Publication and non-publication of clinical trials in PTSD: an overview

Table 6 Cox regression of factors involved in publication time

Variable	B	Wald χ²	Significance	Relative risk (95% CI)
Randomised	− 0.21	0.20	0. 653	0.82 (0.33; 1.99)
Treatment arms^a		1.87	0.392
2 arms	0.55	0.87	0.351	1.74 (0.54; 5.56)
3–6 arms	0.36	0.34	0.562	1.43 (0.43; 4.73)
Any control used	0.095	0.07	0.797	1.10 (0.53; 2.27)
Intervention type^b		18.64	< 0.001*
Behavioural	0.51	8.65	0.003*	1.67 (1.19; 2.34)
Drug and behavioural	1.16	16.79	< 0.001*	3.20 (1.84; 5. 58)
Procedure/device	0.74	5.54	0.019*	2.09 (1.13; 3.87)
Any sponsor, university	0.26	3.88	0.049*	1.30 (1.00; 1.70)
Registered before completed	0.33	4.10	0.043*	1.40 (1.01; 1.93)

Model χ2(9) = 32.69 (p < 0.001*)
− 2 log-likelihood = 2271
*Significance set at p < 0.05
^aOne arm was the number of arms against which other arms were analysed
^bDrug intervention was the type of intervention against which other interventions were analysed

ISSN: 2058-8615