TL;DR / Summary: Letter-to-number coding is a training drill, not a security system. Used in handover exercises, it forces staff to slow down, confirm the right patient, and read back a value — three habits that reduce real errors.
Why a simple cipher works as a drill
Handover errors are usually not knowledge failures. They are attention failures: the wrong bed number, a misheard dose, a name that sounded like another name.
A drill that converts a short identifier into letters and numbers forces attention at the exact point where attention usually slips. It is deliberately slow, and the slowness is the point.
Nothing is encrypted. The exercise simply removes the ability to coast on expectation.
Designing the drill
Take a non-clinical identifier — a bed number, a shift code, a training room — and encode it as letters. Teams decode it, read it back, and confirm.
Keep the encoded strings short. Six characters is enough to require care and short enough to complete in a few seconds.
A workable exercise
- Encode: bed and bay as letters, shift as a number
- Transmit: one person reads the code aloud once
- Decode: the receiver writes the answer
- Read back: the receiver states the decoded value aloud
- Confirm: the sender confirms or corrects
Five steps, under a minute. Run it for ten minutes at the start of a training day and the read-back habit carries into the ward.
Read-back is the real lesson
The cipher is a vehicle. Read-back is the skill, and it is the one that prevents medication and identification errors.
In the drill, a wrong decode is caught by the confirm step. In practice, the same step catches a misheard dose before it is administered.
The value of a training drill is the habit it leaves behind, not the puzzle it poses.
Make the confirm step non-optional. A receiver who guesses instead of asking is the failure mode the drill exists to expose.
Common decoding errors
Letters that sound alike cause most mistakes when codes are spoken: B and P, D and T, M and N. Choose words for the alphabet that are phonetically distinct.
Numbers behave similarly. Fifteen and fifty, thirteen and thirty. In spoken drills, write the digits rather than saying them, or use a phonetic alphabet deliberately.
Watch for transposition. AB becomes BA under time pressure, and transposition is exactly the error class the drill is designed to reveal.
Keeping it non-clinical
Use training data only. Never put a real patient identifier into an exercise, even a fictional coded one, because the habit of encoding real identifiers is the opposite of data protection.
Label the exercise clearly as training, and keep the answer sheets in the training folder rather than anywhere clinical.
If the session runs in a clinical area, use a room number that does not exist. That removes any chance of confusion during a live shift.
Scaling it to a team
Run it in pairs first, then in a chain of four where the message passes through two relays. Chains expose the degradation that handovers suffer in real life.
Time each round and compare accuracy, not speed. Fast and wrong is the behaviour the drill is meant to discourage.
Debrief with one question: where did the message change? The answer is usually at a relay, which is precisely where real handover errors occur.
For the encoding itself, the letters to numbers converter saves preparation time, and printed code cards keep the exercise offline and quick to reset.
Then stop the drill while people are still engaged. Ten focused minutes leaves a habit; forty becomes an ordeal nobody repeats.
Debriefing the drill
After the exercise, ask each pair which step they nearly got wrong. The answers point at the same three places: hearing a digit, writing it down, and reading it back.
Do not ask who got it right. Correct answers teach nothing; near-misses teach the whole room.
Keep the debrief to five minutes. Longer and it becomes a lecture, and the drill’s value is in the doing.
The near-miss is the curriculum.
Record the common near-misses on a flip chart and keep the sheet. Repeating the same drill in a month shows whether the habits held.
If a specific digit pair caused most errors, add it deliberately to the next session. Targeted repetition works faster than general encouragement.
Making it part of induction
A drill run once is entertainment. A drill in the induction pack is a standard.
Add the exercise to the first-week schedule with a one-page script: encoding rules, the read-back step, and the three common errors to watch for. New staff then arrive knowing the routine.
Training that is not in the induction is training that depends on enthusiasm.
Ask a recent starter to run it for the next cohort. Peer delivery is more credible than a manager repeating the same slides.


