Road to Recovery: protocol for a mixed-methods prospective cohort study evaluating the impact of a new model of substance use care in a Canadian setting

Abstract

Introduction

The Road to Recovery (R2R) Initiative is an innovative model of substance use care that seeks to increase treatment capacity by creating approximately 100 new addiction treatment beds to provide on-demand addiction care in Vancouver, British Columbia, for patients with substance use disorders. The new model also coordinates the region’s existing clinical substance use services to support patients across a care continuum that includes traditional office-based addiction treatment and harm reduction services, early withdrawal management and more intensive abstinence-based treatment programming. To understand the impact of offering on-demand and coordinated substance use care, an observational cohort of individuals who access any R2R clinical service will be created to examine health and social outcomes over time.

Methods and analysis

This prospective mixed-methods study will invite individuals from Vancouver, Canada, who access substance use treatment through the R2R model of care to (1) complete a baseline and 12-month follow-up quantitative questionnaire that solicits sociodemographic, substance use and previous addiction treatment data and (2) provide consent to the use of participants’ personal identifiers to access health records for chart review and for annual linkage to select health and administrative databases to allow for ongoing (virtual) community follow-up over 5 years. Additionally, a purposive sample of cohort participants will be invited to participate in baseline and 12-month follow-up qualitative interviews to share their experiences accessing R2R and identify challenges and opportunities associated with the implementation of R2R.

Ethics and dissemination

The study was approved by the University of British Columbia Providence Health Care Research Ethics Board in September 2023. Results from the proposed study will be published in peer-reviewed journals, presented at national and international scientific conferences and disseminated through regular meetings with policymakers, individuals with lived and living experience, and other high-level stakeholders, academic presentations and lay media.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Leveraging the implementation of an innovative model of substance use care, the proposed study provides an opportunity to generate insights from high-quality data that will improve the delivery of a continuum of care to individuals living with substance use disorders.

• Employing a mixed-methods approach, this study will triangulate quantitative data on substance use patterns and healthcare utilisation with qualitative data from individuals accessing harmonised substance use services, offering a detailed understanding of the outcomes, patient experiences and public health implications of a comprehensive, coordinated substance use care programme.

• Although the study lacks a formal control group, we will employ multiple sources of data to enable comparative analyses.

• Generalisability beyond Vancouver and to other healthcare systems which employ non-public pay models may be limited because of the important influence of the context-specific availability of social and community substance use services in different locations.

Introduction

Since the declaration of a public health emergency in Canada in 2016 and the USA in 2017 in response to rising overdose deaths, more than 400 000 Americans and 38 000 Canadians have died from drug overdoses.¹,⁵ Within Canada, British Columbia (BC) shoulders a substantial burden of that loss, exceeding 13 000 overdose deaths since 2016.⁶ BC’s unpredictable and toxic drug supply combined with the province’s siloed and fragmented substance use disorder (SUD) treatment system are thought to be key contributors.⁷ The burden of disease is not restricted to criminalised substances. The prevalence of alcohol misuse across Canada is at an unprecedented high, between 2015 and 2016 over 77 000 hospitalisations were attributed to alcohol-related harms, with the highest rates observed in BC.⁸ Accordingly, the BC provincial government and representative health organisations have urgently called for the development and implementation of a coordinated system of substance use care that spans a continuum ranging from harm reduction services to abstinence-based programming.^{9 10}

To address care gaps, Providence Health Care (PHC; a regional health authority in the province of BC), in collaboration with the provincial government’s Ministry of Mental Health and Addictions (MMHA), and Vancouver Coastal Health (VCH; another regional health authority), are spearheading the Road to Recovery (R2R) Initiative, a first-of-its-kind model of substance use care.¹¹ BC’s existing substance use treatment system is composed of both publicly and privately funded service providers in bed-based and outpatient settings. Quality and standards vary substantially between service providers, and interorganisational communication is extremely limited. The R2R model of care seeks to address this by (1) improving access to care through the creation of approximately 100 new substance use treatment beds and (2) offering comprehensive substance use management along the entire care continuum (ie, harm reduction, withdrawal management, bed-based recovery and outpatient longitudinal follow-up). Please refer to online supplemental table S1 for the list of R2R Services.

R2R offers a unique opportunity whereby the centralisation of the triage and delivery of substance use care facilitates coordination across the region to address key service gaps (ie, access bed-based services). Individuals living with a SUD will be medically assessed and triaged across various healthcare sites (eg, clinics, emergency departments) to the most appropriate care setting, and standardised, evidence-based care practices will be available across the care continuum. The discharge will be coordinated with a long-term recovery bed-based intake or bridged with the use of a transitional care bed—a temporary care bed for those who no longer require withdrawal management but whose disposition remains unclear. Beyond this, a strong focus of R2R is inclusivity, and honouring what matters most to individual patients to create a safe environment for recovery. Efforts to support this include the incorporation of cultural and traditional healing practices to address the unique needs of Indigenous Peoples, who have been disproportionality impacted by the substance use crisis as a consequence of colonialism, systemic racism and the unbalanced concentration of major health inequities.

To determine outcomes related to the offering of comprehensive and coordinated substance use treatment, the aims of this new cohort are (1) to evaluate over time the impact of harmonised addiction services on substance use trajectories, healthcare access and utilisation, and health and social outcomes; (2) to characterise access and journeys through the regional substance use treatment system and (3) to explore patient experiences, identify facilitators and barriers to substance use treatment and gather suggestions for improvement to inform future scale-up of R2R services across BC.

Methods and analysis

Study summary

The proposed study will constitute a cohort of patients accessing substance use treatment services parallel to the phased operationalisation of R2R (see figure 1 for planned phased roll-out) in the VCH region. VCH is one of five regional health authorities in BC.

Figure 1. Overview of the phased implementation of the Road to Recovery Initiative services from 2022 to 2027.

In this mixed-methods prospective cohort design, patients receiving R2R services (online supplemental table S1) will be observed through primary data collection via (1) interviewer-administered baseline and 12-month questionnaires and (2) qualitative interviews. Participants will be asked to provide consent for the use of personal identifiers (ie, a unique and persistent identifier issued to all provincial residents to access healthcare) to support secondary data collection through (1) 30-day prospective follow-up from the time of R2R admission using electronic medical record chart review and (2) linkage annually to provincial health and administrative databases over a 5-year follow-up period. A purposive sample of cohort participants will also be invited to participate in baseline and 12-month follow-up qualitative interviews.

Employing a convergent parallel design, quantitative and qualitative data collection will occur in parallel, of which results are later compared and related to one another (figure 2).¹² This will be an iterative process, as findings from each convergent analysis will inform adaptations and additions to data collection processes across implementation phases of R2R.

Figure 2. A schematic summary of the convergent parallel mixed-methods design being implemented in the Road to Recovery (R2R) Initiative Research Evaluation.

Primary study objectives and hypotheses

Below, we briefly summarise the primary study objectives and corresponding hypotheses. Table 1 summarises the proposed methods and data sources used to address the study objectives.

Table 1. Summary of the R2R evaluation study objectives, hypotheses and proposed methods and data sources.

Objective		Hypotheses	Methods	Data source
1	To describe substance use patterns and harms related to ongoing substance use following R2R access	Hypothesis 1: R2R access, including both frequency and duration of R2R treatment time will be associated with a reduction in substance use and substance-related harms; the greatest reduction is anticipated in underserved subgroups.Hypothesis 2: R2R access will be associated with positive changes to exposures associated with increased morbidity and mortality engagement with high-risk income-generating activities (eg, sex work). We anticipate effects will be attenuated for patients using legal substancesHypothesis 3: R2R retention will be associated with reduced rates of relapse and overall longer durations of abstinence.	Quantitative	Baseline and 12-month questionnaires 30-day chart review Provincial Administrative Health Databases
2	To characterise acute healthcare utilisation patterns following R2R access	Hypothesis 1: R2R access will be associated with a reduction in acute healthcare utilisation (ie, emergency department visits).Hypothesis 2: R2R retention (eg, completion of withdrawal management) will be associated with a longer time and fewer hospital presentations/readmissions.	Quantitative	Provincial Administrative Health Databases
3	To evaluate substance use treatment trajectories over time and identify key predictors associated with favourable outcomes (eg, longer time to relapse, prolonged treatment adherence, reduced healthcare utilisation).	Hypothesis 1: R2R retention will be positively associated with subsequent enrolment in a bed-based substance use treatment programme, short-stay withdrawal management programmes, primary care, community-based programmes, harm reduction services, and support networks. Access to services will positively impact substance use outcomes and support patients to meet their treatment goals.Hypothesis 2: R2R services will observe lower rates of unplanned discharges, even during high-risk periods including days associated with synchronised income assistance payments.13Hypothesis 3: R2R retention will be positively associated with a longer duration of substance use disorder medication adherence (eg, methadone, alcohol relapse prevention medications).Hypothesis 4: Increased access to R2R services (ie, number of admissions for withdrawal management) will be concentrated in more severely unwell patients, however, will be positively associated with increased access to services (eg, opportunities to make referrals, manage medical comorbidities).	Quantitative	Baseline and 12- month questionnaires 30-day chart review Provincial Administrative Health Databases
4	To investigate health and well-being (eg, social functioning, physical health) among study participants over time	Hypothesis 1: R2R access will improve social outcomes (eg, shifts towards less harmful income generation, more secure housing, improved food security), through reduced expenditure for money on drugs and increased institutional engagement.Hypothesis 2: R2R access will improve quality of life, with distinct gains in health domains, including perceived access to services), safety and shelter.Hypothesis 4: Patients accessing R2R services will experience improved physical health (eg, reduced hospitalisations for infections).	Quantitative	Baseline and 12- month Questionnaires 30-day chart review
5	To qualitatively capture how R2R service access has impacted study participant’s health (eg, substance use) and social outcomes (eg, housing) over time	Participant experiences in R2R will demonstrate a range of pathways to improved outcomes through positive and sustained institutional interactions, improved health behaviour, improved social relationships and increased material security	Qualitative	Semistructured open-ended interviews

R2RRoad to Recovery

Objective 1: to describe substance use patterns and harms related to ongoing substance use following R2R access

Hypothesis 1: R2R access, including both frequency and duration of R2R treatment time, will be associated with a reduction in substance use and substance-related harms; the greatest reduction is anticipated in underserved subgroups (eg, women, youth and Indigenous populations).

Hypothesis 2: R2R access will be associated with positive changes to exposures associated with increased morbidity and mortality (eg, injection alone), engagement with high-risk income-generating activities (eg, sex work). We anticipate effects will be attenuated for patients using legal substances (nicotine and alcohol).

Hypothesis 3: R2R retention will be associated with reduced rates of relapse and overall longer durations of abstinence.

Objective 2: to characterise acute healthcare utilisation patterns following R2R access

Hypothesis 1: R2R access will be associated with a reduction in acute healthcare utilisation (ie, emergency department visits).

Hypothesis 2: R2R retention (eg, completion of withdrawal management, connection to other R2R services) will be associated with a longer time and fewer hospital presentations/readmissions.

Objective 3: To evaluate substance use treatment trajectories over time and identify key predictors associated with favourable outcomes (eg, longer time to relapse, prolonged treatment adherence and reduced healthcare utilisation).

Hypothesis 1: R2R retention will be positively associated with subsequent enrolment in a bed-based substance use treatment programme, short-stay withdrawal management programmes, primary care, community-based programmes, harm reduction services and support networks. We anticipate the aggregation of services will positively impact substance use outcomes and support patients to meet their treatment goals.

Hypothesis 2: R2R services will observe lower rates of unplanned discharges, even during high-risk periods including days associated with synchronised income assistance payments.¹³

Hypothesis 3: R2R retention will be positively associated with a longer duration of SUD medication adherence (eg, methadone and alcohol relapse prevention medications).

Hypothesis 4: Increased access to R2R services (ie, number of admissions for withdrawal management) will be concentrated in more severely unwell patients, however, will be positively associated with uptake of other health services (eg, connection to primary care and management of infectious diseases).

Objective 4: to investigate health and well-being (eg, social functioning, physical health) among study participants over time

Hypothesis 1: R2R access will improve social outcomes (eg, shifts toward less harmful income generation, more secure housing, improved food security), through reduced expenditure for money on drugs and increased institutional engagement.

Hypothesis 2: R2R access will be associated with an improvement in reported quality of life, with distinct gains in health domains, including perceived access to services), safety and shelter.

Hypothesis 4: Patients accessing R2R services will experience improved physical health (eg, reduced hospitalisations for infections and medical comorbidities, etc).

Objective 5: To qualitatively capture how R2R service access may have impacted study participant’s health (eg, unregulated drug use, risk factors for overdose, treatment trajectory, quality of life) and social outcomes (eg, custody of children, housing, employment, social functioning) over time.

Participant experiences in R2R will demonstrate various pathways to improved outcomes through positive and sustained institutional interactions, improved health behaviour, improved social relationships and increased material security.

Study setting and participants

Intervention

The R2R model of care distinguishes itself from previous systems of addiction care through its comprehensive portfolio of available services (online supplemental table S1) and further through restructuring of the regional coordination and delivery of addiction services. Prior to R2R, centralisation of the triage and delivery processes through a single health provider (ie, PHC) did not occur. Previously, delivery of withdrawal management services was not guided by triaging protocols (ie, protocols considering comorbid medical illness, withdrawal severity or housing status), which limited accessibility and led to major inefficiencies including long waitlists and frequently missed intakes. Although hospital settings (ie, emergency departments, medical wards) can provide on-demand withdrawal management services, this often occurs outside of recovery-focused units and, in many cases, without the support of addiction medicine specialists or robust aftercare options. Through generating a unified system, with improved capacity to support patients in need of monitored withdrawal management, the R2R model of care seeks to relieve hospitals and emergency departments from managing single addiction-based issues (ie, withdrawal). Ultimately, it is thought that centralising the triage and delivery of R2R services through a single provider will strengthen coordination of care, enhance accessibility by streamlining service delivery and raise the quality of care by establishing a uniform standard that is applied across the entire treatment continuum.

Furthermore, the R2R model of care will increase treatment capacity through the addition of 100 treatment beds (over the phased implementation). These beds are being added to the existing substance use system of care (ie, including other public and privately funded addiction services not delivered by PHC). R2R treatment beds will also include new types of recovery-based options that previously did not exist. These include transitional care beds, which afford patients the ability to remain in a recovery-focused environment while they await access to longer stay addiction treatment beds (ie, 30–90 days intensive recovery-based treatment facilities) or recovery-focused housing. Before the introduction of these beds through R2R, structurally vulnerable individuals without stable housing would often return to unstable environments (ie, shelters, street) after completing withdrawal management, increasing the likelihood of relapse, further destabilisation and repeated admissions to withdrawal management facilities. R2R seeks to address this gap by supporting patients throughout their entire journey—from initial contact (ie, emergency rooms and community addiction clinics), through acute stabilisation and withdrawal management (ie, hospital or community-based withdrawal management beds) and ultimately into more stable, long-term recovery-based settings. Access to transitional care beds helps minimise the risk of destabilisation during this process. Ultimately, the addition of new treatment beds and services under a single health provider are intended to function collectively as comprehensive system of addiction care, where, for example, PHC can deliver the entire spectrum of substance use services, which will be publicly funded, widely accessible and support seamless transitions across all phases of an individual’s recovery journey.

Setting

The VCH region has consistently observed the highest number of drug overdose deaths as well as the second-highest rate of overdose deaths in BC.¹⁴ Further, approximately 74% of all overdose presentations in the VCH region receive medical care at St. Paul’s Hospital’s (SPH) emergency department.¹⁵ Given this, SPH was chosen as the preferred site for the initial implementation of R2R and study recruitment.

Participants

Eligible participants include individuals who have accessed one or more R2R services, are 18 years or older, can provide written informed consent and currently reside or plan to remain living in BC. Individuals will not be eligible for inclusion if they have an acute, severe mental health or medical condition limiting to provide informed consent, as determined by the clinical team. Acknowledging multiple R2R encounters can occur, participants will only be allowed to enrol once.

Recruitment

Recruitment will occur across R2R services (eg, withdrawal management, bed-based recovery) with healthcare providers (eg, nurses, physicians) informed about the study and connecting suitable participants to a member of the research team, who will assess eligibility and then obtain informed consent.

Participant remuneration

Participants will be paid $C30.00 for completing each quantitative questionnaire ($C60.00 total for completing both questionnaires; $C30.00 for completing the baseline questionnaire and $C30.00 for 12-month follow-up questionnaire). Participants participating in the qualitative portion of the study will be paid $C40.00 per interview ($C80.00 total for participating in both interviews; $C40.00 for participating in the baseline interview and $C40.00 for the follow-up interview).

R2R quantitative study

Below is a summary of measures taken to address quantitative study objectives. Data collection and management process undertaken to promote data quality are summarised in online supplemental table S2.

Baseline questionnaire

The baseline questionnaire solicits sociodemographic, social/structural and some clinical (eg, age, gender, ethnicity education, income, housing status, income generating activity, social support and primary care provider) information as well as substance use history (eg, duration of use for each substance), current substance use patterns (eg, substance-specific use, routes of use, frequency of use), health measures (eg, overdose, harm reduction access), previous SUD treatments and current treatment goals, and criminal justice system involvement. Validated tools incorporated into the questionnaire include the recovery capital¹⁶ and EUROQOL-5-Dimension 5-level quality of life instruments.¹⁷

Study follow-up and outcome measures

Longitudinal follow-up for patients entering the study will occur at 30 days (electronic medical record chart review), 1-year (12-month follow-up questionnaire) and annually over a 5-year follow-up period (health and administrative database linkages). Outcomes were selected to reflect patient-centred, clinically relevant and other commonly employed endpoints identified in previous literature.¹⁸ Furthermore, PHC’s Indigenous Wellness and Reconciliation (IWR) team successfully led a patient-journey mapping exercise with Indigenous individuals with lived/living experiences of SUDs to gain an understanding of the health and social outcomes that matter most to them, which informed selection of study outcomes. Please refer to table 2 for a summary of the study outcomes, timing of measurement and data source.

Table 2. Summary of prospective cohort study outcomes for the R2R evaluation.

Outcome domain, mapped to Study Objective	Outcome	Measurement	Follow-up time	Methods:study design	Methods comparative data source(s)
Substance useobjectives 1–3	Substance use patterns	1. Baseline and 12-month questionnaire	1. Baseline and 1-year follow-up	Quantitative	Self-controlleddesigns, BCCSU cohorts
	Complications from substance use (eg, overdose, hospitalisation)	1. Baseline and 12-month questionnaire2. Baseline and 12-month qualitative interview3. 30-day chart review4. Administrative data linkage	1. Baseline and 1-year follow-up (questionnaire interviews)2. 30 days post enrolment (chart review)3. 5-year follow-up (administrative data linkage)	Quantitative and qualitative	Self-controlled designs, BCCSU cohorts, BC SUD population
	Substance use treatment	1. Baseline and 12-month questionnaire2. Baseline and 12-month qualitative interview3. 30-day chart review4. Administrative data linkage	1. Baseline and 1-year follow-up (questionnaire, interviews)2. 30 days post enrolment (chart review)3. 5-year follow-up (administrative data linkage)	Quantitative and qualitative	Self-controlled designs, BCCSU cohorts, BC SUD population
Patient important outcomesobjective 3	Goals of treatment	1. Baseline and 12-month questionnaire5. Baseline and 12-month interviews	4. Baseline and 1-year follow-up (questionnaire, interviews)	Quantitative and qualitative	Self-controlled designs
	Ranking of treatment goals	6. Baseline and 12-month questionnaire	5. Baseline and 1-year follow-up	Quantitative	Self-controlled designs
Physical healthobjective 4	ED visits, hospitalisation	1. Baseline and 12-month questionnaire2. Baseline and 12-month qualitative interview3. 30-day chart review4. Administrative data linkage	1. Baseline and 1-year follow-up (questionnaire interviews)2. 30 days post enrolment (chart review)3. 5-year follow-up (administrative data linkage)	Quantitative and qualitative	Self-controlled designs, BCCSU cohorts, BC SUD population
	Health utilisation for chronic medical comorbidities	1. Baseline and 12-month questionnaire2. Baseline and 12-month qualitative interview3. 30-day chart review4. Administrative data linkage	1. Baseline and 1-year follow-up (questionnaire interviews)2. 30 days post enrolment (chart review)3. 5-year follow-up (administrative data linkage)	Quantitative and qualitative	Self-controlled designs, BCCSU cohorts, BC SUD population
Social and structural outcomesobjective 4	Housing status	1. Baseline and 12-month follow-up	1. Baseline and 1-year follow-up	Quantitative	Self-controlled designs, BCCSU cohorts
	Employment	1. Baseline and 12-month follow-up	1. Baseline and 1-year follow-up	Quantitative	Self-controlled designs, BCCSU cohorts
	Interaction with criminal justice system	1. Baseline and 12-month questionnaire2. Baseline and 12-month qualitative interview3. Administrative data linkage	1. Baseline and 1-year follow-up (questionnaire, interviews)2. 5-year follow-up (administrative data linkage)	Quantitative	Self-controlled designs, BCCSU cohorts
	Recovery capital	1. Baseline and 12-month questionnaire	1. Baseline and 1-year follow-up	Quantitative	Self-controlled designs, BCCSU cohorts
	Quality of life	1. Baseline and 12-month questionnaire2. Baseline and 12-month interviews	1. Baseline and 1-year follow-up (questionnaire, interviews)	Quantitative and qualitative	Self-controlled designs, BCCSU cohorts
Experiences with R2R (facilitators, barriers)objective 5	Treatment experience	1. Baseline and 12-month interview	1. Baseline and 1-year follow-up	Qualitative	N/A
	Facilitators to R2R treatment	1. Baseline and 12-month interview	1. Baseline and 1-year follow-up	Qualitative	N/A
	Barriers and challenges with R2R treatment model	1. Baseline and 12-month interview	1. Baseline and 1-year follow-up	Qualitative	N/A

Self-controlled designs account for (1) pre–post designs (comparison of baseline and 12-month follow-up) and (2) where possible comparative ITS analyses using linkage to existing BCCSU cohort data for R2R participants who previously participated in BCCSU existing cohort studies.

BCCSUBritish Columbia Centre on Substance UseEDemergency departmentITSinterrupted time seriesN/Anot applicable R2RRoad to RecoverySUDsubstance use disorder

1. 30-day electronic medical record chart review: A chart review of participants’ hospital electronic medical records will be completed 30 days following study enrolment to document a comprehensive summary of treatments provided while accessing R2R services. Outcomes of interest include medications prescribed and administered for withdrawal management, additional clinical services offered, referrals made and disposition (eg, planned or unplanned discharge).

2. 12-month follow-up questionnaire: A follow-up questionnaire will be administered 12 months following study enrolment. Outcomes of interest include past year, past 6 months and past 30-day substance use patterns, substance use treatment access, retention and goals, complications from substance use (eg, overdose) and social functioning (eg, housing status, employment status, involvement with the criminal justice system, quality of life, recovery capital).

3. Annual database linkages: Primary study data generated through the interviewer-administered questionnaires will be linked annually to provincial health and administrative databases over a 5-year follow-up period. Outcomes of interest include prescription drug information, healthcare utilisation (eg, hospital admissions), use of provincial medical services (eg, primary care, emergency health services), overdoses (fatal and non-fatal) and deaths. A summary of the linked administrative databases is included in online supplemental table S3. While the performance characteristics for substance use outcomes using BC administrative data are not reported, validation in US-based studies (which similarly rely on International Classification of Disease codes) demonstrate excellent performance for the classification of opioid overdose (sensitivity=97.2%, specificity=84.6%).¹⁹

Proposed comparator groups

While randomised controlled trials remain the gold standard for evaluations of clinical interventions, ethical concerns exist in adopting this approach to evaluate the R2R model of care. There is no clinical equipoise in randomly assigning drug-using individuals to comprehensive and coordinated substance-use care during the overdose crisis. Accordingly, we propose the following approaches to constitute comparison groups to facilitate:

1. Self-controlled study designs through retrospective database linkages: When obtaining informed consent, participants will be asked to agree to retrospective data linkage for 5 years prior to study enrolment and prospective annual database linkages over a 5-year follow-up period. This approach will allow participants to act as their own ‘control’, allowing specific health, social outcomes, and healthcare utilisation patterns to be compared in the 5-year pre-R2R and post-R2R engagement.

2. Propensity score matching (PSM) with population-level data: PSM will be undertaken between R2R participants and a control group of the wider BC SUD population. The BC SUD control population will be identified using SUD-related ICD codes in health service and death data, as well as SUD-related medications, and who have not used R2R services. Conducting PSM analysis to a province-wide cohort provides a unique opportunity to enable a comparison of vital health and substance use outcomes at a population level.

3. PSM with existing local BCCSU cohort studies: The Vancouver Injection Drug Users Study (VIDUS)²⁰ and the AIDS Care Cohort to Evaluate exposure to Survival Services (ACCESS)²¹ are two long-running community-recruited prospective cohort studies of more than 2500 adults who are at risk of HIV and use drugs (ACCESS). The At-Risk Youth Study (ARYS)²² is a multiyear study of street-involved youth between the ages of 14 and 26. Drawn from the same population of interest as individuals eligible for R2R services, protocols for these cohorts (the ‘BCCSU cohorts’) are harmonised to facilitate pooled analyses when appropriate. BCCSU cohort participants complete biannual study interviews that are compatible with time periods for our data gathering. We propose to match R2R cohort participants and treatment-seeking BCCSU cohort participants not exposed to R2R. This is intended to prevent selection bias resulting from differences in substance use goals, as R2R participants will be, by definition, substance use care-seeking individuals. To avoid contamination, existing BCCSU cohort participants will be ineligible to participate in matched analyses if they received R2R services during the analysis period. R2R participants who participated in one of the BCCSU cohort studies will be withdrawn from the respective BCCSU cohort for comparative analyses.

4. Data linkage with existing BCCSU cohort study questionnaires: Among R2R cohort participants who previously participated in either the VIDUS, ACCESS or ARYS studies, we also propose R2R data linkage with the participant’s responses to the BCCSU cohort study questionnaires. This will provide the opportunity to complete a quasi-experimental study using comparative interrupted time series analyses (ITS), whereby existing detailed assessments of substance use behaviours and social functioning and substance use treatment history (captured from existing BCCSU cohort study questionnaires) will be compared pre-enrolment and postenrolment in R2R services.

Sample size

Power calculations were conducted for the hypothesis evaluating engagement in hospital substance use treatment through R2R on time to mortality, assuming a conservative overall event rate (p) between 0.03 and 0.18 based on previous literature evaluating community methadone administration and mortality.²³,²⁵ Methadone literature guided these calculations given its common use for opioid use disorder (OUD) treatment and the high prevalence of OUD presentations to SPH’s emergency department.²⁶ A total of 1348 participants and 162 events are required to achieve a power of 80% to detect a difference of 27% (HR of 0.73, based on previous Vancouver-based studies²³) between the participants receiving R2R services (intervention) and those not (comparison) group, assuming an event rate (mortality) of 12%, at a level of significance set at α=0.05. These estimates were made using the power command in STATA V.17.0, allowing for the adjustment of ~20 covariates.²⁷ Online supplemental table S4 summarises the sample size needed given various parameter values.

Statistical methods

Analyses will be performed separately for each proposed comparator group (eg, existing BCCSU cohorts vs provincial substance use cohort) given varying data sources and availability of variables. For PSM analysis to a prespecified comparison group, scores will be generated using established clinical and demographic variables of prognostic significance used in previous propensity scores for patients who use substances,^{28 29} including age, sex/gender, race/ancestry, residence location, substance use treatment history, comorbidities and severity of use. Given the phased approach to R2R implementation, analyses will be adjusted for year of study enrolment. As some participants may be readmitted to R2R services multiple times, all results will be adjusted for the number of exposures to acute-care R2R services (ie, hospital-based services) during follow-up. Analyses for binary outcomes (eg, 30-day hospital readmission (yes vs no)) will be evaluated using generalised estimating equations with a logit link function to account for the study’s repeated measures clustered within individuals. For time-to-event analyses, multivariable survival models will be constructed using extended Cox models that consider both time-fixed covariates (eg, sex, ethnicity/ancestry) and time-dependent covariates (eg, substance use patterns).^{30 31} We also use parametric survival regression under the Weibull distribution to calculate adjusted HRs if the assumptions are unmet. We will also use marginal structural modelling analyses using inverse probability of treatment weights to assess a causal pathway whereby receipt of in-hospital addiction care through R2R promotes involvement of subsequent ongoing engagement in substance use treatment as ascertained through confidential record linkages.³²

Finally, for the planned ITS analysis for R2R participants who previously also participated in existing BCCSU cohort studies (described in the ‘Proposed comparator groups’ section), analyses will be performed using the restricted maximum likelihood (REML) approach, given the robustness of REML estimates over the different series lengths compared with other estimators (eg, ARIMA).³³,³⁵

Methods to evaluate robustness and address bias

We will use multiple imputations to manage missing data.³⁶ Sensitivity analyses will be undertaken to evaluate the robustness of findings by applying imputation methods (eg, presuming patients lost to follow-up were more likely to experience negative outcomes than those followed). Investigating differences between individuals who do and do not consent to record linkage will be undertaken to identify potential selection biases in the administrative data analyses. Given the risk for false-negative findings in subgroup analyses due to inadequate power, Bayes’s rule will be applied for the design and interpretation of subgroups,³⁷ including a priori selected categorical variables with an established prior probability threshold of 20% for the outcome of interest.

Timeline and follow-up

Figure 3 depicts a timeline for the proposed project. Data collection will occur in parallel to R2R implementation. Participants are intentionally recruited over 5 years to ensure the inclusion of patients across the distinct phases of the R2R roll-out (figure 1). Primary data collection (1-year follow-up of phase 5 recruited participants) will be completed in year 6. Years 6–10 will continue with annual data linkages to facilitate 5-year follow-up.

Figure 3. A visualisation of the recruitment and data collection timelines followed for the Road to Recovery Initiative Research Evaluation. REB, Research Ethics Board.

R2R qualitative study

A mixed-methods study design combines quantitative and qualitative elements in order to strengthen, enhance and expand the study’s findings.³⁸ A sample of cohort participants (n=240) will be recruited to participate in an in depth qualitative interview at baseline (ie, within 30 days of R2R enrolment) and 12-month follow-up. Qualitative data collection will be facilitated by staff trained in qualitative interviewing techniques, which include probing themes that emerge during the interview. A semistructured interview guide has been developed and includes supplemental sections for participants facing the unique challenges described below. Participants will be queried about how they accessed the R2R, treatment goals and motivation, substance use and treatment history, living situation at the time of R2R enrolment, experiences with the R2R and how the R2R could be improved. Participants will be asked to describe if and how R2R access impacted their substance use, treatment, health and social trajectories at baseline and follow-up.

Qualitative recruitment

A purposeful sample will be invited to participate in the qualitative portion of the study. Using a purposive sampling strategy, we can identify cohort participants who reflect a diverse range of substance use treatment experiences, ensuring groups under-represented in substance use research are interviewed. Recruitment will be stratified by age (eg, youth participants aged 24 years and under), gender identities (eg, cisgender men and women, transgender individuals), prenatal and postpartum status, and ethnicity (eg, Indigenous individuals). Experts have developed supplemental interview guides to ensure the unique experiences of groups often underrepresented in health study outcomes are captured. We will recruit roughly equal numbers of people who have and have not accessed community-based substance use care previously (eg, methadone, buprenorphine).

Sample size estimates

A large sample of 240 participants was selected to safeguard the representation of a diverse range of patient perspectives and experiences over time, ensuring patient perspectives and experiences are adequately captured across the entire implementation phase of R2R (2023–2026) as new services are added and changes are made. It is anticipated that this will be accomplished via the proposed staggered recruitment of five patients per month, including 1-year follow-up interviews, over the planned 5-year qualitative component of the study.

Data collection

Qualitative interviews will be digitally recorded and sent to an external professional transcriptionist for verbatim transcription. Audio files will be transmitted to transcriptionists via a secure file transfer service. Transcripts will be checked for accuracy, and any identifying information will be removed.

Qualitative data interpretation and analysis

Electronic transcripts will be managed using NVivo V.10 software. The study team will develop a codebook using deductive (a priori) and inductive (emergent) reasoning for organisation and categorisation of the dataset. Developed by Braun and Clark,³⁹ the steps of thematic analysis will be followed sequentially to examine emergent findings. For example, deductive codes including ‘past treatment experiences,’ ‘postdischarge health barriers’ and ‘impacts of racism’ allow for broad organisation and data familiarisation. We will develop interpretations about key concepts, recurring, converging and contradictory themes, and develop illustrative examples within the interview data across various points of the R2R’s implementation. Trustworthiness of the data will be ascertained using traditional measures of qualitative assurance such as member checking, interrogation of contradictory ideas and data audits.¹²⁴⁰,⁴³ Member checking will be completed directly with participants from the qualitative study.

Critical to achieving overarching study aims, we will use constant comparative analysis to facilitate an ongoing assessment of the utility of the R2R. The ‘point of integration’ with quantitative findings will occur by analysing results, respectively, in a stepwise process.

Data statement

The statistical code and a deidentified dataset for primary data collected for the R2R cohort (ie, quantitative questionnaires) will be available on reasonable request on study completion. Given the sensitivity of the data collected (eg, criminal justice system involvement, substance use practices and pregnancy or family status) the data governance committee will review all data requests from external institutions. Qualitative datasets will not be made available to protect participant confidentiality.

Patient and public involvement

To ensure the R2R evaluation is undertaken in full collaboration with patient, family and Indigenous partners in a culturally safe and trauma-informed manner, a patient and family advisory committee was established. The committee meets monthly to discuss plans and challenges related to the implementation and operations of the R2R clinical service. Additionally, research team members are involved in these meetings to discuss the proposed evaluation and seek feedback on important aspects of the study, including a selection of important patient outcomes and review of study materials. The advisory committee will be involved in supporting both quantitative and qualitative data interpretation. Advisory committees associated with the existing BCCSU cohort studies will be consulted for all research involving these studies.

Furthermore, an R2R data governance committee is being established to review and support the integration of findings from this study into actionable strategies to adapt services based on the needs of providers and patients, while also reviewing how this data can be used to inform questions relevant to the study objectives. The committee will include members of the R2R investigator and operations teams, persons with lived experience, health authority partners, government (MMHA) and the IWR teams. The governance committee role is distinct from the patient and family advisory committee, where this group will meet specifically to review and discuss data that arises from the cohort study. This is intended to ensure patient perspectives and cultural safety are considered throughout the knowledge generation and mobilisation stages.

Ethics and dissemination

The study was approved by the University of British Columbia PHC Research Ethics Board (REB) in September 2023. Participants recruited into the study provided consent per the standard procedural guidelines required by the REB. The study protocol is reported in accordance with the Standardised Protocol Items Recommendations for Observational Studies (SPIROS) for Observational Study Protocol reporting guidelines.⁴⁴ Results will be published in peer-reviewed journals, presented at national and international scientific conferences and disseminated through regular meetings with policy-makers, individuals with lived and living experience, and academic presentations, and lay media. Findings will be disseminated to key stakeholders through the existing infrastructure of the BCCSU’s Knowledge Translation Team and the regional and provincial health authority communications teams.

Funders had no role in the study’s design nor any study conduct, data collection, management, analysis, and interpretation; and/or preparation, review, or approval of this manuscript.