Table 4. Multivariable multinomial logistic regression of factors associated with delayed colonoscopy, no colonoscopy, or no information about colonoscopy, with early colonoscopy (n=998) as the reference category.

	Groupsa
	Delayed colonoscopy n=989		No colonoscopy (NC) n=102		No information (NI) n=119		NC vs NI	NC and NI pooled n=221
	OR (95% CI)	P-valueb	OR (95% CI)	P-valueb	OR (95% CI)	P-valueb	P-valuec	OR (95% CI)	P-valueb
Level 1: individual
FOBT number
First screening FOBT	1.61 (1.16–2.25)	<0.01	1.96 (0.77–4.99)	0.16	2.06 (0.79–5.35)	0.14	0.97	2.01 (1.02–3.97)	0.04
Second screening FOBT	1.00		1.00		1.00			1.00
FOBT delivery mode
Delivered during a GP visit	1.00	0.85	1.00	<0.01	1.00	<0.001	0.28	1.00	<0.001
FOBT mailed to the individual's home	1.02 (0.82–1.28)		1.94 (1.25–3.01)		2.74 (1.78–4.21)			2.29 (1.67–3.14)
Level 2: Area
Townsend index
Quintile 1 (least deprived)	1.00	0.60	1.00	<0.05	1.00	<0.001	0.68	1.00	<0.001
Quintile 2	1.04 (0.77–1.41)		1.39 (0.71–2.72)		1.70 (0.83–3.52)			1.52 (0.92–2.51)
Quintile 3	1.01 (0.74–1.38)		1.80 (0.94–3.45)		2.45 (1.22–4.89)			2.07 (1.28–3.36)
Quintile 4	0.91 (0.66–1.26)		1.33 (0.66–2.66)		1.52 (0.72–3.23)			1.42 (0.84–2.38)
Quintile 5 (most deprived)	1.19 (0.86–1.64)		2.29 (1.20–4.37)		4.37 (2.23–8.55)			3.17 (1.98–5.08)
GP density per 100 000 population
⩽52.7	1.00	0.11	1.00	0.46	1.00	0.37	0.90	1.00	0.27
>52.7	1.18 (0.96–1.46)		1.16 (0.76–1.76)		1.21 (0.78–1.81)			1.19 (0.88–1.61)

Abbreviations: CI=confidence interval; FOBT=faecal occult blood testing; GP=general practitioner; NC=no colonoscopy; NI=no information; OR=odds ratio.

^aEarly colonoscopy was the reference category.

^bP-trend for ordinal variables and p of heterogeneity for non-ordinal variables.

^cP of heterogeneity obtained by multinomial logistic regression with robust variance.