Some procedures are used every shift, such as a guest’s allergy question, a temperature log or a cash handover, and they stay sharp because the work itself rehearses them. Others are trained once, filed, and then wait: the choking guest, the fire evacuation, the chemical spill, the cold room that fails overnight, the needle-stick injury, the refund override that only a manager may approve.
For those, the question an operations leader actually faces is not whether to refresh the training. Everyone agrees it should be refreshed. The question is how often, and most organisations answer it with whatever interval a certificate happens to carry. That is one number. The rate at which the skill actually fades is a different number, and the gap between the two is where rare-procedure training quietly stops working.
The compliance interval and the retention interval are not the same number
Malaysian practice gives you one clear figure. DOSH’s Guidelines on First-Aid in the Workplace%202004.pdf) state that first-aiders “need to undergo a recognized course once every three (3) years,” and that the employer should keep records of when each first-aider was trained and when they were refreshed. That is a reasonable administrative cycle, and usually the only interval anyone hands you. The UK Health and Safety Executive is less specific still: employers “must train everyone in emergency procedures,” and people respond more reliably where plans are “clearly agreed, recorded and rehearsed” and staff “take part in regular and realistic practice.” Regular is not a number.
The retention research gives a much shorter figure. An editorial in Arquivos Brasileiros de Cardiologia, citing the European Resuscitation Council guidelines, notes that basic life support skills decay within three to twelve months of initial training. In the study it discusses, medical students’ correctly completed steps fell from 12.5 to 10.8 within six months. Its recommendation is low-dose, high-frequency practice distributed across the months after a course rather than one course and a long wait.
Both numbers are correct; they answer different questions. Three years keeps the record valid. Three to twelve months is when the person stops being able to do it. Planning only around the first means the training stays compliant through most of the period in which it no longer works.
What it looks like when the rare procedure is the one nobody rehearsed
In May 2017 a dust explosion and fire at the Didion Milling corn mill in Cambria, Wisconsin killed five employees and seriously injured fourteen. Among the causes identified in the US Chemical Safety Board’s final report, published in December 2023, was that the company “failed to inform or train its employees to safely respond to a smoldering fire.” One of its recommendations was to train all employees on an updated emergency response plan.
A smouldering fire in a mill is exactly the kind of event that sits in a procedure for years without being used. The lesson for a multi-branch operation is not about dust. It is that the procedures least likely to be practised are often the ones with the worst consequence when they are needed, and nothing in the normal working day will tell you they have gone stale.
Sort rare procedures by how fast the skill fails
Not every seldom-used procedure needs the same attention, so rank them rather than refreshing all on one cycle. Three questions do most of the sorting. How often does anyone actually perform it? How bad is a wrong first move? And how much of it is a physical sequence rather than a decision?
A procedure that is never used, carries severe consequences and is mostly physical, such as evacuating a floor, using an extinguisher or responding to a choking guest, decays fastest and needs the shortest interval. A procedure that is rarely used but is essentially a decision with a clear escalation point, such as refusing a suspect delivery or handling a customer allergy disclosure after the order has gone in, holds up better and can sit on a longer cycle. Write the list once, with your own procedures on it, and the refresh calendar follows from it instead of from habit.
Separate the part that can be refreshed on a phone from the part that needs hands
Short digital training is good at the parts that fade first in a decision: which step comes first, who to call, what not to do, where the escalation line sits. It can deliver the same approved version to every branch and record who answered what.
It cannot confirm that someone can lift a casualty, operate an extinguisher, isolate a machine or perform compressions. For those, the refresh has to be physical, with a supervisor or trainer watching, and its record is separate from any digital result. Deciding which half of a procedure you are refreshing is what stops a quarterly module from being mistaken for a drill.
Let an event trigger a refresh, not only the calendar
A fixed cycle misses the moments that actually raise risk. A near miss, a complaint, a new layout after a renovation, different equipment, a change of supplier, a branch that has taken on several new hires, a long holiday period staffed mostly by part-timers: each of these makes a rarely used procedure more likely to be needed and less likely to be remembered. Treat them as triggers in their own right, and record the refresh against the trigger.
Rehearse with the shift that would actually be on duty
Drills tend to be scheduled when management is present, which usually means a weekday morning. Rare emergencies are not scheduled at all. If the evening and overnight crews have never rehearsed the evacuation, the organisation has tested the shift least likely to face it alone. Rotate which shift the drill lands on, and treat a never-rehearsed shift as a gap in the same way as an expired certificate.
What improves when the interval is set deliberately
The refresh calendar stops being a list of certificate expiry dates and starts reflecting which procedures are actually at risk. You find out that a sequence has been lost before an incident finds out for you. Branches and shifts with gaps become visible individually rather than averaged away, and when an auditor asks why the interval is what it is, there is a reason on the record rather than a convention.
Where Wisteria fits
Wisteria is built for the part of this that is repeatable. It checks connected sources and flags documents that may need training, so a revised emergency procedure does not sit unnoticed, then prepares draft flashcards and questions from the approved document: automatically found, human approved, with the trainer reviewing and publishing before anything reaches staff. Short refreshers reach the relevant roles on the phones employees already use, including the shifts a trainer rarely meets, assessments record results rather than attendance, and recertification can be reassigned when the interval or the source document changes. The physical rehearsal stays with your supervisors, and the observation record sits alongside it.
What to do next
If you would like to see what this looks like with your own material, send us one SOP for a procedure your teams seldom use. We will build a short set of sample flashcards and understanding-check questions from it, split by the roles it affects, and mark which parts we would not attempt to cover digitally. You can review the result before deciding anything.
Carmen Lee — Instructional Designer, Wisteria
Six years in the training industry, and someone who cares about frontline learning.
