Table 4.

Logistic regression analyses for the primary outcomes: ACR20 and 35% improvement in patient GA^a

Contrast	Contrasts evaluated		Odds ratio	s.e.	95% Cl	P-value
	Groups 2 and 3	Groups 4 and 5
Consultation vs no consultation	n = 30	n = 31
No. of patients who achieved ACR20, n (%)	7 (23.3)	4 (12.9)	0.404	0.298	0.095, 1.714	0.219
No. of patients who achieved 35% patient GA, n (%)	12 (40.0)	10 (32.2)	0.725	0.424	0.231, 2.280	0.582

	Groups 2 and 4	Groups 3 and 5
Complex vs placebo	n = 29	n = 32
No. of patients who achieved ACR20, n (%)	4 (13.8)	7 (21.9)	2.050	1.494	0.492, 8.548	0.324
No. of patients who achieved 35% patient GA, n (%)	10 (34.5)	12 (37.5)	1.142	0.652	0.373, 3.497	0.816

	Group 1	Group 2
Individual vs complex	n = 16	n = 14
No. of patients who achieved ACR20, n (%)	5 (31.3)	2 (14.3)	0.258	0.259	0.036, 1.842	0.177
No. of patients who achieved 35% patient GA, n (%)	6 (37.5)	6 (42.9)	1.076	0.863	0.224, 5.181	0.927

	Group 1	Group 3
Individual vs placebo	n = 16	n = 16
No. of patients who achieved ACR20, n (%)	5 (31.3)	5 (31.2)	1.262	1.045	0.249, 6.394	0.778
Number of patients who achieved 35% patient GA, n (%)	6 (37.5)	6 (37.5)	1.047	0.811	0.229, 4.781	0.953

^aAdjusted for practitioner, baseline disease severity, baseline disease activity, sex, age, duration of illness, number of medications for RA and number of other medical complaints.