Method note
In developmentPreference fidelity
The proportion of a person's stated preferences that were honoured in the care they received, and where they were not, whether a reason was recorded.
It is a measurement method, not a feature. It measures the organisation.
How the figure is derived
01
Countable preferences
Only preferences that were stated and recorded before the decision are counted. A preference inferred after the fact is not a preference; it is a reconstruction.
02
Matched decisions
Each recorded preference is paired with the decision that touched it. A preference with no corresponding decision is out of scope for the period, not a failure.
03
Honoured or departed
The pairing resolves one way or the other. Partial cases — seen in English with an interpreter where Urdu was asked for — are counted as departures, because a reason is owed.
04
Reasoned or unreasoned
Each departure is then either reasoned, with a named person and a time, or unreasoned. The second figure is the one that matters to an inquiry.
fidelity = honoured / (honoured + departed) reason coverage = reasoned departures / departures
Two figures, reported together. A service with low fidelity and full reason coverage is behaving defensibly. A service with high fidelity and thin reason coverage cannot show that it was.
What it does not do
- It does not score a patient.
- It does not measure symptoms, mood, wellbeing, or anything a person reports about themselves.
- It does not assess anyone's mental state.
- It is not a quality-of-care measure, and a departure is not an error. Services depart from preferences for sound reasons every day.
- It does not produce a league table. The figures are only interpretable alongside the reasons.
Why this figure, and why now
- Mental Health Act 2025 — duty to give reasons where an advance choice is not followed
- Care Quality Commission — mental health assessment framework, continuous oversight model
- Lampard Inquiry — reporting 2027, implementation forum established