Neighbourhood Profile

“Expected”is a model, not a target — and it runs entirely on data we already hold. For each metric we fit an ordinary least-squares regression of the metric on each practice’s catchment-weighted IoD2025 deprivation and its registered %65+, across 6,142 English practices, then read off what a practice with thiscatchment’s deprivation and age profile would typically achieve. The residualis the actual result minus that expectation: a fair, like-for-like view that doesn’t penalise a practice for serving a more-deprived or older list, and is not a flat national average.

Model used here: catchment deprivation (age added nothing measurable here). Where deprivation does not vary enough to model, the expected value degrades transparently to the plain average. It is a modelled estimate (labelled as such), not a “feed required” item — no new data is needed to compute it.