Table 1.

Differences in practice characteristics between practices that were invited into the study versus practices that were not invited into the study, practices that accepted participation into the study versus practices that declined to participate among those invited, and practices that recruited women into the study versus those that did not recruit women into the study among those which had agreed to participate.

		Practice invited
		Practice accepted participation				Practice did not accept participation		Practice not invited
		Practice recruiting		Practice not recruiting
		N = 125	12.97%	N = 65	6.74%	N = 111	11.51%	N = 663	68.78%
Number of patients			3.23 (0.26–9.63)		2.13 (0.19–6.01)		0.71 (0.06–1.96)
	<2000	36	28.80	31	47.69	53	47.75	328	49.47
	≥2000	82	65.60	29	44.62	43	38.74	277	41.78
	Missing	7	5.60	5	7.69	15	13.51	58	8.75
Type of practice			3.67 (0.92–9.24)		2.96 (0.86–6.96)		1.73 (0.74–3.35)
	Single	42	33.60	36	55.38	71	63.96	430	64.86
	Group	59	47.20	15	23.08	27	24.32	180	27.15
	Other	24	19.20	14	21.54	13	11.71	53	7.99
Geographic location			0.00 (0.00–2.23)		2.14 (0.22–6.14)		4.86 (2.61–7.79)
	City	56	44.80	28	43.08	68	61.26	523	78.88
	Country	60	48.00	31	47.69	36	32.43	140	21.12
	Missing	9	7.20	6	9.23	7	6.31	0	0.00
Average age of doctors			2.47 (0.38–7.94)		0.02 (0.01–1.65)		0.20 (0.02–1.01)
	≤50 years	47	37.60	14	21.54	33	29.73	190	28.66
	51–60 years	49	39.20	25	38.46	39	35.14	237	35.75
	>60 years	22	17.60	21	32.31	25	22.52	186	28.05
	Missing	7	5.60	5	7.69	14	12.61	50	7.54
Sex of doctors			6.22 (2.71–12.33)		2.44 (0.84–6.04)		0.45 (0.11–1.44)
	Male	23	18.40	32	49.23	41	36.94	238	35.90
	Female	36	28.80	14	21.54	35	31.53	222	33.48
	Both sexes	59	47.20	14	21.54	21	18.92	153	23.08
	Missing	7	5.60	5	7.69	14	12.61	50	7.54

The boxes show the uncertainty coefficient in % (95% confidence interval in brackets), which quantifies the difference in distribution and thereby the relative strength of the selection effect in each step for each characteristic of the women. The uncertainty coefficient builds on Goodman and Kruskal’s classic review of association measures [18].