Table 2.

Differences in incidence rates of diagnoses of non-diabetic hyperglycaemia and type 2 diabetes by general practices’ coverage of the NHS Health Check Programme and patients’ baseline diabetes risk score.

	Incidence Rate	Total Sample			DRS ≥ 10
		HR	95% CI		HR	95% CI
Non-Diabetic Hyperglycaemia
Programme Coverage§
Low	0.19 per 1000 person-years (N = 2,096)	ref			ref
Medium	0.25 per 1000 person-years (N = 2,693)	1.15***	1.08	1.22	1.15**	1.04	1.27
High	0.29 per 1000 person-years (N = 2,337)	1.19***	1.11	1.27	1.23***	1.11	1.37
Type 2 Diabetes	Crude Rates	HR	95% CI		HR	95% CI
Programme Coverage§
Low	0.35 per 1000 person-years (N = 4,018)	ref			ref
Medium	0.42 per 1000 person-years (N = 4,482)	1.10***	1.05	1.15	1.13**	1.05	1.21
High	0.45 per 1000 person-years (N = 3,671)	1.09***	1.03	1.14	1.10*	1.02	1.19

Time period was Jan 2009-May 2016. Results are shown from multivariable Cox regression models. All models have been adjusted for the baseline values of the following independent variables: age, gender, ethnicity, smoking status, body mass index, antihypertensive medication, general practice deprivation score, and region. Models have also been adjusted for propensity score based on patients’ probability of being registered with a practice with low, medium or high coverage of the Health Check Programme. Abbreviations: DRS = diabetes risk score, HR = Hazard ratio. Legends: *p < 0.05, **p < 0.01, ***p < 0.001, ^§Tertiles of general practices coverage of the NHS Health Check Programme were defined based on programme coverage during the first three years after its implementation (2009–11).