Table 5 The efficacy as the overestimated percentage of patients with an ACR20 response to anti-TNF in the active drug group (Active) and the DREAM patients, eligible and ineligible for the RCTs

Study	Active	Eligible DREAM patients				Ineligible DREAM patients
Study	Active	n	ACR20	Δ	p Value*	n	ACR20	Δ	p Value*
Adalimumab + MTX
Weinblatt, 200316	67,2	115	67.8	0.6	0.5332	38	31.6†	−35.6	0.0002
Furst, 200318	52,8	123	67,5	14.7	0.9971	30	23.3†	−29.5	0.0011
Keystone, 200417	63,3	51	68.6	5.3	0.7605	102	53.9	−9.4	0.0562
Adalimumab mono
Van de Putte, 200319	57	9	50.0	−7.0	0.3527	17	35.3	−21.7	0.0540
Van de Putte, 200420	46	8	22.2	−23.8	0.0833	22	31.8	−14.2	0.1096
Etanercept + MTX
Weinblatt, 199910	71	109	54.1	−16.9	0.0164	37	32.4	−38.6	0.0001
Lan, 200411	90	80	55.0	−35.0	0.0004	71	35.2	−54.8	0.0000
Etanercept mono
Moreland, 19974	75	14	41.2	−33.8	0.0064	25	37.5	−35.0	0.0020
Moreland, 199912	59	17	35.7	−23.3	0.0531	24	40.0	−21.5	0.0323
Keystone, 200413	63	34	47.1	−15.9	0.0434	8	12.5	−50.5	0.0022
Infliximab + MTX‡
ATTRACT, 99-0415	50	23	39.1	−10.9	0.1763	36	36.1	−13.9	0.0799

DREAM, Dutch Rheumatoid Arthritis Monitoring; MTX, methotrexate; RCT, randomised controlled trial.
*Bonferroni correction resulted in a p value <0.0045 (0.05/11) being considered significant compared with the active drug group.
†Significant (p value <0.0045 (0.05/11)) difference between response percentages of eligible and ineligible patients.
‡Maini et al14 presented only Paulus criteria in their articles.