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

Background The COVID-19 pandemic led to an unprecedented relaxation of restrictions on take-home doses in opioid agonist treatment (OAT). We conducted a mixed methods systematic review to explore the impact of these changes on program effectiveness and client experiences in OAT. Methods The protocol for this review was registered in PROSPERO (CRD42022352310). From Aug.–Nov. 2022, we searched Medline, Embase, CINAHL, PsycInfo, Web of Science, Cochrane Register of Controlled Trials, and the grey literature. We included studies reporting quantitative measures of retention in treatment, illicit substance use, overdose, client health, quality of life, or treatment satisfaction or using qualitative methods to examine client experiences with take-home doses during the pandemic. We critically appraised studies using the Mixed Methods Appraisal Tool. We synthesized quantitative data using vote-counting by direction of effect and presented the results in harvest plots. Qualitative data were analyzed using thematic synthesis. We used a convergent segregated approach to integrate quantitative and qualitative findings. Results Forty studies were included. Most were from North America (23/40) or the United Kingdom (9/40). The quantitative synthesis was limited by potential for confounding, but suggested an association between take-home doses and increased retention in treatment. There was no evidence of an association between take-home doses and illicit substance use or overdose. Qualitative findings indicated that take-home doses reduced clients’ exposure to unregulated substances and stigma and minimized work/treatment conflicts. Though some clients reported challenges with managing their medication, the dominant narrative was one of appreciation, reduced anxiety, and a renewed sense of agency and identity. The integrated analysis suggested reduced treatment burden as an explanation for improved retention and revealed variation in individual relationships between take-home doses and illicit substance use. We identified a critical gap in quantitative measures of patient-important outcomes. Conclusion The relaxation of restrictions on take-home doses was associated with improved client experience and retention in OAT. We found no evidence of an association with illicit substance use or overdose, despite the expansion of take-home doses to previously ineligible groups. Including patient-important outcome measures in policy, program development, and treatment planning is essential to ensuring that decisions around take-home doses accurately reflect their value to clients. Supplementary Information The online version contains supplementary material available at 10.1186/s13011-023-00564-9.


Introduction
Opioid use disorder affects an estimated 21.4 million people worldwide [1].It is associated with significant morbidity and mortality, attributable in part to the stigmatization, social marginalization, and criminalization of people who access the unregulated drug supply [2,3].Regionally, opioid use disorder is most prevalent in high-income North America [4].In 2022, a total of 83,827 deaths in the United States and 7,328 deaths in Canada were attributed to opioid toxicity [5,6].This is a substantial increase over 2016, when 43,149 deaths were reported in the United States and 2,831 in Canada [5,6].The severity of the overdose crisis in this region of the world is the result of historical overprescribing, social factors, and an unregulated drug supply that is heavily contaminated with fentanyl, benzodiazepines, and other adulterants [7][8][9].
Opioid agonist treatment (OAT) using methadone or buprenorphine is an effective and well-established approach to reducing the harms associated with opioid use disorder [10][11][12][13].Both methadone and buprenorphine suppress use of unregulated opioids when prescribed at adequate doses [11,14] and are associated with substantial reductions in rates of fatal and non-fatal overdose [13,15].Despite these benefits, retention in OAT is low; it ranges from 19 to 86% at six months, with a median retention rate of 58% [16].Mortality rates rise steeply after treatment cessation [13].
Burdensome treatment conditions, particularly for clients on methadone, may contribute to low retention in OAT [17].These conditions commonly include supervised dosing, in which OAT clients must travel to their clinic or pharmacy each day so that their medication can be ingested under the observation of a health care provider [18].Take-home doses, which can be carried out of the clinic and stored safely elsewhere, may be granted to clients who meet specific criteria.
In the United States, pre-pandemic guidelines for methadone programs required clients to meet eight criteria reflecting 'stability' and to remain in treatment for a minimum of six months before becoming eligible to receive more than two take-home doses per week [19].Factors affecting eligibility for take-home OAT in other jurisdictions include time in treatment, abstinence from illicit substance use, housing stability, distance from the treatment facility, and provider discretion [18,20].
Restrictions on take-home doses are driven by concerns over the potential for diversion, injection, and overdose [21].Methadone is approached with particular caution; as a full agonist with a long half-life, it has the potential to cause serious respiratory depression if taken in excess or in conjunction with alcohol, unregulated opioids, or other sedatives [21].For this reason, careful titration is necessary to initiate methadone safely.However, systematic reviews of supervised versus unsupervised dosing have found insufficient evidence to determine whether restrictions on take-home doses are effective in reducing diversion [22,23].Recent research has drawn attention to the role of unmet treatment need in the market for diverted medication [24][25][26] and highlighted the potential for benefits as well as harms [27,28].
Though some OAT clients appreciate the structure of daily supervised dosing [29,30], inflexible restrictions on take-home doses have repeatedly been identified as a source of dissatisfaction with treatment [31].In addition to "[obstructing] the basic day-to-day functioning of life" [32] (p.S118), supervised dosing has been described as humiliating, degrading, and stigmatizing [29,33,34].Commentators have argued that supervised dosing is part of a treatment paradigm that reinforces institutional stigma and power imbalances, serving as a form of social control as well as a medical intervention [35][36][37][38].
The COVID-19 pandemic led to the relaxation of restrictions on take-home doses on an unprecedented scale.The risks of viral infection to clients and providers in medical settings, as well as the dangers of treatment discontinuation for clients who might stop OAT to avoid exposure to COVID-19, were deemed to outweigh the potential harms of take-home doses.Regulations and guidelines to encourage use of take-home doses during the pandemic were developed in Canada [39], the United States [40,41], Australia [42], England [43], Spain [44], Italy [45], and India [46].Other changes to OAT during COVID-19 included the suspension of urine testing or a reduction in testing frequency, increased emphasis on naloxone distribution, medication delivery for clients in isolation or quarantine, and the use of virtual care in place of in-person visits [39, 41-43, 45, 46].Though implementation of the new flexibilities around takehome doses varied [47], their introduction created an unparalleled opportunity to assess the impact of relaxing restrictions on take-home doses in OAT.
Previous reviews of changes to take-home guidance during COVID-19 have focused on providers' experiences [48] and changes within the United States [49].To our knowledge, this is the first systematic review of international scope to focus on how relaxing restrictions on take-home doses during the COVID-19 pandemic affected program effectiveness and client experiences in OAT.Results from this study can support clinicians, policymakers, and stakeholders in making informed decisions around the implementation and expansion of takehome doses in OAT.

Design
We conducted a mixed methods systematic review to address the following questions: Q1 (quantitative): What was the impact of relaxing restrictions on take-home doses during the COVID-19 pandemic on program effectiveness in OAT, as defined by (1) retention; (2) illicit substance use; (3) fatal and non-fatal overdose; (4) client health (e.g., measures of physical, mental, or emotional health); (5) quality of life; and (6) treatment satisfaction?Q2 (qualitative): What was the impact of relaxing restrictions on take-home doses during the COVID-19 pandemic on clients' experiences with OAT? Q3.What are the integrated findings of Q1 and Q2, and what are their implications for OAT?
Mixed methods approaches have the potential to generate a more complete and nuanced understanding of a phenomenon than quantitative or qualitative evidence alone.Qualitative evidence can suggest explanations for quantitative findings, help policymakers predict the impact of an intervention in a specific context, and illuminate aspects of human experience that are not captured by quantitative research [50].We used a convergent segregated approach in which the quantitative synthesis (Q1) and qualitative synthesis (Q2) are conducted separately before being integrated through 'configured analysis' (Q3) [51].Reporting of the methods was guided by the PRISMA and PRISMA-S statements for reporting systematic reviews and the Synthesis Without Metaanalysis (SWiM) reporting guideline [52][53][54] (Additional file 1).The protocol for this review was registered in PROSPERO (CRD42022352310; https:// www.crd.york.ac.uk/ prosp ero/).

Search strategy
We used the PICO (Population, Intervention, Comparator, Outcomes) and PICo (Population, phenomenon of Interest, Context) frameworks to structure our search strategy and define our inclusion criteria (Table 1).
The search strategy was developed by a member of the research team with expertise in systematic searching (AA) and reviewed by a professional research librarian.Substantive elements of the search strategy for this review were used in a previously published review [48].We restricted all searches to articles published after January 1, 2020 because the review focuses on actions taken in response to the COVID-19 pandemic.

Table 1 PICO and PICo criteria for review questions Q1 and Q2
Abbreviations: OAT opioid agonist treatment a As specified in the review protocol, we included studies in which relaxed restrictions on take-home doses formed part of a broader intervention or context Q1: What was the impact of relaxing restrictions on take-home doses during the COVID-19 pandemic on program effectiveness in OAT, as indicated by (1)  We searched six electronic databases and registers on Aug. 26, 2022 to retrieve peer-reviewed literature: Medline (Ovid), Embase (Ovid), CINAHL Complete (EBSCOhost), PsycInfo (EBSCOhost), Web of Science Core Collection (Web of Science), and Cochrane Central Register of Controlled Trials (Ovid).See Additional file 2 for a sample search strategy.We conducted a grey literature search of selected websites and databases from Oct. 27-Nov.7, 2022.We conducted forward and backward citation chaining from Dec. 1-2, 2022.We updated the searches through an additional round of forward citation chaining conducted on Mar.31,2023.Full search strategies can be found in the OSF data repository [55].

Screening, data extraction, and critical appraisal
We imported all searches into Covidence, an online platform for supporting systematic reviews [56].Screening, data extraction and critical appraisal were completed in Covidence by two reviewers working independently and blinded to each other's assessments (AA, SB, RF, TM).See Table 2 for eligibility criteria used in screening.Disagreements were resolved through discussion or by a third reviewer (JL, SB).We used a standardized, pre-piloted form to extract information on study characteristics and findings, including geographical region, study aim, study design, and sample characteristics.We used the Mixed Methods Appraisal Tool (MMAT) version 2018 to appraise study quality and validity [57].The MMAT is designed specifically for mixed methods systematic reviews.We used the results of the appraisal to assess the strengths and weaknesses of the evidence base and conducted a sensitivity analysis excluding lowquality studies, which we defined as studies meeting fewer than three of five criteria on the MMAT.

Quantitative synthesis
For the quantitative synthesis, we grouped study findings by outcome to improve comparability.We did not conduct meta-analysis or summarize effect estimates because the diversity of outcome measures precluded calculation of a common effect estimate.Nor was it possible to summarize p-values with the data available.Instead, we synthesized data using vote counting based on direction of effect to answer the question "Is there any evidence of an effect?"[58,59].This method is an acceptable alternative to meta-analysis when it is not possible to calculate a standardized estimate of effect, as is often the case in reviews of complex interventions [58][59][60].For each outcome, we compared the number of studies showing a beneficial effect with the number showing a harmful effect.As per guidance, we did not take statistical significance or magnitude of effect into account [59].

Inclusion Criteria
For all studies: • Includes findings on the impact of relaxed restrictions on take-home doses of opioid agonist medication for opioid use disorder, either alone or in conjunction with other interventions/exposures, during the Covid-19 pandemic on program effectiveness in opioid agonist treatment • Written in English, French, Spanish, Portuguese, or Italian

For quantitative component:
• A randomized or non-randomized study reporting quantitative data OR a mixed methods study where the quantitative component can be cleanly extracted • Assesses one or more of the following client outcomes: (1) Retention in treatment, using any quantitative measure; (2) illicit substance use, using any quantitative measure; (3) fatal and non-fatal overdose, using any quantitative measure; (4) client health, using any quantitative measure; (5) client quality of life, using any quantitative measure; (6) client satisfaction with treatment, using any quantitative measure

For qualitative component:
• A qualitative study using any qualitative approach (e.g., grounded theory, critical theory, ethnography) OR a mixed methods study where the qualitative component can be cleanly extracted • Includes findings on OAT clients' experiences with relaxed restrictions on take-home doses of OAT during the Covid-19 pandemic

Exclusion Criteria
For all studies: • OAT clients are a subgroup of the study population, but findings specific to this group cannot be extracted; • Take-home doses intended to be supervised remotely or in person (e.g., witnessed daily delivery; take-homes witnessed through videoconferencing systems) • Commentaries, editorials, or letters to the editor, unless original empirical research is presented • Conference abstracts, posters, or slide decks, unless meeting three predefined conditions designed to limit retrieval to relevant studies for which sufficient information can be obtained • The study is a preprint that has become available in peer-reviewed form

For qualitative component:
• The study uses quantitative methods (e.g., questionnaires, fixed-choice surveys) to collect qualitative data When a study used more than one measure for the same outcome, we used Boon & Thomson's revised method [58] to determine the overall direction of effect supported by the study.If the direction of effect was the same (e.g., all beneficial or all harmful) for ≥ 70% of measures, we considered this the overall direction of effect.We recorded the direction as mixed if less than 70% of measures reported a consistent effect direction.We described the results of the synthesis using harvest plots displaying direction of effect, study quality, and sample size [61][62][63].
We planned to investigate heterogeneity through subgroup analyses based on treatment type (buprenorphine, which had considerably fewer restrictions on take-home doses before the pandemic, versus methadone) and on race and ethnicity.However, formal statistical investigation was not feasible because of insufficient data.Where possible, we explored the effects of treatment type through informal methods; more specifically, through visual inspection of harvest plots in which studies were shaded according to treatment type (methadone vs. buprenorphine).

Qualitative synthesis
We synthesised qualitative findings using thematic synthesis, which consists of (1) coding studies line-by-line; (2) grouping codes into descriptive themes; and (3) integrating the descriptive themes into analytical themes that address the review question more directly [64].Thematic synthesis preserves a clear audit trail from data to analytical themes, making it particularly suitable for systematic reviews [65].
Two members of the research team (AA, SB) coded the same four studies line-by-line in NVivo 1.7 [66].AA and SB compared and reconciled their coding to create a set of codes and descriptive themes that were used to code/re-code all studies (AA, SB).After coding was completed, AA and SB discussed conceptual links between the descriptive themes and generated analytical themes.These themes were then reviewed with a third member of the research team (EOJ).See Fig. 1 for an illustration of theme development.

Certainty of evidence
There is no consensus around whether appraising the certainty of the evidence is appropriate in mixed methods reviews, with some organizations supporting this step [67] and others advising against [51].Methodologists have raised concerns about the use of GRADE and similar methods in mixed methods reviews because of the complexities and uncertainties around incorporating these assessments into the integrated findings of the Fig. 1 Example of the development of an analytical theme.For visual simplicity, only descriptive themes and codes contributing to Theme #1 are shown review [51,68].In light of these concerns, we did not formally appraise the certainty of the evidence supporting the qualitative or quantitative findings.

Integrated analysis
To develop the integrated analysis, we juxtaposed the qualitative and quantitative syntheses and considered how they might complement, explain, or contextualize each other [51].After drafting the analysis, we discussed our preliminary findings with seven community members with lived experience of OAT to help us assess the credibility of our findings and inform further interpretation.

Results
After excluding duplicates, we retrieved 2,888 records from databases and registers and 20 records from citation chaining and the grey literature search.Of these, 42 records (representing 40 studies) met our eligibility criteria and were included in the review  (hereafter referred to as S1-S40; see Table 3) (Fig. 2).

Study characteristics
Most studies were from the United States (16/40), the United Kingdom (9/40), or Canada (7/40).Twenty-four studies included participants on a variety of OAT medications.Fourteen focused exclusively on methadone clients and two were limited to buprenorphine clients.For additional details on study design and participant characteristics, see Tables 3 and 4.
Eighteen studies contributed data to the quantitative synthesis.As specified in our review protocol, we included studies in which the relaxation of restrictions on take-home doses formed part of a broader intervention or exposure.Other pandemic-related changes to OAT described in the quantitative studies included increased use of telehealth and virtual care (S2, S6, S7, S11, S13, S22, S30, S31), reduced in-person appointments (S6, S7, S11, S13, S15-17, S22), cessation or reduced frequency of urine testing (S2, S6, S11, S17, S22, S37), home delivery of medication for clients who were self-isolating and/or at high risk (S7, S22, S30), rapid or remote protocols for OAT induction (S2, S30, S31), and increased naloxone provision (S7, S22).Of the 18 studies, nine were intended to assess only the impact of changes to take-home policies.Five of these studies (S3, S4, S13, S15, S22) used methods to control for the impact of co-exposures or other factors associated with the receipt of take-home doses (e.g., regression modelling) in their analysis.Six studies defined their intervention of interest as pandemic-related changes to OAT, including, but not limited to, increased take-home doses.Two studies defined their exposure/intervention as the pandemic together with associated changes to OAT.
Twenty-five studies contributed to the qualitative synthesis.Three focused exclusively on OAT clients' experiences with take-home during the COVID-19 pandemic.Many were designed to explore participants' experiences with any and all pandemic-related changes to OAT (15/25).A smaller number explored how people who use drugs experienced life during the pandemic (7/25).Though all studies met our inclusion criteria, some contributed little data to the synthesis.

Quantitative synthesis
Visual inspection of harvest plots (see Fig. 3) suggested an association between take-home doses and improved retention, but showed no clear evidence of an association with overdose or illicit substance use.The small number of studies reporting client health or quality of life precluded meaningful synthesis.We did not identify any studies reporting treatment satisfaction.Brief narrative summaries are provided below.

Retention
Seven studies reported measures of retention, including one finding a negative direction of effect (S15), one with mixed direction of effect (S11), and five supporting a positive direction of effect (S7, S9, S13, S26, S31).See Table 5.Two were high-quality (S13, S15), two were moderate-quality (S7, S11), and three were low-quality (S9, S26, S31).Our main concerns about the quality of studies contributing to this outcome were failure to account for confounding, unplanned co-interventions, and generalizability (Table 6).
Negative direction A before-and-after study (S15) found that treatment discontinuation increased following the relaxation of restrictions on take-home doses, regardless of time in treatment.However, logistic regression showed that the odds of treatment discontinuation decreased with each additional take-home dose.

Mixed direction
The overall direction of effect was mixed in a study using statistical modelling to test for changes in OAT discontinuation after pandemic-related treatment changes (S11).Though there was an immediate decrease in treatment discontinuation for all clients, tests for gradual changes showed no change among stable clients and a negative trend for non-stable clients.
Positive direction Five studies reported a positive direction of effect (S7, S9, S13, S26, S31).A cohort study of buprenorphine clients (S7) found that clients referred to treatment during the pandemic, when prescription durations increased, had a higher rate of retention at 90 days than clients referred prior to the pandemic.Another cohort study (S13) assessed the risk of OAT discontinuation in a sample stratified by treatment type and number of take-home doses at baseline.In all four subgroups, clients who received additional take-home doses during COVID-19 had a lower risk of treatment discontinuation.Two before-and-after studies reported increased retention following the relaxation of restrictions on take-home doses (S9, S26), and a time series study using data on buprenorphine prescriptions in the United States (S31) reported a reduction in treatment disruptions of 28 days or more during the pandemic.
Negative direction One time series study (S25) and two before-and-after studies (S4, S37) found an increase in the percentage of positive urine tests among OAT clients following pandemicrelated treatment changes.One study (S4) used statistical modelling to examine whether urine test positivity was associated with number of takehome doses, but found no clear association.
Mixed direction A cross-sectional study (S6) reported that clients receiving additional take-home doses during the pandemic were less likely to report increased or decreased opioid use since COVID-19.In a beforeand-after study (S30), the total percentage of positive urine tests among OAT clients increased following the COVID-19 pandemic.However, the percentage of clients testing positive decreased.
Positive direction Three before-and-after studies (S8, S15, S22) reported a decrease in the percentage of positive urine tests (S8, S15) or self-reported substance use (S22) following pandemic-related treatment changes.In one study (S15), a linear regression analysis limited to clients in treatment for at least three months before the pandemic found that the probability of a positive urine test decreased as takehome doses increased.
Negative direction A cross-sectional study (S6) found that self-reported opioid overdoses were higher for OAT clients who received extra take-home doses during the pandemic than for those who did not.
Mixed direction A modelling study (S2) found that actual methadone-related deaths did not far exceed projected deaths among people prescribed methadone during England's first COVID-19 lockdown, when most clients received two-week take-home doses.The count of buprenorphine-related deaths among people prescribed buprenorphine was unchanged compared with previous years.A retrospective, propensity-weighted cohort study found that increased take-home doses were associated with a lower risk of overdose among methadone clients and a higher risk among buprenorphine/naloxone clients (S13).In a commentary with data on overdoses reported to health care providers at opioid treatment programs in New York (S17), there was a higher count of non-fatal overdoses and a lower count of fatal overdoses after changes to take-home guidelines.A preprint with data on fatal overdoses among methadone clients (S26) reported that receiving additional take-home doses during the pandemic was associated with a higher rate of fatal overdose for clients without take-home doses at baseline.However, for clients who had take-home doses at baseline, those who received additional take-home doses during the pandemic had a lower rate of fatal overdose than those who did not.
Positive direction One before-and-after study (S3) reported fewer overdose-related emergency department visits among methadone clients following changes to take-home guidelines.Statistical modelling showed that the odds of overdose decreased with each one-dose increase in take-home doses after controlling for age, gender, education, and employment.Another before-and-after study (S8) did not specify an outcome measure, but reported reduced overdoses among methadone clients following the relaxation of take-home guidelines.

Client health
Three studies reported on client health, which included measures of physical, emotional, and mental well-being and measures of infection and disability related to substance use.See Table 11.One study found a negative direction of effect (S8) and two reported a mixed direction of effect (S6, S2).One study was moderatequality (S22) and two were low-quality (S6, S8).Sources of downgrading included generalizability, appropriateness of outcome measurements, failure to account for confounding, and unplanned co-interventions (see Table 12).

Favours intervention
Where adjusted and unadjusted effect estimates were reported, we present adjusted estimates.Where weighted and unweighted effect estimates were reported, we present weighted estimates.In no case did this change the estimated direction of effect Acronyms: HR hazard ratio, NR not reported, OR odds ratio, RR relative risk Negative direction A study of methadone clients (S8) found that the incidence of infections associated with substance use was higher in the three months following the relaxation of restrictions on take-home doses than in the three months prior.
Mixed direction A cross-sectional study (S6) found increased hospital admissions for substance use among OAT clients who received additional take-home doses during the pandemic, but no difference in emergency department visits for substance use.A before-and-after study (S22) using self-reported data reported a decrease in mean physical health scores and an increase in mean psychological health scores following pandemic-related changes to OAT.

Quality of life
Quality of life was reported in two studies.See Table 13.Direction of effect was negative in one low-quality study (S16) and positive in one moderate-quality study (S22).Both studies were downgraded for unplanned co-exposures (see Table 14).
Negative direction A cross-sectional survey (S16) using the WHOQOL-BREF, a 26-item instrument for assessing quality of life, found that clients who received take-home doses had lower scores that those who continued to pick up their medication daily.
Positive direction A before-and-after study (S22) found that OAT clients had higher scores on quality of life scales following pandemic-related changes to OAT.

Subgroup analysis
Subgroup analysis by treatment type (Fig. 4) showed no clear difference between methadone and buprenorphine in changes in retention and overdose.For all other outcomes, it was not possible to investigate differences between treatment types because of insufficient data (fewer than two buprenorphine studies).An unplanned subgroup analysis of illicit substance use by substance type (opioids versus other unregulated substances) was inconclusive, though the direction of effect was more often positive or mixed for use of unregulated opioids than for use of other unregulated substances (Fig. 5).

Sensitivity analysis
We explored the sensitivity of our findings to study quality by excluding low-quality studies (shown in light gray in Fig. 3).Visual inspection of harvest plots showed a  decrease in the proportion of studies supporting a positive direction of effect for retention, although the overall trend was preserved.There were no notable changes in other outcome domains.

Qualitative synthesis
We identified four analytical themes describing clients' experiences with the relaxation of restrictions on takehome doses during COVID-19 (see Table 15).Clients' quotes reflected a holistic view of treatment and indicated that access to take-home doses influenced self-perception, treatment experience, and mental health during the pandemic.

Theme #1: feeling trusted to self-manage treatment
Alongside meeting client's physical needs, take-home doses increased client confidence.Offering this "bit of trust" (S33) made it possible for clients to reach a level of agency that previous medication policies did not allow.
With more ownership of their medication, clients had the space and time to exercise their expertise in their own care and look after their needs (S23, S32-33, S35, S40).However, though some clients found it "very easy" (S20) to adapt to take-home doses and wanted to protect their right to keep them (S1, S15, S27), a few stated that they "had trouble with take-home doses" (S14) or were not "ready for it" (S27).
Reassurance and responsibility Take-home doses were overwhelmingly seen as an indicator of trust (S1, S5, S15, S20, S33-34) between the clinician and the client.Takehome doses provided reassurance, signifying that clients "must be doing well" (S15) or were "on the right track" (S1, S15, S20) in their recovery.

When you get your take-home doses it's like you feel you are being trusted to take care of yourself, and do the right thing…it felt great…that I was on the right track in my recovery. (Client in S20, p. 5)
In some cases, clients felt that take-home doses helped them move forward and gave them a sense of pride and personal achievement (S15, S20, S27).

I feel that it's given me a sense of responsibility. I wasn't sure if I was ready to handle-but of course, I rose to the challenge. That makes me feel proud of myself. (Client in S15, p. 4) I was much more physically stable because I wasn't missing doses and also felt …it was sort of empowering
Table 8 Critical appraisal of quantitative studies reporting illicit substance use a The MMAT (Mixed Methods Appraisal Tool) Qualitative Checklist is designed specifically for mixed methods systematic reviews (Hong et al., 2018 [57]).It consists of five sections specific to various study designs, each with five quality criteria.All quantitative studies included in this review, including quantitative components of mixed-methods studies, were appraised under Sect.3: Quantitative non-randomized studies b This review included studies in which the intervention of interest (relaxed restrictions on take-home doses) formed part of a broader intervention (e.g., pandemicrelated changes to OAT treatment).To increase the relevancy of the quality assessments, we interpreted questions 4   Medication self-management With more control over their medication, clients had the flexibility to adapt their dosing schedule to meet their individual needs (S5, S23, S32, S33, S35a, S40).Take-home doses functioned as a bridge to more autonomous care by enabling clients to take a more active role in managing their treatment (S5, S15, S19-20).Self-governance made it possible for clients to take their medication at a time that suited needs, with some opting to take it later in the day (S15, S36a, S40) or preferring to split their dose (S12, S21, S23, S33).These aspects of medication ownership promoted better sleep (S5, S15, S36a) and helped clients navigate urges to use unregulated substances (S12, S15).

That has been quite a … luxury to be able to have what I need at home and be able to dose at my convenience. I found that I like to take it at night, (it makes me feel better), but I can't do that if I'm going to the clinic every day. (Client in S40, p. 1108) I was able to take my medication the way I was supposed to. I didn't have to think of taking extra, I didn't want to take extra. (Client in S15, p. 4)
Though most clients associated take-home doses with positive experiences, a few felt "overwhelmed" (S20), selfidentified as "addicts" (S1, S27), and were unsure of their ability to self-manage (S10, S20, S27).

[…] For me [access to take-home doses] just wasn't good at the time because I was still pretty new in my sobriety, you have to trust in yourself and everybody is different. (Client in S20, p. 6) I basically told on myself and told [the clinic] that I was having trouble with the take-home doses, so they stopped giving them to me…I like it better because [going to the clinic] gets me up and ready for the day. (Client in S14, p. 5)
A number of these clients had difficulty spacing out their doses and ran out of medication early (S1, S10, S16, S27).In some instances, they turned to unregulated opioids to ease the resulting withdrawal symptoms (S1, S27).

Table 10 Critical appraisal of quantitative studies reporting fatal and non-fatal overdose
a The MMAT (Mixed Methods Appraisal Tool) Qualitative Checklist is designed specifically for mixed methods systematic reviews (Hong et al., 2018).It consists of five sections specific to various study designs, each with five quality criteria.All quantitative studies included in this review, including quantitative components of mixedmethods studies, were appraised under Sect.3: Quantitative non-randomized studies b This review included studies in which the intervention of interest (relaxed restrictions on take-home doses) formed part of a broader intervention (e.g., pandemicrelated changes to OAT treatment).To increase the relevancy of the quality assessments, we interpreted questions 4 and 5 relevant to the research question posed in this review

No
Study MMAT Section 3 a for quantitative non-randomized studies
Wanting "less exposure" Before the COVID-19 pandemic, clients were not given the option of distancing themselves from the "triggers on the street" (S1, S32) that some encountered during their clinic visits.Takehome doses acted as a protective "barrier" (S15), creating space between clients and the "old people" (S15, S21) and places that they preferred to "stay away" from (S20).Clients were able to manage their environments to protect their wellbeing and recovery by choosing to avoid situations where they were "reminded of [their] drug history all the time" (S21, p. 37).

Cause when I would come here every day, I see people that I used with every day. And so when I am not seeing them every day I am getting a different type of habit. I am growing a different type of a habit outside of the clinic and so it's better for me that way I guess. (Client in S15, p. 6)
Additionally, some clients with take-home doses stated that picking up their medication less frequently protected them from threats of theft or coercion (S23, S33).

Table 11 Studies reporting measures of client health
Acronyms: HR hazard ratio, NR not reported, OR odds ratio, RR relative risk * Not reported in the original study; inferred or calculated by authors a Control group: OAT clients with at least one take-home dose during the pandemic, but without additional take-home doses b Intervention group: OAT clients with additional take-home doses during the pandemic  Reduced stigma and anxiety Compulsory clinic attendance for supervised dosing was seen as a "form of control" (S18) that created a constant fear of missing appointments and losing access to medication (S18, S34).With room to breathe, clients could create experiences and environments that were free from the stigma associated with receiving OAT (S18, S23, S34, S38).

The good thing is I don't have to keep going to the chemist which is a pain, a real pain [. . .] like they keep changing the pharmacist so you have to go through all the rigmarole of it being controlled and that, proving who you are and where you live and stuff. (Client in S34)
Not all clients felt more protected from environmental risks, and some preferred to pick up their medication on a more frequent basis (S1, S28).In one example, a lack of safe and reliable housing increased the risk of medication theft (S1), while others had concerns around medication loss and spillage (S27, S36) or accidental consumption of medication by others (S37).

Theme #3: life/treatment balance
Take-home doses reduced treatment burden and permitted clients to create space in their lives for employment, family, and rewarding daily activities.This facilitated a more "normal" life and made it easier for some clients to adhere to treatment (S5, S19, S24, S34).Reducing commutes to the clinic or pharmacy was particularly beneficial for clients balancing treatment with caregiving responsibilities (S15, S20, S38), physical disabilities (S15, S21, S34), mental health challenges (S19, S36a), or limited incomes (S15, S20-21, S34).
Employment For working clients, daily supervised dosing created recurrent conflicts between treatment and employment (S1, S15, S19, S24, S29, S34, S38).Many contended with lengthy commutes (S21, S24), limited hours of service (S1, S24), and unpredictable wait times (S24, S34) to get their medication.Some clients reported that it was challenging to obtain or keep employment (S1, S29); others had missed doses (S19, S34) or been driven to give up treatment (S24, S34): And there was no way I could make it all the way there to talk to the doctor and get everything set up, and then make it to work on time.There's no possible way.(Client in S24, p. 8) Take-home doses made it possible for clients to meet their work commitments without compromising their treatment, and vice-versa.

Table 15 Analytical themes resulting from qualitative synthesis
Theme #1: Feeling trusted to self-manage treatment Clients felt trusted when they were provided with take-home doses.Take-home doses reassured them that they were doing well in treatment and increased pride, responsibility, and treatment autonomy.
Theme #2: Navigating environmental risks Take-home doses allowed clients to reduce their exposure to triggers of illicit substance use and stigma.Reduced anxiety created space for experiences and environments that promoted client well-being.
Theme #3: Life/treatment balance Take-home doses eliminated daily conflicts between treatment obligations and employment.With treatment consuming less of their time and energy, clients gave their attention to family and other rewarding activities.
Theme #4: Emotional and psychological impact of not receiving take-home doses Clients who did not receive take-home doses during COVID-19 felt punished and exposed to unnecessary risk.Housing stability was a barrier to equitable treatment.

limited and I work and have depression so I couldn't get there every day. With take homes I'm far less likely to miss a dose and less likely to use. (Client in S19, p. 5)
Family and rewarding daily activities With greater control over their schedules, clients were "free" (S34) to give more attention to their families (S15, S20, S34, S36a) and to pursuing other rewarding activities (S1, S14-15, S20, S35a).These ranged from enjoying a leisurely morning coffee (S14) to going to the gym (S35a) and spending time outdoors:

Theme #4: emotional and psychological impact of not receiving take-home doses
Though some clients received additional take-home doses during the pandemic, others were required to continue with daily supervised dosing (S5, S40).Although their treatment was unchanged on the surface, the relaxation of restrictions on take-home doses had a profound emotional and psychological impact on many of these clients.
Anger and frustration with differential treatment Clients who continued to pick up their medication daily were acutely aware of the risk of COVID-19 infection during these visits (S14, S19, S38).Being forced to run "that germ gauntlet" (S19, p. 4) spurred anger and frustration, particularly given that other aspects of society had been radically overhauled to protect the general public: Clients whose take-home doses were revoked after the early phases of the pandemic also expressed dissatisfaction: I don't like [going from one month to 2 weeks] at all but, honestly, you don't rattle the cage too much…I feel kind of put upon in a way because…I shouldn't be in there with all the people.I am staying away from the grocery stores and everything but my methadone-of course.Anyway, I am not happy, but I'm not mad either.Just disappointed … (Client in S20, p. 5) Supervised dosing as punitive The feeling that supervised dosing was "punishment", either for substance use generally or for the behaviour of a minority of people using substances, was pervasive among clients (S21, S24, S35a, S40): [. ..] heroin addicts are, I believe, hated by society so there's a whole idea that you have to suffer … or be controlled.Otherwise, you're gonna do yourself some harm.(Client in S40, p. 1108) While some clients viewed daily supervised dosing as appropriate in certain cases (S1, S15, S34), particularly for those who were just beginning treatment [34], a common sentiment was that restrictions on take-home doses were crudely applied and needed to accommodate greater consideration of individual circumstances (S19-20, S24, S34).
In contrast, a few individuals felt that restrictions on take-home doses encouraged clients to be "dedicated" (S15) to their treatment adherence or abstinence (S15, S34), with one client explaining that having their takehome doses rescinded "gave me time to really acknowledge where I really messed up" (S1, p. 5).
Compounding inequities Clients who remained on takehome doses found clinics busier than usual (S18, S38), perhaps because of shorter hours of operation, social distancing measures, and reduced transit schedules (S14).Social distancing meant that some clients had to line up outside, where they felt conspicuous and exposed to judgment: Since the whole virus thing they've been like it has been like really packed, so to have to wait on line outside a lot it's embarrassing and I'm feeling things oh, look at them the drug addicts.(Client in S14, p.

1148)
Housing stability influenced access to take-home doses (S14, S35b).For clients with stable housing, the pandemic brought take-home doses into the realm of possibility; for those without, it cast their ineligibility into sharp relief: It's also been very difficult trying to stay clear of the virus…I didn't qualify for take-home doses.I don't have a home to take [methadone] to.I didn't qualify for a lockbox full of meds that I could give to anybody that was in a position of being able to watch me.Because nobody's in that position over me, I'm homeless [. ..] (Client in S14, p. 5) In this way, the liberalization of take-home doses increased treatment inequity for clients with unstable housing.

Sensitivity analysis
The majority of studies contributing to each qualitative theme were appraised as high-quality (see Tables 16,17,18 and 19).Excluding low-and moderate-quality studies from the synthesis did not change the findings appreciably.

Integrated analysis
We juxtaposed the quantitative and qualitative syntheses and found that the qualitative findings provided a plausible mechanism for the increased retention observed in the quantitative studies.We did not observe any evidence of an association between take-home doses, illicit substance use, and overdose risk in the quantitative synthesis.However, the qualitative findings suggested that this apparent lack of association may conceal individual variation in the impact of take-home doses.We identified a critical gap in the quantitative literature on quality of life, client health, and treatment satisfaction.See Fig. 6 for a visual representation of the integrated findings.

Reduced treatment burden observed in qualitative evidence may explain increased retention
The qualitative evidence suggests that reduced treatment burden may account for the increased retention observed in the quantitative synthesis.Definitions of treatment burden vary; however, it has been characterized as a multidimensional concept that includes the "physical, financial, temporal, and psychosocial" demands that treatment imposes on patients (Sav et al., 2013).Tran et al. (2014) take a similarly holistic view, describing treatment burden as "the 'work of being patient and its effect on the quality of life [...] the challenges associated with everything patients have to do to take care of themselves" (p.2).
In the qualitative synthesis, the burden of treatment included the costs of travelling to the clinic and the opportunity costs of losing or being unable to obtain employment because of conflicts with daily supervised dosing requirements.Several clients explicitly linked employment to missed doses or treatment discontinuation.Others discussed the physical and time burden of treatment; physical disabilities, mental health challenges, and caregiving responsibilities were described as challenges to frequent clinic attendance.
In addition, daily supervised dosing generated significant psychosocial burden.Inflexible treatment conditions forced clients to repeatedly subject themselves to environments where they felt mistrusted, stigmatized, and anxious about encountering substance use triggers.Takehome doses, in enabling clients to avoid negative experiences that reinforced "addict" identities, may have made them more likely to stay in treatment.

Individual variation in illicit substance use and overdose risk
The quantitative synthesis showed no evidence of an association between take-home doses and illicit substance use or overdose.It is possible that this finding conceals differences between subgroups, as the qualitative analysis showed individual variation in the relationship between take-home doses, illicit substance use, and overdose risk.
Some clients stated that take-home doses reduced their exposure to people and environments associated with use of unregulated substances.Others noted that takehome doses meant fewer missed doses and allowed them to administer their medication in a way that increased its perceived efficacy: for instance, through splitting their dose or taking it a preferred time of day.A few of these clients reported reduced withdrawal symptoms, allowing them to reduce their use of unregulated substances and, by extension, risk of overdose.
However, though most clients described positive experiences with take-home doses, a small number of individuals preferred the structure and accountability of daily dosing and had difficulty regulating their use of medication when given a multi-day supply.Two studies described instances of clients turning to the unregulated drug market, increasing their overdose risk, after consuming their medication before their next scheduled pick-up date.
Taken together, the qualitative and quantitative syntheses suggest that take-home doses may have decreased illicit substance use for some clients while increasing use within the smaller group of clients who experienced take-home doses as destabilizing.One of the primary studies in this review (S30) supports this hypothesis.The authors of this study reported that the percentage of urine tests positive for opioids in a cohort of OAT clients increased by an average of 10.6% during COVID-19, but that the percentage of clients abstinent from opioid use (defined as zero positive urine tests) increased from 26.5% to 53.7%, despite no significant change in the median number of urine tests per month.

Key facets of client experience not captured by quantitative studies
We identified no quantitative studies reporting on treatment satisfaction and very few studies reporting on client health or quality of life.The findings of the qualitative synthesis suggest that this is a significant gap.In describing the impact that take-home doses had on their lives, most clients focused on how take-home doses affected their perceptions of themselves, their experiences of treatment, and their mental health.Relatively few focused on the impact of changes on their use of illicit substances or risk of overdose, which, together with retention, were the most frequently reported outcomes in the quantitative studies.

Discussion
In this review, the relaxation of restrictions on takehome doses during the COVID-19 pandemic was associated with improved client experience and increased retention in OAT.We found no evidence that offering take-home doses to previously ineligible clients altered rates of illicit substance use or overdose in this population.We note that the risk of overdose in the community (i.e., from diverted medication) is also an important consideration.However, the scope of the present review was limited to the impact of take-home doses on individuals in treatment.Our findings align with the results of a recent policy review of the evidence on pandemicrelated regulatory changes to methadone treatment in the United States [49].Previous systematic reviews of supervised versus unsupervised dosing did not identify any studies of overdose and found no evidence of a difference in retention or illicit substance use [22,23].In both reviews, however, the authors concluded that the size and quality of the evidence base prevented them from drawing conclusions [22,23].

Treatment burden and retention in treatment
Our qualitative findings suggested that reduced treatment burden may explain the association between takehome doses and increased retention.There is growing recognition of the impact of treatment burden on people managing chronic conditions [111][112][113].Studies show a significant association between treatment burden and medication adherence; as burden increases, adherence decreases [113,114].In OAT, lower adherence may translate into lower retention because missed doses reduce medication effectiveness.In addition, the substance use that may result from missed doses can result in treatment dismissal in some OAT programs [18].Validated instruments for measuring treatment burden are a relatively recent development and have rarely been used in OAT [115].However, research supports an association between various dimensions of treatment burden and retention in OAT.For instance, retention decreases when the time burden of treatment is increased, as when treatment includes mandatory counselling [116] or when clients travel more than 30 min to reach their clinic [117].
The difficulty of balancing treatment and employment is widely recognized as a barrier to retention [36,118,119].In addition to anecdotal evidence of clients leaving treatment because of work conflicts [117,[120][121][122][123], a recent cohort study found that employment was a significant predictor of "sub-optimal care trajectories" in OAT [124].Stigma is a compounding factor, as reluctance to disclose OAT may discourage clients from seeking accommodations from their employers [125].
Commentators have responded to the growing body of research on treatment burden with calls for "minimally disruptive medicine" that recognizes the impact of treatment demands, such as supervised dosing requirements, on clients' lives [17,126].The findings of the present review suggests a need for further research using validated instruments to measure treatment burden in OAT.Optimizing the benefits of take-home doses In the integrated analysis, we concluded that an apparent lack of association between take-home doses, illicit substance use, and overdose may obscure differences in the impact of take-home doses on individual clients.Previous qualitative studies also show divergence in client experiences, with some clients preferring supervision [127] or stating that a short period of supervision is helpful upon treatment entry [29,128].
In the present review, as in previous studies [29,129], clients had insight into their ability to manage takehome doses.These findings suggest that the benefits of take-home doses can be optimized by treating clients as active participants in care planning.Retaining flexibilities around take-home doses in the post-COVID-19 era would allow providers and clients to evaluate the merits of take-home doses relative to individual treatment needs and preferences.Research supports the value of client engagement in improving experiences of treatment [130][131][132], enhancing therapeutic relationships [131,133], and determining effective dosages in OAT [134].
Based on the qualitative synthesis, factors that warrant discussion between providers and clients include the client's level of comfort with a higher degree of self-management, the benefits and disadvantages of decreased clinic attendance, and the impact of supervised dosing on the client's life/treatment balance.These discussions may occur in conjunction with consideration of other factors affecting individual risk, such as ongoing use of unregulated opioids.Findings also suggest that the option to return to supervised dosing if desired should be available to clients who request take-home doses.

Split dosing and medication effectiveness
In the qualitative synthesis, some OAT clients identified their ability to time their medication or split their dose as an advantage of take-home doses.Methadone is typically offered to OAT clients once a day because its average half-life approximates 24 h [135].However, medication interactions and wide variations in individual metabolism mean that some people on this regimen will have breakthrough withdrawal symptoms that cannot be resolved through a simple increase in dose [135].In a recent pharmacokinetic study, serum testing showed that 8.5% of the sample were ultra-rapid methadone metabolizers who would benefit from split dosing [136].
Increased access to split dosing may also benefit the 55-61% of methadone clients who report chronic pain [137].Management of pain in OAT clients is complicated by uncertainties around best practices [138], stigma and distrust from health care providers [139], and the complex relationship between pain and opioid use [140].Though methadone is not a first line treatment for pain in the general population, a recent systematic review suggests that a divided dose of methadone may be preferable to other opioid analgesics for some methadone clients with chronic pain [138].However, research in this area consists primarily of case series and case reports [138].For OAT clients using methadone for analgesia, multiple daily doses are necessary because methadone does not provide pain relief for as long as it suppresses withdrawal.Clients who use unregulated substances to alleviate chronic pain are unlikely to get the same benefit from once-daily methadone.
Relaxed restrictions on take-home doses, in making split dosing more accessible to clients, may increase medication effectiveness for rapid metabolizers while supporting treatment regimens that combine opioid maintenance with methadone for analgesia.

Patient-important outcomes
Of the quantitative outcomes included in this review, the most frequently reported were retention, substance use, and overdose.Given that clients in the qualitative synthesis focused primarily on the impact of take-home doses on their psychological state and life/treatment balance, relatively few quantitative studies reported on client health, quality of life, or treatment satisfaction.This is consistent with previous research demonstrating that common measures of effectiveness in OAT do not necessarily reflect the outcomes valued by clients [141][142][143][144][145].
Though reducing use of unregulated substances is a common treatment goal [142], many OAT clients also seek improved psychological wellbeing, improved relationships, improved role functioning, and decreased stigma and shame [132,143,146].Reed et al. (2023) found that clients asked to rate the importance of predefined recovery goals considered "having a sense of selfworth" as important as "not using opioids" [146] while Treloar et al. (2007) reported that clients valued takehome doses for making them feel trusted [147].Numerous studies have found that clients also value "feeling normal" or "living a normal life" [142,143,147,148] sentiments echoed by clients in the present review.
Recent studies have highlighted the limitations of traditional outcome measures and established the need for greater consideration of outcomes important to clients [142,144].In the qualitative studies included in the present review, clients valued the take-home doses that they received during the pandemic in part for their impact on psychological well-being.Substance use disorders are closely intertwined with anxiety, mood disorders, and other mental health challenges [149], and there is clear value in treatment delivery models and outcome measures that reflect the importance of meeting clients' mental health needs during OAT.To our knowledge, there is not yet a widely accepted set of patient-important outcomes for use in recovery from substance use disorder, although at least one such instrument has been developed [150].Involving people with lived and living experience of substance use in the development of patient-important outcome measures is essential to ensuring that they are relevant and meaningful to clients [150][151][152][153].

Strengths and limitations
The relaxation of restrictions on take-home doses occurred in conjunction with other changes to program delivery, such as increased use of telehealth and reduced frequency of urine testing.In the case of buprenorphine, which was subject to fewer restrictions than methadone pre-pandemic, the impact of these changes may have exceeded the impact of the relaxation of restrictions on take-home doses.However, few of the studies identified in this review focused exclusively or primarily on buprenorphine.The pandemic itself was associated with social upheaval, changes to the unregulated drug supply, and disruptions to harm reduction services [154,155].It was not possible to control for these confounders in the quantitative synthesis.However, we were able to mitigate this limitation by using a mixed methods approach that allowed us to triangulate the quantitative findings with qualitative data.In this review, the qualitative findings were consistent with an association between take-home doses and retention and suggested treatment burden as a plausible explanation.Nevertheless, the association that we observed between take-home doses and retention should be interpreted with caution, particularly given that a sensitivity analysis excluding low-quality studies weakened the evidence supporting a positive direction of effect for retention.We also note that the impact of take-home doses may have been influenced by factors that we could not fully account for in this review, such as the level of pre-pandemic restrictions, the flexibilities provided by guidelines issued during the pandemic, and the extent to which flexibilities were implemented.These are known to have varied substantially [47].
We synthesized the quantitative findings using vote counting based on direction of effect.This method is preferable to simple narrative synthesis in that it reduces bias in the presentation and interpretation of findings [59].It also has limitations.First, it provides no information about magnitude of effect [59].Though we found evidence of a positive association between take-home doses and retention, we are unable to conclude whether the size of this increase would be considered meaningful in a clinical setting.
Second, vote counting based on direction of effect is less powerful than other methods of synthesis [59].Compounding this limitation is the fact that a number of the quantitative studies used a before-and-after design that did not distinguish between clients who benefited from relaxed restrictions and those who remained on supervised dosing during the pandemic.This may have masked any associations between takehome doses and program effectiveness.Our finding of no association between take-home doses, illicit substance use, and overdose cannot be considered conclusive, particularly as the qualitative synthesis indicated that take-home doses were widely perceived as facilitating recovery.

Deviations from protocol
This review deviated from our protocol in that we did not contact subject matter experts to solicit unpublished manuscripts or re-run all searches prior to the final analysis.However, several of the databases that we searched included preprints (e.g., Ovid MEDLINE ALL; Embase) and we conducted an additional round of forward citation chaining on Mar.31, 2022, to capture articles published after the initiation of this review.We engaged with OAT clients by discussing our preliminary findings with seven community members with lived experience of OAT rather than through the town hall approach specified in our original research protocol.

Conclusions
In this mixed methods systematic review, we found that the relaxation of restrictions on take-home doses during the COVID-19 pandemic was associated with increased retention in OAT.See Fig. 7 for a summary of the implications of our findings for opioid agonist treatment.Qualitative evidence suggested that changes in retention may be attributable to reduced treatment burden.We found no evidence of an association between take-home doses and illicit substance use or overdose, despite the expansion of take-home doses to individuals who were ineligible to receive them prior to the pandemic.
Previous qualitative studies have demonstrated that daily supervised dosing is burdensome, stigmatizing, and viewed with disfavour by many clients [33,34,156,157].This review builds on that body of research by illuminating the ways in which more liberal provision of take-home doses altered clients' experiences of treatment during the COVID-19 pandemic.Though some clients reported challenges with managing their medication, the dominant narrative was one of appreciation, reduced anxiety, and a renewed sense of agency and identity.
Crucially, these benefits are not captured by traditional measures of effectiveness in OAT.This suggests that prepandemic policies on take-home doses severely underestimate their value to clients.Including patient-important outcome measures in policy, program development, and treatment planning is critical to ensuring that decisions around take-home doses accurately reflect their impact on people in opioid agonist treatment.
retention; (2) illicit substance use; (3) fatal and non-fatal overdose; (4) client health; (5) quality of life; and (6) treatment satisfaction?P (Population): People receiving OAT via any route of administration (e.g., oral, sublingual, buccal, injectable) I (Intervention): Relaxation of restrictions on take-home doses of OAT during the COVID-19 pandemic a C (Comparator): (1) No comparator OR (2) restrictions on take-home doses prior to the COVID-19 pandemic O (Outcomes): Program effectiveness, as indicated by incidence of (1) retention; (2) illicit substance use; (3) fatal and non-fatal overdose; (4) client health; (5) client quality of life; and (6) client treatment satisfaction Q2: What was the impact of relaxing restrictions on take-home doses during the COVID-19 pandemic on clients' experiences with OAT? P (population): People receiving OAT via any route of administration (e.g., oral, sublingual, buccal, injectable) I (phenomenon of Interest): Client experience (e.g.satisfaction with treatment, relationship with provider, self-efficacy, alignment of service with personal treatment goals, other patient-reported outcomes) Co (Context): Relaxation of restrictions on take-home doses of OAT during the COVID-19 pandemic a

Fig. 3
Fig. 3 Harvest plots showing results of synthesis by direction of effect

*
Not reported in the original study; inferred or calculated by authors a Retention defined as active buprenorphine prescription at least 90 days after treatment initiation b Control group: OAT clients initiating treatment after referral before pandemic c Intervention group: OAT clients initiating treatment after referral during pandemic d Control group: OAT clients from Jul.-Dec.2019 e Intervention groups: OAT clients from (a) Jan.-Jun.2020, (b) Jul.-Dec.2020, and (c) Jan.-Mar.2021 f The step transfer function was used to test for immediate change g The ramp transfer function used to test for gradual change h OAT discontinuation defined as a gap in therapy exceeding 14 days i Control group: OAT clients with no change in take-home doses during pandemic j Intervention group: OAT clients with increased take-home doses during pandemic k Control group: OAT clients pre-pandemic l Intervention group: OAT clients post-pandemic m Analysis limited to OAT clients with at least three months of pre-pandemic data and one month of post-pandemic data n "Stable clients" defined as clients with six months or more of buprenorphine prescriptions without a treatment disruption."Treatment disruptions" defined as gaps of 28 days was a pain in the ass because the closest ([methadone] clinic) is in Bullhead.So they got to pick you up at five o'clock in the morning, drive you down there in the bus [. ..] you have to go all the way down there to see the doctor [45 minutes].And there's no guarantee you're going to get your dose that day.And you have to sit there and wait and you make the bus wait.Well, after picking everybody up, you're looking at like two hours, something like that. ...That's why I stopped going to them because I had to go to work.

Fig. 5
Fig. 5 Exploratory subgroup analysis of illicit substance use

Fig. 6
Fig. 6 Visual representation of integrated findings

Table 2
Eligibility criteria used to screen studies Acronyms: OAT opioid agonist treatment

Table 3 (
continued) a Acronyms: MMT methadone maintenance treatment, MOUD medication for opioid use disorder, NR not reported, OAT opioid agonist treatment, OUD opioid use disorder, PWID people who inject drugs, PWLLE people with lived and living experience [of substance use], PWUD people who use drugs, SAMHSA Substance Abuse and Mental Health Administration b Qualitative findings were extracted from a preprint version of this manuscript.Comparison with the peer-reviewed publication showed no appreciable changes to the data extracted for this review c Qualitative findings from this mixed-methods preprint were later published in peer-reviewed form (Suen et al., 2022/Wyatt et al., 2022) and were therefore not extracted from the preprint

Table 4
Characteristics of participants in included studies

Table 4 (continued) No Study Sample a No. of OAT Clients in Sample Opioid Medication(s) Used Age b Sex Race and Ethnicity e
a Acronyms: GE gender-expansive, IQR interquartile range, NR not reported, SD standard deviation b Ages presented as mean (SD), unless otherwise specified.All ages are rounded to the nearest year c Based on 53 posts reporting age d Prescriber data excluded from sample characteristics e Race and ethnicity were initially extracted in dichotomized form (White/Non-White) to facilitate subgroup analysis.As subgroup analysis was not possible, one reviewer (AA) subsequently extracted a more detailed breakdown of the 'Non-White' category using the terms used in the original studies.All figures are rounded to the nearest percent.Some figures sum to more than 100% because of rounding and/or selection of multiple race and ethnicity categories f Qualitative findings from this mixed-methods preprint were later published in peer-reviewed form (Suen et al., 2022/Wyatt et al., 2022) and were therefore not extracted.Sample characteristics in this table are based on participants in the quantitative analysis g Inferred from type of treatment facility, general description of treatment, or participant quotes; may not be exhaustive list of participants' medications

Table 5
Studies reporting measures of retention

Table 6
Critical appraisal of quantitative studies reporting retention (Hong et al., 2018)ethods Appraisal Tool) Qualitative Checklist is designed specifically for mixed methods systematic reviews(Hong et al., 2018).It consists of five sections specific to various study designs, each with five quality criteria.All quantitative studies included in this review, including quantitative components of mixedmethods studies, were appraised under Sect.3: Quantitative non-randomized studies b This review included studies in which the intervention of interest (relaxed restrictions on take-home doses) formed part of a broader intervention (e.g., pandemicrelated changes to OAT treatment).To increase the relevancy of the quality assessments, we interpreted questions 4 and 5 relevant to the research question posed in this review

Table 7
Studies reporting measures of illicit substance use

Table 7 (
continued) NegativeWhere adjusted and unadjusted effect estimates were reported, we present adjusted estimates.In no case did this change the estimated direction of effect Acronyms: NR not reported, OR odds ratio.Where bivariate and multivariate analyses were reported, we present the results of the multivariate analysis * Not reported in the original study; inferred or calculated by authors a Control group: OAT clients pre-pandemic b Intervention group: OAT clients post-pandemic c Control 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) d Intervention group: 2020 values from a fitted model that removed the main effect of year e Control group: All OAT clients f Intervention group: OAT clients with additional take-home doses during pandemic g Adjusted for years in treatment, age, substance use disorder diagnosis, psychiatric disorder diagnosis, and % reduction in visit frequency h Analysis limited to clients with three months of pre-COVID-19 data and one month of post-COVID-19 data i Defined as an increase of 4 or more days in the previous 28 days j Control group: OAT clients with no take-home doses at follow up k Intervention group (a): OAT clients with 1-5 take-home doses/week at follow up l Intervention group (b): OAT clients with 6 + take-home doses/week at follow up m Based on proportion of comparisons favouring control n Based on mean control group value versus mean intervention group value and 5 relevant to the research question posed in this review

Table 9
Studies reporting measures of fatal and non-fatal overdose

Table 9 (
continued)Where adjusted and unadjusted effect estimates were reported, we present adjusted estimates.Where weighted and unweighted effect estimates were reported, we present weighted estimates.In no case did this change the overall estimated direction of effectAcronyms: HR hazard ratio, NR not reported, OR odds ratio, RR relative risk * Not reported in the original study; inferred or calculated by authors a Control group: Projected deaths of methadone clients, Mar.-Jun.2020 b Intervention group: Actual deaths of methadone clients, Mar.-Jun.2020 c Control group: Buprenorphine clients, 2016-2019 d Intervention group: Buprenorphine clients, 2020 e Control group: OAT clients pre-pandemic f Intervention group: OAT clients post-pandemic g Control group: OAT clients without additional take-home doses during pandemic h Intervention group: OAT clients with additional take-home doses during pandemic i ≥ 1 emergency department visit or inpatient hospitalization for opioid toxicity j Coroner-confirmed fatal opioid overdoses k Control group: OAT clients without additional take-home doses during pandemic l Intervention group: OAT clients with additional take-home doses (any number) during pandemic m Intervention group: Clients with additional take-home doses (at least a two-week supply) during pandemic n Could not be modelled because of small numbers o Overdoses reported to clinical personnel or documented in medical records p Fatal overdoses ascertained from electronic health records; defined as death [over 10-month follow-up period] with any or multiple illicit substances (including opioids) listed as any of the potential causes of death q Control group: Clients who never had take-home doses (neither before nor during pandemic) r Intervention group: Clients newly started on take-home doses during pandemic s Control group: Clients with no change or a decrease in take-home doses during pandemic t Intervention group: Clients with additional take-home doses during pandemic as well, because it means they are trusting you to have the six takeaways, you felt more like a normal person, more like an adult, being trusted with some responsibility and that was quite empowering.(Client in S5, p. 4)

Table 12
(Hong et al., 2018)ted data collected through the Australian Treatment Outcome Profile Critical appraisal of quantitative studies reporting client health a The MMAT (Mixed Methods Appraisal Tool) Qualitative Checklist is designed specifically for mixed methods systematic reviews(Hong et al., 2018).It consists of five sections specific to various study designs, each with five quality criteria.All quantitative studies included in this review, including quantitative components of mixedmethods studies, were appraised under Sect.3: Quantitative non-randomized studies b This review included studies in which the intervention of interest (relaxed restrictions on take-home doses) formed part of a broader intervention (e.g., pandemicrelated changes to OAT treatment).To increase the relevancy of the quality assessments, we interpreted questions 4 and 5 relevant to the research question posed in this review c Control group: OAT clients pre-pandemic d Intervention group: OAT clients post-pandemic e

Table 13
Studies reporting measures of quality of life a Control group: OAT clients who attended the OST site for medication every day during the pandemic b Intervention group: OAT clients receiving take-home doses during pandemic c Control group: OAT clients pre-pandemic d Intervention group: OAT clients post-pandemic

Table 14
(Hong et al., 2018)of quantitative studies reporting quality of life a The MMAT (Mixed Methods Appraisal Tool) Qualitative Checklist is designed specifically for mixed methods systematic reviews(Hong et al., 2018).It consists of five sections specific to various study designs, each with five quality criteria.All quantitative studies included in this review, including quantitative components of mixedmethods studies, were appraised under Sect.3: Quantitative non-randomized studies b This review included studies in which the intervention of interest (relaxed restrictions on take-home doses) formed part of a broader intervention (e.g., pandemicrelated changes to OAT treatment).To increase the relevancy of the quality assessments, we interpreted questions 4 and 5 relevant to the research question posed in this review Fig. 4 Subgroup analysis of retention and overdose by treatment type

Table 16
(Hong et al., 2018)of qualitative studies supporting Theme #1: feeling trusted to self-manage treatment a The MMAT (Mixed Methods Appraisal Tool) Qualitative Checklist is designed specifically for mixed methods systematic reviews(Hong et al., 2018).It consists of five sections specific to various study designs, each with five quality criteria.All qualitative studies included in this review including qualitative components of mixedmethods studies, were appraised under Sect.1: Qualitative studies

Table 17
(Hong et al., 2018)of qualitative studies supporting Theme #2: navigating environmental risks a The MMAT (Mixed Methods Appraisal Tool) Qualitative Checklist is designed specifically for mixed methods systematic reviews(Hong et al., 2018).It consists of five sections specific to various study designs, each with five quality criteria.All qualitative studies included in this review including qualitative components of mixedmethods studies, were appraised under Sect.1: Qualitative studies

Table 18
(Hong et al., 2018)of qualitative studies supporting Theme #3: life/treatment balance a The MMAT (Mixed Methods Appraisal Tool) Qualitative Checklist is designed specifically for mixed methods systematic reviews(Hong et al., 2018).It consists of five sections specific to various study designs, each with five quality criteria.All qualitative studies included in this review including qualitative components of mixedmethods studies, were appraised under Sect.1: Qualitative studies

Table 19
(Hong et al., 2018)of qualitative studies supporting Theme #4: emotional and psychological impact of not receiving takehome doses a The MMAT (Mixed Methods Appraisal Tool) Qualitative Checklist is designed specifically for mixed methods systematic reviews(Hong et al., 2018).It consists of five sections specific to various study designs, each with five quality criteria.All qualitative studies included in this review including qualitative components of mixedmethods studies, were appraised under Sect.1: Qualitative studies