Handover under pressure: reframing a 'medical vocabulary' brief

Reframe: this is a genre problem, not a lexis gap. The nurses already retrieve the words; the failure is selecting, sequencing and delivering information to a listener under time pressure. Handover is a spoken genre with recognisable moves - identify the patient, presenting problem, relevant history, current observations and trend, risks, and the action or request. Breaking down is a discourse and strategic competence problem plus fluency under pressure. The previous course tested recognition of terms, which is why it produced a pass rate and no behavioural change: the assessment did not sample the target performance.

Instruments, fitted to shift work

1. Diagnostic performance sample. Ask each nurse to give a 60-second handover of a supplied case. Record it, rate it against a rubric. Twelve nurses × 15 minutes can be done across shift patterns and gives a genuine baseline plus a visible gap for the learner.
2. Observation of real handovers at two or three shift changes, with consent and anonymised notes. Note where summaries unravel and whether interruptions derail them. This is the richest data and the least convenient.
3. Document collection. Ten anonymised handover sheets or notes to identify the language the genre actually uses.
4. Separate interviews with three nurses and the ward manager. Treat their accounts as different data, not as one truth; the manager sees outcomes, the nurses see constraints.
5. Self-assessment against a handover-specific can-do grid, which also builds buy-in.

Design implication of the constraints: rotating attendance means the course must be modular and each session self-contained but cumulative. No homework means all practice is in the room, so the two-hour block has to be mostly active use, not input.

Organising principle: target genres, recycled in a spiral. Select three or four performance genres - shift handover, telephone referral, explaining a procedure to a patient, and written documentation - and revisit each twice at increasing demand rather than 'covering' it once. Lexis is taught because a move requires it, which also answers the ward manager: she will see the vocabulary in use, which is what she actually wants.

Representative sequence: handover (about 90 minutes)

1. Notice the genre. Two short recorded handovers of the same case, one expert and one weak. Learners identify what the expert did and where the weak one lost the listener. No vocabulary list yet.
2. Language focus drawn from those texts. Tense and aspect for recent events versus current state, quantifying observations and trends ('BP 90 over 50, down from 110'), signposting, and the specific chunks that carry the moves.
3. Controlled practice with a twist. Case cards where one detail changes at the last moment, so learners must re-select rather than recite.
4. Pushed output. 60-second timed handover, recorded on a phone, rated by a peer against the shared rubric and then by the teacher.
5. Feedback and re-run. Same case, second attempt, comparing recordings. Progress is audible, which matters for adults who feel their English is fixed.
6. Transfer with interference. A new case, and the listener interrupts with a question mid-handover. This is what actually happens on a ward.

Why it addresses the failure: it trains selection and sequencing under time constraint, it rehearses interruption, and it is short enough to survive a two-hour release. The rubric is the same one used in the diagnostic, so the learner can see the distance closed.

The demand is legitimate but needs precision. A level on a certificate is a claim, and a claim requires evidence that will survive scrutiny. An end-of-course test on covered content is not that evidence; it is a test of attendance. The previous provider's pass rate is the cautionary case: it certified vocabulary and the ward still failed.

Proposal

1. Performance-based assessment - two recorded handovers and one telephone referral, rated on an analytic rubric mapped to CEFR spoken production and interaction descriptors at B1. This supports a defensible statement such as 'performs reliably at B1 in the assessed occupational tasks'.
2. Rater reliability. Two raters on a sample, a moderation meeting, and standardised rating conditions. Without this, a level claim is personal opinion with a logo on it.
3. State the limits in writing. The exit report to the trust should say what was assessed and what was not. A certificate that overclaims will eventually be tested by a real ward, and the provider will lose the contract.

Washback risk. If the trust insists on a written vocabulary test because it is easy to administer, teaching will revert to lists within a week; the assessment will become the syllabus. Argue this explicitly, and offer the performance assessment as the alternative rather than simply refusing.

The two agency nurses. Design the course so assessment points fall at the midpoint as well as the end, so a leaver still has certified evidence of what they achieved. Avoid extended pair or group projects that collapse when one member leaves, and avoid making any single learner the linchpin of a task. Confirm with the trust at the outset who pays for a repeat assessment if an agency nurse returns.