Skip to main content

Table 3 Estimates (95 % CIa) from the joint longitudinal-competing risks model for scenario 1 (conservative) and scenario 2 (worst)

From: Modelling variable dropout in randomised controlled trials with longitudinal outcomes: application to the MAGNETIC study

Scenario

Longitudinal outcome

Dropout due to good prognosis

Dropout due to poor prognosis

β 1 (95 % CI)

HR(1) (95 % CI)

γ (1) (95 % CI)

HR(2) (95 % CI)

γ (2) (95 % CI)

1 (n = 502 case 1: 25 case 2: 14)

−0.165 (−0.336 to 0.011)

1.915 (0.820 to 3.507)

−0.768 (−1.340 to −0.299)

0.801 (0.122 to 1.872)

0.200 (−0.436 to 0.715)

2 (n = 502 case 1: 15 case 2: 24)

−0.163 (−0.363 to 0.010)

2.125 (0.845 to 3.904)

−1.389 (−2.021 to −0.938)

1.069 (0.350 to 2.148)

0.159 (−0.361 to 0.588)

  1. a95 % bias-corrected percentile CIs are obtained from 1000 bootstrap resamples