Assessing fragility of statistically significant findings from randomized controlled trials assessing pharmacological therapies for opioid use disorders: a systematic review

Table 1 Characteristics of included RCTs, in order of increasing FI (k = 10)

Authors	Year	Outcome	Total sample size [n (n_{arm 1}, n_{arm 2})]	No. lost to follow-up (%)	p-value	Journal impact factor	Fragility index	Overall ROB
Krupitsky et al. [22]	2004	Treatment retention	52 (25, 27)	0 (0)^a	<0.05	3.77	1	Low
Fischer et al. [25]	1999	Treatment retention	60 (31, 29)	0 (0)^a	0.03	6.53	2	Some concerns
Petitijean et al. [17]	2001	Treatment retention	58 (31, 27)	0 (0)^a	0.002	4.50	4	Low
Krook et al. [23]	2002	Treatment retention	106 (51, 55)	7 (6.6)^a	<0.001	6.53	6	Low
Sees et al. [18]	2000	Treatment retention	179 (91, 88)	0 (0)^a	0.01	56.27	7	Some concerns
Kakko et al. [19]	2003	Treatment retention	40 (20, 20)	0 (0)^a	0.0001	79.32	8	Low
Yancovitz et al. [26]	1991	Heroin use	301 (94, 75)	132 (43.9)	<0.001	9.30	11	High
Hartnoll et al. [24]	1980	Treatment retention	96 (52, 44)	8 (8.3)^a	0.001	8.48	12	Low
Hulse et al. [21]	2009	Maintaining therapeutic level of naltrexone in blood	69 (34, 35)	9 (13.0)	<0.001	8.48	15	Low
Oviedo-Joekes et al. [20]	2009	Treatment retention	226 (111, 115)	6 (2.7)^a	<0.001	91.25	26	Low

^aShould be interpreted with caution as primary dichotomous outcome is treatment retention, thus we only report on losses to follow-up for known reasons unrelated to treatment, as it is otherwise difficult to distinguish loss to follow-up from the outcome treatment retention

ISSN: 1745-6215