Table 4.

Random effects moderator analysis results for PNS.

Moderator	k1	Effect Size (d)2	95% CI	Z Value	P value	Q betweeen (df)3	P of Q between
Drug targeted for intervention						6.43 (4)	0.17
Cocaine	4	0.4	0.13–0.67	2.89	<0.01
Opiates	3	0.18	−0.11–0.46	1.23	0.22
Opiates and cocaine	2	0.43	0.18–0.67	3.42	<0.01
Tobacco	2	1.02	0.37–1.67	3.09	<0.01
Poly substance	1	0.49	0.23–0.74	3.74	<0.001
Study decade						1.10 (2)	0.58
1990–1999	2	0.51	0.25–0.77	3.83	<0.001
2000–2009	3	0.30	0.01–0.59	2.01	0.05
2010 onwards	7	0.40	0.20–0.60	3.93	<0.001
Study Quality						0.36 (2)	0.84
Stong	1	0.48	0.21–0.75	3.43	<0.01
Moderate	5	0.36	0.06–0.66	2.32	0.02
Weak	6	0.44	0.30–0.58	0	<0.001
Itntervention Duration						0.32 (2)	0.85
<12 Weeks	5	0.47	0.28–0.67	4.73	<0.001
12 Weeks	2	0.42	0.18–0.67	3.35	0.04
>12 Weeks	5	0.37	0.02–0.71	2.06	<0.01
Reinforcer type						0.41 (1)	0.52
Monetary Vouchers	9	0.39	0.23–0.54	4.82	<0.001
Other'	3	0.51	0.17–0.85	2.94	<0.01
Opiate treatment						0.35 (1)	0.55
Methadone	8	0.45	0.30–0.60	6.00	<0.001
Other	4	0.32	−0.08–0.72	1.58	0.12

¹Number of studies, ²Weighted random effects, ³A significant value of Q-between indicates significant differences among effect sizes between the categories of the moderator variable.