Analysis approaches to address treatment nonadherence in pragmatic trials with point-treatment settings: a simulation study

Table 5 Parameterization of simulation for data generating process for simulation setting 1 outlines in DAG 1 (exclusion restriction satisfied and no unmeasured confounding)

#	Arm	α ₀	α ₁	α ₂	α ₃	Nonadherence	θ ₀	θ ₄
1	Z = 1	0.72	0.6	0.4	0.35	10	-1	0.5
	Z = 0	-4.06	0.6	0.4	0.35	10	-1	0.5
2	Z = 1	-0.23	0.6	0.4	0.35	20	-1	0.5
	Z = 0	-3.14	0.6	0.4	0.35	20	-1	0.5
3	Z = 1	-1.47	0.6	0.4	0.35	40	-1	0.5
	Z = 0	-1.92	0.6	0.4	0.35	40	-1	0.5
4	Z = 1	-2.52	0.6	0.4	0.35	60	-1	0.5
	Z = 0	-0.85	0.6	0.4	0.35	60	-1	0.5
5	Z = 1	-3.76	0.6	0.4	0.35	80	-1	0.5
	Z = 0	0.39	0.6	0.4	0.35	80	-1	0.5
6	Z = 1	-4.72	0.6	0.4	0.35	90	-1	0.5
	Z = 0	1.35	0.6	0.4	0.35	90	-1	0.5
7	Z = 1	0.72	0.6	0.4	0.35	10	-5.5	8
	Z = 0	-4.06	0.6	0.4	0.35	10	-5.5	8
8	Z = 1	-0.23	0.6	0.4	0.35	20	-5.5	8
	Z = 0	-3.14	0.6	0.4	0.35	20	-5.5	8
9	Z = 1	-1.47	0.6	0.4	0.35	40	-5.5	8
	Z = 0	-1.92	0.6	0.4	0.35	40	-5.5	8
10	Z = 1	-2.52	0.6	0.4	0.35	60	-5.5	8
	Z = 0	-0.85	0.6	0.4	0.35	60	-5.5	8
11	Z = 1	-3.76	0.6	0.4	0.35	80	-5.5	8
	Z = 0	0.39	0.6	0.4	0.35	80	-5.5	8
12	Z = 1	-4.72	0.6	0.4	0.35	90	-5.5	8
	Z = 0	1.35	0.6	0.4	0.35	90	-5.5	8

ISSN: 1471-2288