, timers, status and history, can be reconfigured as the ward changes, and integrates with reporting and mobile alerting. It depends on a display and software, which is a dependency a lamp matrix does not have.

In practice the graphical panel is now standard for the working interface, and where a simple always-available indication is also wanted, lamps or a basic repeater provide it.

Where staff actually are during every shift - which is not always where the drawing puts it.

The obvious location is the nurses' station, and that is usually right for the day. The question that decides the design is where staff are at night and at quiet periods, because that is when a single unanswered call matters most and when the station may be unoccupied.

So panels are frequently needed in more than one place: the main station, a night station or treatment room, and sometimes a staff rest area - with the understanding that a panel in a room where staff cannot see it is not covering that period.

Sight lines and glare matter. A panel facing a window is unreadable at certain times of day.

Walk the ward with nursing staff at design stage and ask where they are at three in the morning. The answer frequently differs from the assumption.

A member of staff acknowledges the call at the panel or at the point of call, which tells the system - and everyone else - that it is being handled.

Without acknowledgement, every display continues to show the call and several staff may respond to the same room while another call waits.

Acknowledging typically changes the indication rather than clearing it: the call remains visible as attended, and the timer may continue, so it is clear that the call has been accepted but not yet resolved.

The important design rule is that a call must be cleared at the point of call - in the room - not remotely. That ensures someone has actually attended. A system that allows calls to be cancelled from the panel can be cleared without anyone going, which defeats its purpose and is generally prohibited by the applicable healthcare guidance.

Confirm the reset behaviour against the standard for the market, since it is one of the requirements most often specified explicitly.

Call and response times per event, retained long enough to show trends, and exportable for reporting.

At minimum the panel or its system should record when each call was raised, its type and location, when it was acknowledged, when it was cleared, and by whom where staff identification is used.

From that come the response-time figures that are increasingly required by inspectors and used internally to support staffing cases.

Retention and access need defining, since these records concern identifiable patients and staff and fall under data protection obligations.

The practical specification point is to state what reports are needed and see them demonstrated before purchase. Systems vary widely in whether the data is genuinely usable or merely present, and a log that can only be read event by event on the panel does not support the reporting a ward manager actually needs.

Yes, and larger panels commonly do - but the display has to make clear which zone a call belongs to, and coverage responsibility has to match.

A panel covering several bays or a whole floor is efficient and gives a wider overview, which helps at night when one team covers more area.

The risk is dilution. A panel showing calls from areas the staff at that position are not responsible for produces alerts they must mentally filter, which is alarm fatigue in another form. And a call in a zone that everyone assumes someone else is covering is exactly the failure mode to design out.

So multi-zone panels should show zone clearly, and the escalation configuration should define who is responsible for each zone at each time of day.

Where staffing changes between day and night, the system should support different configurations for each rather than being set once for the busiest case.

Regular functional testing of every call point through to the panel, with results recorded - because a call that never reaches the panel is invisible.

The test that matters is end to end: raise a call of each type from each location and confirm it appears correctly at every display that should show it, with the right type and zone, and that it clears only from the point of call.

That is more work than testing the panel alone, and it is the only test that finds a call point wired to the wrong zone or a room whose labelling is wrong - which is a fault that produces staff attending the wrong room in an emergency.

On a graphical panel, also confirm the display itself is not showing stale information after a software or power event.

Keep the test records. On a system with life-safety relevance they form part of the compliance evidence, and they are asked for at inspection.