What the evidence says
The strongest and most consistent finding in this evidence set is a gap between detection and referral. Across a screening population of 1,152 attendees, formal signposting or referral occurred for only about 9% of attendees (97 of 1,152). Referral rates for specific detected risks were smaller still: 6% of identified smokers were referred to smoking cessation support, 4.7% of attendees (54 of 1,152) were referred for social isolation, and 0.6% (7 of 1,152) were referred for alcohol use. These are exact counts from service activity data reported by a single tier-A source, so confidence in both the direction and magnitude of the referral shortfall is high, but the finding is not corroborated elsewhere in this evidence set.
Screening itself was not fully completed across the population it did reach: about 30% of attendees declined BMI measurement. The source attributes this to a mix of patient discretion and the sensitivity of weight-related discussion, a explanation drawn from staff and student observation and fidelity checks rather than from a separate causal test.
Where screening was completed, it detected substantial prevalence of cardiovascular and metabolic risk: 44% of participants had elevated blood pressure, 62% were overweight or obese, and 36% had cholesterol above 5 mmol/L. The same screening data showed a correlation between area deprivation and these risk indicators — a negative correlation between deprivation and BMI (r = −0.23) and between deprivation and glucose (r = −0.12) — which the source interprets as participants from more deprived areas exhibiting higher risk. These coefficients are weak to moderate in magnitude, and the source does not report their statistical significance, so magnitude confidence is only medium even though the direction is reported plainly.
On reach, the screening population included attendees from all socioeconomic deciles, with 31% drawn from the most deprived quintile.
No claim in this evidence set measures screening or referral rates under varying referral conditions — different referral models, opt-in versus opt-out designs, or staff-prompted versus self-initiated referral pathways. Every figure here comes from a single clinic's service activity data under one referral process, so the comparative question the topic asks about — how referral rates change as the referral condition changes — is not addressed by the current sources.