T08 Audit and improvement project record — Worked example

Hypothetical worked example

Read with chapter 12

Use chapter 12 to define the measure and collection rule. Establish local classification and data arrangements before collecting or sharing records.

Question and governance

QI01 is wholly hypothetical. The team examines whether discharge summaries identify the follow-up action and intended recipient. A real project requires local classification, authorisation and proportionate data handling before collection.

Measure

The numerator is summaries containing both required elements; the denominator is summaries reviewed in that sample. This is a process measure, not a measure of harm prevented. The same explicit definition should be used across samples.

Baseline and change

Baseline: 12 of 30 summaries, 40%. The doctor helped define the question and review findings; colleagues helped design and use the prompt. The prompt was revised after staff reported duplication.

Later samples

First review: 22 of 30, approximately 73%. Later review: 24 of 30, 80%. The difference between baseline and the later sample is 40 percentage points. The figures are invented teaching data.

Interpretation

The small samples and absence of a comparison group limit causal claims. Differences in cases or recording may contribute. The project does not establish improved patient outcomes, and no statistical significance is claimed. Duplication feedback is useful but does not quantify workload.

Ownership and portfolio

Agree who maintains the prompt and checks the measure. Retain the project plan, definitions, aggregate results and a clear account of personal contribution. Do not claim sole ownership of the team's improvement.