We randomly assigned a total of 508 patients to receive sequential monotherapy (group 1, n = 126), step-up combination therapy (group 2, n = 121), initial combination therapy with prednisone (group 3, n = 133), or initial combination therapy with infliximab (group 4, n = 128) (Figure 1). At baseline, the groups were balanced with respect to demographic and disease characteristics (Table 1). Enrolled patients had a median disease duration of 23 weeks (interquartile range, 14 to 53 weeks) and had active disease with mean disease activity and HAQ scores of 4.4 (SD, 0.9) and 1.4 (SD, 0.7), respectively. Seventy-two percent of patients had joint erosions at baseline. Over time, 27 patients who were equally distributed across the treatment groups (P = 0.474) were lost to follow-up: 12 withdrew consent (7 declined follow-up, 4 discontinued all medications despite having no adverse events, and 1 moved from the area), 7 had a revised diagnosis, 1 discontinued treatment because of an adverse event, 4 died, and 3 were lost to follow-up for other reasons (2 were admitted to a nursing home and 1 wanted to become pregnant) (Figure 1). Furthermore, 12 (10%), 11 (9%), 14 (11%), and 6 (5%) patients in groups 1, 2, 3, and 4, respectively (P = 0.343), did not adhere to the treatment protocol but were included in the intention-to-treat analysis.