Accuracy of the Berger-Exner test for detecting third-order selection bias in randomised controlled trials: a simulation-based investigation

Table 3 Pooled diagnostic Odds ratios (DOR) per study parameter

Study parameter	Parameter setting	Alpha 1%		Alpha 5%		Alpha 20%
		DOR*	95% CI*	DOR*	95% CI*	DOR*	95% CI*
N	120	835	396 - 1763	212	110 - 406	62	33 - 117
	240	698	334 - 1455	242	124 - 474	71	37 - 135
	480	668	325 - 1374	234	122 - 451	65	34 - 122
TF	10%	948	443 - 2027	238	124 - 458	64	34 - 120
	20%	785	378 - 1626	251	130 - 484	69	37 - 131
	50%	547	272 - 1101	204	107 - 391	64	34 - 120
RM	Fixed/BS = 4	995	311 - 3186	500	176 - 1420	113	43 - 299
	Fixed/BS = 6	666	225 - 1976	233	86 - 630	86	32 - 227
	Fixed/BS = 8	566	196 - 1638	185	69 - 498	63	24 - 166
	Varying	723	242 - 2166	182	69 - 494	52	20 - 136
	MP/MTI = 2	1456	432 - 4906	395	108 - 805	72	27 - 189
	MP/MTI = 3	712	239 - 2126	224	83 - 604	53	20 - 140
	MP/MTI = 4	474	160 - 1404	144	54 - 382	43	16 – 114

CI = Confidence Intervals; Fixed = Fixed block randomisation; BS = Block size; MTI = Maximum tolerated imbalance; Fixed = Fixed block randomisation; Varying = Block randomisation with randomly varying block size 4, 6, 8 with equal probability (1/3); MP = Maximal procedure; MTI = Maximum tolerated imbalance; N = Subject number; TF = Trait frequency;
RM = Randomisation method.
*Decimals rounded.

ISSN: 1471-2288