Investigating the effects of medical density on health-seeking behaviours using a multiscale approach to residential and activity spaces: Results from a prospective cohort study in the Paris metropolitan area, France

Table 6 Sensitivity analysis of the association between cervical screening and residential medical density according to the delineation of residential neighbourhoods

			Medical density tertiles	Sample size	% with delayed cervical screening	Difference (p-value)	aOR (95% CI)²	AIC
Neighbourhood units	Administrative areas	Census blocks	Highest and middle	440	15.0	NS	Ref.	501.11
			Lowest	222	12.6		0.91 (0.51-1.61)
		Adjacent census blocks	Highest and middle	437	13.7	NS	Ref.	501.09
			Lowest	225	15.1		1.11 (0.63-1.96)
		Municipalities	Highest and middle	424	13.4	NS	Ref.	501.09
			Lowest	238	15.5		1.11 (0.63-1.94)
	Regular grids	600 × 600 metres¹	Highest and middle	441	13.6	NS	Ref.	500.82
			Lowest	221	15.4		1.19 (0.69-2.04)
		1000 × 1000 metres¹	Highest and middle	440	13.9	NS	Ref.	501.21
			Lowest	222	14.9		0.97 (0.56-1.67)
		1400 × 1400 metres¹	Highest and middle	437	12.1	< 0.05	Ref.	498.73
			Lowest	225	18.2		1.56 (0.89-2.75)
		1800 × 1800 metres¹	Highest and middle	434	13.1	NS	Ref.	501.15
			Lowest	228	16.2		1.08 (0.62-1.86)

Results of the bivariate analyses and multilevel logistic regression models.
¹ From an initial 200 x 200 metres grid. Every grid was centred on the cell corresponding to respondent’s place of residence.
² Adjusted odds ratio accounting for age, level of education and activity space structure (as defined using a 0.7 threshold value).

ISSN: 1476-072X