15250-m-cuypers

139 Regret and information satisfaction at one-year follow-up 7 Table 1. Continued DA group Control group P -value Complete participation Lost to follow-up P -value N= 235 N=100 N=270 N=65 Hospitals 1 20 (7%) <.001 9 (3%) 6 (7%) .26 2 1 (1%) 0 (0%) 1 (1%) 3 50 (18%) 37 (14%) 10 (12%) 4 34 (13%) 19 (7%) 12 (15%) 5 16 (6%) 10 (4%) 5 (6%) 6 21 (8%) 13 (5%) 4 (5%) 7 72 (25%) 53 (20%) 13 (16%) 8 38 (14%) 31 (11%) 6 (7%) 9 21 (8%) 18 (7%) 3 (4%) 10 6 (6%) 6 (2%) 0 (0%) 11 20 (18%) 13 (3%) 1 (1%) 12 10 (9%) 8 (3%) 1 (1%) 13 10 (9%) 8 (3%) 2 (2%) 14 24 (22%) 19 (7%) 5 (6%) 15 8 (7%) 5 (2%) 3 (4%) 16 23 (21%) 17 (6%) 3 (4%) 17 8 (7%) 5 (2%) 3 (4%) 18 Merged with hospital 14 No differences were observed in socio-demographic- or clinical characteristics (PSA, Gleason), or the degree of involvement to the treatment decision process between participants from both trial arms. Patients in the DA group chose AS or surgery more often and radiotherapy less often compared to patients from the control group. Although it was aimed to have equal sized samples in both arms and across hospitals, fewer patients were recruited for the control group, and patient enrollment varied across hospitals (Table 1). Post-hoc power analysis revealed sufficient power at all time points (92-87% power) to detect clinically relevant differences ( d =0.4) between trial arms at all time points, but low power (38-34%) to detect small effects ( d =0.2). Overall, 15%of the participants (n=57) indicated regret (with a score >25) at six, and 19% (n=43) at twelve months after treatment decision-making. But the largest proportion of men from both trial arms reported zero regret at both time points (T2: 36%, n=108, T3: