Table 2.

Summary of included studies.

First Author, Year	Country of Study	Aim	Design	Outcome Measures	Sample Size	Demographics	Summary of Findings
Abrahamsson, 2017 [18]	Sweden	Assess whether prescription of sedatives may be associated with mortality in patients in opioid maintenance treatment.	Open cohort	Mortality.	4501	26.2% female. Median (IQR) age at baseline 34.4 (28.7–42.1) yrs.	Association between z-drug and pregabalin prescriptions and overdose deaths in subjects in OAT. For BZD prescriptions (all prescribable BZD included) the association with overdose death was unclear, whereas the association with non-overdose death was significant.
Bakker, 2017 [19]	UK	Describe co-prescribing of BZD and OST, investigate links with retention in care and mortality.	Case note review	Treatment retention, mortality.	278	30.6% female. 81% White, 10% Arab, 4% Asian, 4% Black.	Treatment retention for patients on BZD maintenance treatment was over twice that compared with patients never on BZD prescription (average 72 months vs. 34 months). Lowered mortality in group of patients prescribed BZD briefly/occasionally.
Best, 2002 [20]	UK	Describe patterns of prescribed diazepam and illicit drug use, and levels of anxiety and depression.	Longitudinal	Changes in illicit drug use.	100	30.0% female. Mean age 36.0 yrs (range 22–52 yrs). 96% White.	No change in use of illicit heroin or crack over study period. Prescribed diazepam associated with increased anxiety and depression markers over the 2 years of study; non-prescribed diazepam not associated with changes in anxiety or depressions.
Durand, 2021 [21]	Ireland	Identify determinants of time to dropout from methadone maintenance treatment (MMT) across multiple treatment episodes in specialist addiction services.	Cohort	Time to dropout from MMT at 3 months (90 days) and 12 months (91–365 days).	2035	31.8% female. Median (IQR) age at entry 34.4 (30.2–39.0) yrs.	BZD prescribing (type not specified) was one of several variables studied. Prescribed BZDs were associated with dropout at 12 months as was being male, and number of comorbidities. Low-dose methadone (<60 mg/day) and previous dropout were associated with dropout at 3 and 12 months. Adherence to MMT treatment was protective of dropout at 3 months and 12 months.
Eibl, 2019 [22]	Canada	To assess the impact of prescribed vs. nonprescribed BZD on medication assisted therapy outcomes.	Retrospective cohort study	Retention in treatment.	3692	44.1% female. Mean (SD) age 34.9 (10.03) yrs.	Prescribed BZD (type not specified) had no impact on retention in treatment, but illicit use was associated with poorer outcomes. Non-prescribed BZD use predictive of poorer retention.
Eiroa-Orosa, 2010 [23]	Germany	To analyse the correlation between BZD use, BZD prescription, and treatment outcome among participants in a heroin-assisted treatment trial.	Case–control study	Primary: health improvement (physical and mental); reduction in illicit drug use. Secondary: BZD use; prescription; self-reported addiction severity; psychopathology (anxiety/phobic anxiety focus).	1015	21.1% female. Mean (SD) age 36.5 (6.72) yrs.	Baseline BZD use correlated with lower retention rates but not with poorer outcome. Type of BZD(s) not specified. Ongoing BZD use correlated with poorer outcomes. Significantly better outcomes were found in the course of phobic anxiety symptomatology for those with regular prescription of BZDs. BZD use at entry and during treatment was associated with greater duration and severity of drug problems.
MacLeod, 2019 [24]	UK	To investigate the hypothesis that prescription of BZD in patients receiving OAT would increase risk of mortality overall, irrespective of any increased treatment duration. All 10 BZDs listed in the British National Formulary included.	Observational	All-cause mortality, drug-related poisoning mortality, and mortality not attributable to drug-related poisoning.	12,118	32.7% female. Mean (SD) age at study exit 38.8 (10.4) yrs.	Concurrent prescription of BZD was associated with: (1) Increased duration of methadone treatment (adjusted mean duration of treatment episode 466 days [95% CI 450 to 483] compared to 286 days [95% CI 275 to 297]). (2) Increased risk of drug-related poisoning (adjusted HR 3.34 [95% CI 2.14 to 5.20], p < 0.001), with evidence of a dose–response effect. (3) Significant risk for all cause [HR 1.87, 95% CI 1.55 to 2.25, p < 0.001]. Concurrent prescription of z-drugs showed evidence of an association with increased risk of drug-related poisoning (adjusted HR 1.64 [95% CI 1.02 to 2.64], p < 0.001). Multiple analyses presented in the paper.
Maremmani, 2014 [25]	Italy	Compare long-term outcomes of treatment-resistant heroin addicts (HA) with and without comorbid BZD severe addiction treated with MMT (and CMT for HA + BZD patients). Patients switched to clonazepam.	Controlled cohort study	Retention in treatment, substance use, clinical improvement, general social adjustment, urinalysis.	77	24.7% female. Age range 20-46 yrs.	No differences in survival-in-treatment rates (0.44 vs. 0.58). HA+BZD patients had better outcome results (lower illness severity, better social adjustment) vs. HA patients without BZD severe addiction. HA+BZD patients needed higher methadone dosage in stabilisation phase.
Mijatović, 2013 [26]	Serbia	Assess the safety of low doses of methadone combined with BZD.	Pilot study	QTc interval; self-reported side effects.	20	25.0% female. Mean (SD) age 32.21 (5.63) yrs.	Statistically significant increase in length of QTc interval. Dose-dependent correlation with BZD but not methadone. BZD type not specified.
Mijatović, 2017 [27]	Serbia	Evaluate role of diazepam in methadone -associated QTc prolongation in patients with opioid use disorder during methadone maintenance treatment.	Observational	QTc interval; serum concentration of methadone, diazepam, and electrolytes.	30	20.0% female. Mean (SD) age 32 (5) yrs.	Statistically significant increase in length of QTc interval at 1 month and 6 months. Statistically significant correlation between concentration of methadone and diazepam.
Park, 2020 [28]	USA	To assess whether (1) BZD prescribing during buprenorphine treatment is associated with increased risks of fatal and non-fatal opioid overdose and all-cause mortality; (2) BZD prescribing during buprenorphine treatment is associated with reduced risk of buprenorphine discontinuation (13 prescribable BZD included).	Retrospective cohort study	Primary outcome: fatal opioid overdose. Secondary outcomes: non-fatal opioid overdose, all-cause mortality, buprenorphine treatment discontinuation.	63,345	37.5% female. Mean (SD) age 38 (11.0) yrs.	Of the 67,088 person-years of observation on buprenorphine, 57,825 person-years (86%) represented exposure to buprenorphine alone, and 9263 person-years (14%) represented exposure to buprenorphine and benzodiazepine. 183 people died of an opioid overdose, there were 693 non-fatal opioid overdoses, and 369 people died from any cause. 31% of fatal opioid overdoses occurred during times when people received BZD during buprenorphine treatment. Compared to periods during which people received buprenorphine alone, periods of concurrent BZD and buprenorphine receipt were associated with an increased risk of opioid-related overdose death [HR 2.92, 95% CI 2.10–4.06]. BZD treatment during buprenorphine treatment was also associated with an increased risk of non-fatal opioid overdose (HR 2.05, 95% CI 1.68–2.50) and all-cause mortality (HR 1.90, 95% CI 1.48–2.44). BZD treatment during buprenorphine treatment was associated with a decreased risk of buprenorphine treatment discontinuation (HR 0.87, 95% CI 0.85–0.89). Opioid overdose during buprenorphine treatment was relatively rare: there were only 183 fatal opioid overdoses during the 4-year study period, representing fewer than 4% of the 4754 estimated total opioid overdoses in the state of Massachusetts.
Rapeli, 2009 [29]	Finland	To examine longitudinal (<9 months) changes in memory function among OST patients with BZD use. BZD type not specified.	Longitudinal	Memory function.	43	39.3% female. Mean (SD) age 28.5 (6.55) yrs.	Significantly worse working memory at T1 and T2, and worse immediate verbal memory at T1 in OST patients vs. normal comparison. Both patient groups reported sig. more subjective memory problems vs. comparison at T1 and T2. OST patients with more memory complaints recalled fewer items at T2 from verbal list learned at T1 than patients with fewer memory complaints. No sig. group by time interactions were found.
Schuman-Olivier, 2013 [30]	USA	To evaluate relationship between BZD misuse history, BZD prescription, and both clinical and safety outcomes during buprenorphine treatment.	Secondary data analysis Retrospective chart review with, quasi-experimental design	Clinical outcomes incl. 12-month treatment retention and urine toxicology for illicit opioids); safety outcomes incl. emergency department visits.	328	40.2% female. Mean (SD) age 36.6 (10.65) yrs. 93.0% White ethnicity.	The 12-month treatment retention rate for the sample (N = 328) was 40% (no association between history of BZD misuse OR prescribing AND retention). Poisson regressions of ED visits during buprenorphine treatment revealed more ED visits among those with a BZD prescription versus those without (p < 0.001). BZD type not specified. Odds of an accidental injury-related ED visit during treatment were greater among those with a BZD prescription (OR: 3.7; p < 0.01), with an enhanced effect among females (OR: 4.7; p < 0.01) Overdose was not associated with BZD misuse history or prescription.
Sharma, 2020 [31]	Canada	Estimate the effect of concurrent BZD use on the risk of hospitalisation/emergency department (ED) visits and deaths among people who use opioids.	Secondary analysis Population-based case cross-over study during 2016–2018	Risk of incident all-cause hospitalisation/ED visits; all-cause mortality.	1.06M	55.0% female. Mean (SD) age 48.7 (18.1) yrs.	Concurrent BZD use (type not specified) occurred in 17% of opioid users (179,805/1,056,773). Overall, concurrent use was associated with higher risk of hospitalisation/ED visit (OR 1.13; p < 0.001) and all-cause death (OR 1.90; p < 0.001). The estimated risk of hospitalisation/ED visit was highest in those >65 (OR 1.5; p < 0.001), using multiple health providers (OR 1.67; p < 0.001) and >365 days of opioid use (OR 1.76; p < 0.001). Events due to opioid toxicity were also associated with concurrent use (OR 1.8; p < 0.001). Opioid dose–response effects among concurrent patients who died were also noted (OR 3.13; p < 0.001)
Weizman, 2003 [32]	Israel	To compare two pharmacological modalities, clonazepam detoxification and clonazepam maintenance, for treating long-term BZD dependence in methadone maintenance patients	Open—unblinded clinical study.	Reduction or cessation of BZD use	66	No data	In the clonazepam detoxification group, 9/33 (27.3%) were benzodiazepine-free after 2 months. In the clonazepam maintenance group, 26/33 (78.8%) refrained from using additional BZD over the maintenance dose after 2 months. The same success rate remained over the entire year. Survival analysis showed clonazepam maintenance to be more successful than the clonazepam detoxification. Axis I psychiatric comorbidity was found to be positively related to treatment success in the clonazepam maintenance group, while axis II antisocial personality disorder was found to be negatively related to treatment success in that group. It had no impact on the clonazepam detoxification group. Maintenance strategy with clonazepam is a useful treatment modality for benzodiazepine-dependent methadone maintenance patients. Psychiatric comorbidity may have an important role in choosing the adequate treatment modality.
Xu, 2021 [33]	USA	Evaluate association of BZD (converted to diazepam equivalent dose) and z-drug use with non-fatal drug-related poisonings among buprenorphine-maintained patients.	Observational, case-crossover design	Non-fatal drug-related poisoning.	23,036	49.2% female. Mean (SD) age 30.05 (12.15) yrs.	Buprenorphine treatment days associated with 37% reduction in risk of drug-related poisoning events vs. non-treatment days. BZD and z-drug treatment days associated with 88% increase in risk of poisoning events. 78% and 122% increase in poisonings associated with low- and high-dose BZD and z-drug treatment, respectively. High-dose BZD/z-drug treatment associated with increased poisonings in combination with buprenorphine co-treatment; but lower than risk associated with BZD/z-drug treatment in absence of buprenorphine.

BZD: benzodiazepine/HR: hazard ratio; OR: odds ratio; CI: confidence interval; SD: standard deviation.