Table 2.

Primary outcome and sensitivity analysis (full-analysis set).

Analysis	SoC (n = 156)	BUP-XR (n = 158)	Difference (SE)	Adjusted IRR (95% CI)	p-value
Primary outcome measure—mean days abstinent from opioids	104.37 (9.89)	123.43 (4.76)	19.05 (5.48)	1.18 (1.05–1.33)	0.004
Primary outcome measure—sensitivity analysis
Adjustment of analysis period
Including 1-week grace perioda	111.86 (10.36)	130.44 (4.89)	18.58 (5.84)	1.16 (1.04–1.31)	0.008
TLFB data truncated by 14 days at endpointb	96.72 (9.28)	114.37 (4.32)	17.65 (5.27)	1.18 (1.05–1.33)	0.006
Duration of treatment episode at study enrolmentc	104.27 (9.71)	123.09 (4.68)	18.82 (6.03)	1.18 (1.05–1.32)	0.004
Inclusion of baseline predictors of missing data
ADAPT ‘addiction severity’ score and QUID-SR score	102.47 (5.42)	119.92 (1.19)	17.45 (6.03)	1.17 (1.05–1.31)	0.007
Missing TLFB data scenario
Best case—all missing data indicate opioid abstinence	131.19 (11.00)	152.34 (3.56)	21.15 (7.85)	1.16 (1.03–1.31)	0.018
Worst case—all missing data indicate opioid use	94.37 (7.59)	133.41 (7.28)	39.04 (10.40)	1.41 (1.17–1.71)	0.001
Missing UDS data scenario
Best case—all missing data indicate opioid abstinence	122.04 (12.61)	150.06 (4.55)	28.02 (9.15)	1.23 (1.05–1.44)	0.009
Worst case—all missing data indicate opioid use	121.98 (12.68)	149.96 (4.65)	27.98 (9.16)	1.23 (1.05–1.44)	0.010

All analysis is mixed-effects, negative binomial regression with baseline drug injecting status (fixed effect) and treatment clinic (random intercept).

Difference is calculated as BUP-XR minus SoC.

ADAPT, Addiction Dimensions for Assessment and Personalised Treatment; BUP-XR, extended-release buprenorphine; IRR, incidence rate ratio; QIDS-SR, Quick Inventory of Depressive Symptomatology–Self-Report; SE, standard error; SoC, the standard of care; TLFB, Timeline Follow-back; UDS, urine drug screen.

All means are adjusted.

^aAnalysis period is day 0–168.

^bAnalysis period is day 8–154.

^cDays in treatment at study enrolment.