Skip to main content

Table 7 Studies reporting measures of illicit substance use

From: The impact of relaxing restrictions on take-home doses during the COVID-19 pandemic on program effectiveness and client experiences in opioid agonist treatment: a mixed methods systematic review

Study

Measure

Control Group

Intervention Group

Statistical Test or Model

p-value

Estimate of Effect

Direction of Effect

Overall Effect Direction

(S4) Bart et al., 2022 [72]

Urine test positive for opiates without confirmed prescription

14%a

22%b

NR

 < 0.001

NR

Favours control

Negative

Urine test positive for amphetamines without confirmed prescription

10%a

16%b

NR

 < 0.001

NR

Favours control

Urine test positive for barbiturates without confirmed prescription

0.2%a

0.3%b

NR

p ≥ 0.001

NR

Favours control

Urine test positive for benzodiazepines without confirmed prescription

6.3%a

11%b

NR

 < 0.001

NR

Favours control

Urine test positive for cocaine without confirmed prescription

11%a

12%b

NR

p ≥ 0.001

NR

Favours control

Urine test positive for oxycodone without confirmed prescription

2.6%a

3.2%b

NR

p ≥ 0.001

NR

Favours control

Urine test positive for opioids (opiates or oxycodone) without confirmed prescription

NRa

NRb

Generalized linear mixed model

NR

OR: 2.34 (95% CI 1.78–3.07)

Favours control

Urine test positive for non-opioids without confirmed prescription

NRa

NRb

Generalized linear mixed model

NR

OR: 2.48 (95% CI 1.89–3.25)

Favours control

Proportion of drug tests positive for opioids among clients with 1–2 take-home doses/week

0.435c

0.202d

Generalized linear mixed model

NR

NR

Favours intervention

Proportion drug tests positive for opioids among clients with 3–5 take-home doses/week

0.187c

0.226d

Generalized linear mixed model

NR

NR

Favours control

Proportion of drug tests positive for opioids among clients with 6 take-home doses/week

0.060c

0.121d

Generalized linear mixed model

NR

NR

Favours control

Proportion of drug tests positive for opioids among clients with > 6 take-home doses/week

0.027c

0.036d

Generalized linear mixed model

NR

NR

Favours control

Proportion of drug tests positive for non-opioids among clients with 1–2 take-home doses/week

0.587c

0.398d

Generalized linear mixed model

NR

NR

Favours intervention

Proportion drug tests positive for non-opioids among clients with 3–5 take-home doses/week

0.187c

0.377d

Generalized linear mixed model

NR

NR

Favours control

Proportion of drug tests positive for non-opioids among clients with 6 take-home doses/week

0.119c

0.161d

Generalized linear mixed model

NR

NR

Favours control

Proportion of drug tests positive for non-opioids among clients with > 6 take-home doses/week

0.049c

0.040d

Generalized linear mixed model

NR

NR

Favours intervention

(S6) Corace et al., 2022 [74]

OAT clients reporting increase in opioid use "since COVID-19 (March 2020)"

46%e

28%f

NR

NR

NR

Favours intervention

Mixed

OAT clients reporting decrease in opioid use "since COVID-19 (March 2020)"

21%e

14%f

NR

NR

NR

Favours control

(S8) Ezie et al., 2022 [76]

Urine drug screens positive for opiates

39%a

36%b

Multiple logistic regression

 > 0.05

Adjustedg OR: 0.82 (0.34–1.98)

Favours intervention

Positive

Urine drug screens positive for any non-prescribed substance other than cannabis

45%a

40%b

Multiple logistic regression

 > 0.05

Adjustedg OR: 0.61 (0.25–1.48)

Favours intervention

(S15) Hoffman et al., 2022 [83]

Random monthly urine drug tests positive for opioids among clients in treatment for < 90 days

38% (SD 0.43)a

33% (SD 0.42)b

Wilcoxon rank sum test, Pearson's Chi-squared test

0.6

NR

Favours intervention

Positive

Random monthly urine drug tests positive for opioids among clients in treatment for 90–180 days

19% (SD 0.34)a

33% (SD 0.43)b

Wilcoxon rank sum test, Pearson's Chi-squared test

0.041

NR

Favours control

Random monthly urine drug tests positive for opioids among clients in treatment for > 180 days

23% (SD 0.33)a

20% (SD 0.32)b

Wilcoxon rank sum test, Pearson's Chi-squared test

0.12

NR

Favours intervention

Expected change in random monthly urine drug test positivity per percentage point in take-home dosing above expectedh

NR

NR

Linear regression

0.005

Slope: -0.12 (95% CI -0.21, -0.04)

Favours intervention

(S22) Lintzeris et al., 2022 [90]

Any self-reported cannabis use

33%a

38%n

McNemar test

0.028

χ2: 4.817

Favours control

Positive

Any self-reported benzodiazepine use

28%a

22%n

McNemar test

0.014

χ2: 6.017

Favours intervention

Any self-reported stimulant use

20%a

16%n

McNemar test

0.120

NR

Favours intervention

Any self-reported opioid use

30%a

24%n

McNemar test

0.033

χ2: 4.563

Favours intervention

Any self-reported injection drug use

29%a

22%n

McNemar test

0.077

NR

Favours intervention

Average days used among clients self-reporting cannabis use

Mean: 18.1 (SD 10.8)

Median: 21a

Mean 18.0 (SD 11.0), Median 26b

Wilcoxon signed-rank test

0.020

Z: -2.331

Favours control

Average days used among clients self-reporting benzodiazepine use

Mean: 14.6 (SD 11.7) Median: 12a

Mean: 16.9 (SD 11.4) Median: 20b

Wilcoxon signed-rank test

NR

NR

Favours control

Average days used among clients self-reporting stimulant use

Mean: 6.5 (SD 8.2) Median: 3a

Mean: 5.9 (SD 7.4) Median: 3b

Wilcoxon signed-rank test

NR

NR

Favours intervention

Average days used among clients self-reporting opioid use

Mean: 12.2 (SD 10.7) Median: 8a

Mean: 7.9 (SD 9.1) Median: 4b

Wilcoxon signed-rank test

0.001

Z: -3.445

Favours intervention

Average days used among clients self-reporting injection drug use

Mean: 10.7 (SD 10.5) Median: 5a

Mean: 8.1 (SD 8.9) Median: 4b

Wilcoxon signed-rank test

0.010

Z: 2.577

Favours intervention

Percentage of clients with “statistically reliable” and “clinically relevant” increase in substance use (composite measure)i

43%j

(a) 40%k

(b) 17%l

Logistic regression

(a) p ≥ 0.05*

(b) p < 0.05*

Adjusted OR:

(a) 0.854 (0.39–1.87)

(b) 0.273 (0.10–0.77)

Favours intervention

(S25) Morin et al., 2021 [93]

Routine urine drug screens positive for fentanyl

Jan: 14%

Feb: 13%

Mar: 14%a

Apr: 12%

May: 21%

Jun: 26%

Jul: 29%

Aug: 29%

Sep: 25%b

Fractional logistic regression

NR

OR:

(a) Apr. vs. Jan: 0.9 (95% CI: 0.8–0.9)

(b) May vs. Jan.: 1.7 (95% CI: 0.5–1.89)

(c) Jun. vs. Jan.: NR

(d) Jul. vs. Jan: NR

(e) Aug. vs. Jan: 2.6 (95% CI: 2.3–2.9)

(f) Sep vs. Jan: 2.2 (95% CI: 1.9–2.6)

Favours controlm

Negative

Routine urine drug screens positive for cocaine

Jan: 24%

Feb: 24%

Mar: 24%a

Apr: 23%

May: 29%

Jun: 28%

Jul: 28%

Aug: 26%

Sep: 25%b

NR

NR

NR

Favours controln

Routine urine drug screens positive for methamphetamine

Jan: 18%

Feb: 19%

Mar: 20%a

Apr: 17%

May: 23%

Jun: 23%

Jul: 18%

Aug: 17%

Sep: 19%b

NR

NR

NR

Favours controln

Routine urine drug screens positive for morphine

Jan: 13%

Feb: 13%

Mar: 13%a

Apr: 12%

May: 15%

Jun: 15%

Jul: 15%

Aug: 15%

Sep: 15%b

NR

NR

NR

Favours controln

Routine urine drug screens positive for oxycodone

Jan: 6%

Feb: 6%

Mar: 6%a

Apr: 6%

May: 7%

Jun: 7%

Jul: 6%

Aug: 6%

Sep: 6%b

NR

NR

NR

Favours controln

(S30) Rosic et al., 2022 [98]

Percentage of opioid-positive urine drug screens

Mean: 7.5% (SD 17.2)a

Mean: 18.1% (SD 26.5)b

Paired t-test

p < 0.001

Risk difference: 10.56% (95% CI: 8.17–12.95)

Favours control

Mixed

Percentage of clients with any opioid-positive urine drug screens

73.5%a

46.3%b

NR

NR

NR

Favours intervention

(S37) Vicknasingam et al., 2021 [107]

Percentage of clients with urine toxicology tests positive for any illicit substance

Dec.: 23%

Jan.: 23%

Feb.: 18%a

Jun.: 24%

Jul.: 19%b

NR

NR

NR

Favours controln

Negative

  1. Where adjusted and unadjusted effect estimates were reported, we present adjusted estimates. In no case did this change the estimated direction of effect
  2. Acronyms: NR not reported, OR odds ratio. Where bivariate and multivariate analyses were reported, we present the results of the multivariate analysis
  3. *Not reported in the original study; inferred or calculated by authors
  4. aControl group: OAT clients pre-pandemic
  5. bIntervention group: OAT clients post-pandemic
  6. cControl group: 2019 values from a fitted model that removed the main effect of year to “[capture] the effect of change in take-out schedule” (Bart et al., 2022, p. 3)
  7. dIntervention group: 2020 values from a fitted model that removed the main effect of year
  8. eControl group: All OAT clients
  9. fIntervention group: OAT clients with additional take-home doses during pandemic
  10. gAdjusted for years in treatment, age, substance use disorder diagnosis, psychiatric disorder diagnosis, and % reduction in visit frequency
  11. hAnalysis limited to clients with three months of pre-COVID-19 data and one month of post-COVID-19 data
  12. iDefined as an increase of 4 or more days in the previous 28 days
  13. jControl group: OAT clients with no take-home doses at follow up
  14. kIntervention group (a): OAT clients with 1–5 take-home doses/week at follow up
  15. lIntervention group (b): OAT clients with 6 + take-home doses/week at follow up
  16. mBased on proportion of comparisons favouring control
  17. nBased on mean control group value versus mean intervention group value
Back to article page

Substance Abuse Treatment, Prevention, and Policy

ISSN: 1747-597X