Table 2.

Cox regression – hazard ratios of discontinuation due to inefficacy, to adverse events, and to any reason

Number of subjects=502
Number of observed treatment period=1026
Standard error adjusted for individual clusters
Variables	Discontinuation due to inefficacy			Discontinuation due to adverse events			Discontinuation due to any reason
	HR	Standard error	p-Value	HR	Standard error	p-Value	HR	Standard error	p-Value
Order of treatment	1.272	0.087	<0.001	1.349	0.128	0.002	1.260	0.070	<0.001
Age	1.007	0.004	0.066	1.006	0.006	0.361	1.007	0.003	0.015
Male	0.898	0.153	0.528	0.620	0.142	0.036	0.769	0.097	0.037
RF	1.048	0.138	0.725	0.900	0.154	0.536	1.003	0.094	0.973
Anti-CCP	0.878	0.115	0.317	0.914	0.158	0.601	0.898	0.085	0.257
Steroid	0.931	0.108	0.537	1.112	0.183	0.519	0.985	0.084	0.856
sDMARD	0.818	0.121	0.175	0.585	0.127	0.014	0.707	0.077	0.001
Years of disease	1.000	0.006	0.972	1.015	0.007	0.037	1.006	0.004	0.109
Gap above 30 days	0.604	0.116	0.009	0.829	0.207	0.454	0.730	0.088	0.009
Biological
ETA	1.220	0.444	0.584	0.630	0.261	0.265	0.767	0.181	0.26
ADA	1.772	0.633	0.109	0.554	0.228	0.152	0.976	0.225	0.916
RTX	0.450	0.172	0.036	0.505	0.234	0.139	0.434	0.108	0.001
INF	1.552	0.565	0.227	1.040	0.416	0.923	1.088	0.256	0.72
TCZ	0.597	0.257	0.23	0.474	0.246	0.151	0.455	0.137	0.009
GOL	0.942	0.392	0.886	0.918	0.426	0.854	0.746	0.209	0.295

Note: Bold: significant at 0.05 level.

Abbreviations: ADA, adalimumab; CCP, cyclic citrullinated peptide; ETA, etanercept; GOL, golimumab; HR, hazard ratio; INF, infliximab; RF, rheumatoid factor; RTX, rituximab; sDMARD, synthetic disease-modifying antirheumatic drug; TCZ, tocilizumab.