Syringe service program-based telemedicine linkage to opioid use disorder treatment: the randomized mixed method STAMINA trial

doi:10.21203/rs.3.rs-4752555/v1

Download PDF

Research Article

Syringe service program-based telemedicine linkage to opioid use disorder treatment: the randomized mixed method STAMINA trial

https://doi.org/10.21203/rs.3.rs-4752555/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

The opioid overdose crisis necessitates expanded access to medications for opioid use disorder (MOUD). Treatment programs often struggle to engage individuals at the highest overdose risk. This study, STAMINA, aimed to engage participants at a syringe service program (SSP) and connect them to treatment through a telemedicine platform.

Methods

In a pragmatic mixed methods trial, participants were randomized to either telemedicine (n = 138) or standard referral (n = 136). Baseline and 90-day structured interviews, urine testing, and administrative data from MOUD providers were collected. Additionally, a subset of participants (n = 36) completed semi-structured qualitative interviews. The study took place at two SSP sites in Chicago, IL, involving 274 individuals (78% male, 61% African American, 78% non-Latino, mean age = 48) interested in starting MOUD treatment. Interventions included an immediate telemedicine appointment with transportation to the first in-person MOUD appointment versus standard referral. Participants could choose from methadone, buprenorphine, or extended-release naltrexone.

Results

The primary outcome was measured by attendance at the first in-person appointment, indicating successful treatment linkage. Secondary outcomes included MOUD engagement, retention, and non-MOUD opioid use, assessed through self-report and urinalysis. The key determinant of linkage success was scheduling an in-person appointment within 48 hours of study enrollment (aOR = 3.20, p < .01). Methadone treatment posed significant requirements but showed greater success compared to buprenorphine. Participants reported various reasons for continued illicit opioid use.

Conclusions

Telemedicine alone is insufficient to overcome MOUD access barriers. Differences in MOUD requirements should inform the design of linkage interventions. Additional supports may be necessary to enhance the success of flexible treatments like buprenorphine.

Trial registration

ClinicalTrials.gov (NCT04575324).

Telemedicine

medications for opioid use disorder

syringe service program

embedded qualitative

Opioid overdose is a global crisis, with 83,353 deaths in 2019, 56% in the US [1]. US overdoses increased during COVID-19 due to fentanyl adulteration, with 82,851 opioid-related deaths in 2022 [2]. To reverse this trend, expanding access to medications for opioid use disorder (MOUD) (e.g., methadone, buprenorphine, and naltrexone) is crucial [3]. However, treatment programs struggle to engage those who mistrust formal healthcare [4]. Emergency medicine MOUD linkage interventions engage some of these individuals [5], though typically after an overdose. Alternative interventions that can engage individuals before overdose are needed. The STAMINA (Syringe Service Telemedicine Access for Medication-assisted Intervention through NAvigation) intervention engages people who use opioids in the community, connecting them to MOUD via telemedicine at a local syringe service program (SSP).

The STAMINA intervention engages opioid users in the community, linking them to MOUD via telemedicine at local SSPs, addressing noted barriers related to engagement, trust, and access delays [6]. SSPs provide low-barrier services to reduce drug use risks without requiring formal treatment [7], creating a more comfortable and trusted space for people who use opioids than traditional healthcare [4]. Prior studies show SSP services reduce sexually transmitted infection incidence and other negative outcomes when combined with MOUD and other treatments [8–10]. Although SSP clients often express treatment interest [11–13], initiation ranges from 5%-70% [8, 12, 14–23].

A handful of randomized trials of case management interventions conducted in the US and Sweden have demonstrated mixed SSP-based treatment linkage results using an in-person referral [14, 22, 24], with only one conducted in Boston demonstrating improved outcomes compared to passive referral [22]. Noted weaknesses likely accounting for most prior trials’ poor performance include long waits (5–7 days) between enrollment and treatment initiation [14] and treatment provider counseling mandates [16, 18, 25]. STAMINA attempts to address these weaknesses by utilizing a telemedicine model to immediately link patients who indicate treatment readiness with a telemedicine MOUD provider who does not require additional services.

Telemedicine MOUD models were widely adopted during COVID-19 to accommodate social distancing, facilitated by the rollback of policies against telemedicine-based buprenorphine prescriptions [26]. Since then, observational studies demonstrating high patient satisfaction, increased access, comparable retention, and cost reductions suggest telemedicine MOUD may be as effective as in-person treatment [27, 28]. However, more rigorous evidence is needed.

The primary goal of this study was to test the effectiveness of STAMINA’s telemedicine MOUD linkage approach. Secondary objectives were to understand STAMINA’s potential effect on MOUD engagement and retention and non-prescribed opioid use.

Design

We conducted a mixed-method randomized controlled trial of STAMINA, detailed in a pre-published protocol [29]. The qualitative component gained importance due to unforeseen pandemic-related issues because it can enhance causal evidence by capturing participants’ experiences and identifying specific circumstances impacting outcomes [30, 31]. Figure 1 shows our mixed-method data collection, analysis process, and mixing points. Quantitative and qualitative data were collected separately. After their final quantitative follow-up, a subset of participants completed semi-structured qualitative interviews. Administrative treatment data were collected for all participants, and quantitative analysis was conducted using predefined model parameters. In the first stage of data mixing, qualitative findings informed the selection of additional covariates for an expanded quantitative model. The final data mixing stage occurred during interpretation. We describe methods specific to the main trial and embedded qualitative component separately. The manuscript follows CONSORT reporting guildeines.

-insert Fig. 1 about here-

Main trial

Participants

We recruited 274 participants from two Chicago SSP sites between August 24, 2020, and June 30, 2022, through SSP staff referrals, self-referrals after seeing recruitment materials, and outreach workers. Eligible participants were at least 18 years old, English-speaking, had past-year opioid use disorder (OUD), expressed interest in MOUD treatment, and resided in Cook County, IL. Exclusions included severe opioid withdrawal needing immediate medical referral, current prescribed MOUD, plans to move outside Cook County, impending prison sentence, and inability to understand study procedures. After eligibility screening, research staff used a treatment decision guide to explain medication options. A federally qualified health center (FQHC) with onsite MOUD services—including newly implemented methadone dosing— provided treatment assessments and referrals.

Procedures

Study procedures for SSP and community-recruited participants were identical. After informed consent, research assistants conducted urine screening and structured interviews, repeated at 90-day follow-up. Participants unable to return for follow-up completed the interview by phone without urine screening. Incentives were $50 at baseline and $35 at follow-up. Trained staff retrieved post-enrollment treatment data from electronic health records using enrollment-assigned medical record numbers, aliases, names, and demographics for matching. The county medical examiner provided mortality data via deterministic linkage based on name, demographics, and social security number. The Chestnut Health System’s Institutional Review Board approved all procedures.

Randomization

The project manager used a computer-generated randomization schedule to assign participants to either the telemedicine or control arm, stratified by enrollment location and methadone preference to ensure balanced arms and limited care variance. Arm assignments were revealed to research assistants after data collection. This was an open trial with no concealment after baseline data collection.

Interventions

STAMINA has three components: (a) a brief MOUD interest assessment; (b) an immediate telemedicine appointment; and (c) one-time rideshare assistance. During the telemedicine visit, the provider discusses medication preferences, OUD use, and histories, helping undecided patients choose an appropriate MOUD. For buprenorphine or naltrexone, the provider writes an induction prescription for immediate filling at the FQHC or a chosen pharmacy and schedules an in-person follow-up within 5 business days (10 for naltrexone), and the participant receives an appointment card and home induction instructions. Participants with complex health conditions preventing virtual prescribing are scheduled to see a provider, ideally within one business day. Staff instruct patients choosing methadone to go to the FQHC’s urgent care door for their initial dose any weekday before 3pm (including the same day if possible). The provider refers participants with transportation difficulties to a more accessible methadone provider, aiming to schedule an appointment within one business day. Transportation varies by medication: buprenorphine or naltrexone patients receive immediate rideshare assistance, while methadone patients get pre-scheduled rides and reminder calls. Fifty-seven percent of buprenorphine or naltrexone and 31% of methadone participants accepted rideshares.

Control participants also received the MOUD interest assessment because it was needed to determine eligiblity. Staff referred those interested in buprenorphine or naltrexone for an in-person FQHC intake and those interested in methadone to the FQHC or a more accessible provider. Appointments were arranged based on availability, with an attempt to schedule within two business days and provided two bus passes.

Supplemental Fig. 1 represents the planned intervention pathways for both study arms.

Measures

Primary outcome. In-person treatment linkage (yes/no), defined by attending one treatment appointment within 14 days of enrollment, per national care standards [32], as determined by the electronic health record. We chose 'in-person' over 'medication-based' linkage because FQHC policies required a post-induction in-person appointment for continuing buprenorphine, ensuring all participants needed an in-person visit before engagement and retention phases.

Secondary outcomes. Secondary outcomes included: (1) MOUD treatment engagement: attending at least two OUD-related treatment appointments within 34 days of enrollment [32]; (2) Treatment retention: not exceeding 14 consecutive days without medication during the observation window; (3) past 30-day self-reported non-MOUD opioid use: assessed at follow-up using the National Survey on Drug Use and Health questions; and (4) urinalysis-detected non-MOUD opioid use at follow-up.

Covariates. We used sixteen covariates in our models. Interviews captured seven demographics: age (continuous), gender (male/female), race (African American/other), ethnicity (Latino/other), employment status (unemployed/other), insurance status (insured/uninsured), and housing (stable/unstable). We also recorded if they were an SSP client (yes/no) and the number of lifetime overdoses. Severe psychological distress (yes/no) was assessed using the Psychological Distress Scale (K6) with a threshold of > 12 (max 24 points) [33]. Opioid use severity was measured using a DSM-5 OUD screen (0–11) [34], and withdrawal severity (0–48) using the Clinical Opiate Withdrawal Scale [35]. Variables related to study involvement included: enrollment arm (telemedicine/control); block randomization (open to any MOUD/methadone only), enrollment site (1/2). We included whether the participant had an in-person appointment scheduled within 48 hours of enrollment (yes/no) based on initial qualitative interviews that revealed barriers some participants faced in attending their first in-person appointment.

Statistical methods/analytics

We conducted an intention-to-treat analysis using Stata v17.0 [36]. The primary question was whether immediate telemedicine consultation improved MOUD linkage more than standard referral. Preliminary mixed effects binary logistic regressions showed little variation by study site (var = .007, sd = .048). A likelihood ratio test comparing fixed effects to a random effects model with site as a random effect indicated no improvement (chibar2 = .03, p = .428). Therefore, we used generalized estimating equation (GEE) binary logistic regression models with study site as the clustering variable. GEE models provided population-averaged fixed effect estimates for included covariates [37], assuming an exchangeable correlation structure, robust standard errors, and Stata’s ‘nmp’ method to correct for the small number of clusters [38, 39].

Sample size/statistical power calculations. With no prior research to inform an expected effect size, we used an odds ratio of 2.0 to represent a minimum clinically meaningful effect. A priori power calculations determined that n = 273 was needed to detect a meaningful effect for a binary predictor (the intervention effect) in a univariable logistic regression with 80% power [29]. Factoring in an expected 20% attrition, a sample of n = 350 was required.

Primary outcome. We included two covariates in the initial analysis: treatment condition and randomization block. Treatment linkage was definitively determined for 206 (75.2%) participants. For others, we assumed no treatment linkage, assigning a value of zero under the assumption of missing not at random (MNAR). We then analyzed the primary outcome regression model with the protocol-specified reduced predictor set using complete case analysis for the covariates, assuming the outcome data were MNAR.

We conducted GEE binary logistic regressions to screen for additional covariates, retaining treatment condition and randomization block. Covariates with p ≤ .20 were included in the final multivariable GEE model. Preliminary analyses showed strong associations between 48-hour appointment scheduling and both treatment linkage (aOR = 3.09, p = .002) and arm assignment (26.8% telemedicine vs. 55% control, χ² = 22.8, p < .001), so we included an interaction term. Only three covariates had missing data (insurance [3.7%], opioid use severity [1.1%], and withdrawal severity [1.8%]), so we ran the multivariable GEE model using complete case analyses. We reran the model with multiple imputations by chained equations (MICE) using 75 imputations to assess the impact of excluding missing cases [40]. Comparative estimates assuming data are missing at random (MAR) are provided alongside complete case analyses.

Secondary outcomes. We used the same approach for modeling secondary outcomes. Based on preliminary analyses, we included any covariate significant at p ≤ .20 in the multivariable regression models. We also reran all secondary outcome regression models using complete case analyses to assess the impact of the MNAR assumption on parameter estimates.

Explanation of protocol modifications

The procedures included protocol modifications due to COVID-19, shifting the study toward a pragmatic approach and elevating the importance of qualitative data. Due to reduced client volume at our SSP and others worldwide (56), we did not recruit at a third planned site with low referrals, instead increasing recruitment days at two higher volume locations. We added outreach-based recruitment and increased baseline interview incentives from $25 to $50 to improve early recruitment rates. The originally planned 6-month interview was eliminated due to social distancing constraints. The FQHC partner's methadone dosing implementation near recruitment sites was delayed by the pandemic, leading us to refer participants to alternate methadone providers. This increased missing data, so we used multiple imputation as described above.

Embedded qualitative component

The embedded qualitative component provided process data to complement the quantitative results and inform covariate selection for the expanded regression model. We invited a convenience sample of 36 trial participants for audio-recorded semi-structured qualitative interviews during their 3-month follow-up. Any participant who completed a follow-up when a trained qualitative interviewer was available was eligible, and all who interacted with the researcher agreed to participate. Two experienced qualitative researchers led data collection and analysis. Participants received $35 for their time. Researchers used a combined deductive-inductive coding process, starting with primary research questions and the interview schedule as an organizing structure [41], then generating subcodes inductively. Saturation was defined as the point at which no new insights emerged [42]. Both analysts applied the expanded code list to a subset of transcripts to establish intercoder reliability (kappa = .71). They then used the coded data to develop higher-level themes, remaining open to new configurations of knowledge and experiences. Analysis was facilitated using MAXQDA data analysis software [43].

Quantitative results

Descriptive statistics

Figure 2 displays the study’s CONSORT diagram. Supplemental figures 2 and 3 display participants’ treatment engagement throughout the observation window.

-Insert Figure 2 about here -

Tables 1 and 2 provide descriptive statistics for baseline characteristics, appointment scheduling, and outcomes. Enrollment was evenly split between sites, with most participants recruited via street outreach. Most were male, African American, non-Latino, unemployed, stably housed, and insured, with an average age of 48. One-third had severe psychological distress, and participants had an average of just over four lifetime overdoses. Baseline scores indicated high OUD severity and mild withdrawal. Two factors significantly differed by arm: telemedicine participants were less likely to have an appointment within 48 hours of enrollment, and fewer completed their initial appointment compared to controls. Four participants died (two telemedicine, two control), with two deaths being opioid-related.

-Insert Table 1 about here –

-Insert Table 2 about here –

Primary outcome

Table 3 presents treatment linkage results for base and covariate-adjusted models. Due to similar results, we discuss the more conservative MNAR results. In the base model, telemedicine participants had significantly lower treatment linkage odds, while methadone preference had a small positive effect. These results held in the adjusted model. Insurance moderately increased linkage odds, and SSP client status slightly increased odds. Being Latino and having unstable housing lowered linkage odds. The interaction term for 48-hour scheduling and treatment arm was significant, indicating a more pronounced effect for telemedicine participants, accounting for much of the observed difference between arms (Figure 3).

-Insert Table 3 about here-

-Insert Figure 3 about here-

Secondary outcomes

Table 4 presents secondary outcome MAR assumption results, supported by the degree of missingness and imputation results (MNAR estimation results are in Supplemental Table 1). The methadone preference block had higher treatment engagement and retention odds in the base model, but this only held for engagement after covariate adjustments. Factors significantly associated with increased odds of both retention and engagement included 48-hour appointment scheduling (indicating effects carried over from initial linkage), the interaction between scheduling and arm, and being older. SSP clients had higher engagement odds, while Latino participants had lower engagement odds.

-Insert Table 4 about here-

Arm assignment significantly affected urine-detected use, with telemedicine participants having lower odds. The methadone preference block was associated with higher urine detection odds in the base model, but this did not hold in the expanded model. The number of prior lifetime overdoses was significantly associated with lower odds of both opioid use measures. For covariates associated with only one use measure, lower self-reported opioid use odds were linked to being female and in the other racial group, while SSP clients had higher odds. Lower urine detection odds were linked to higher baseline opioid withdrawal scores and being employed, while being older was associated with higher detection odds. The interaction between 48-hour scheduling and arm indicated increased odds for self-reported opioid use but not urine detection.

Qualitative findings

As with the main trial sample, qualitative interview participants were mostly middle-aged (mean = 51 years), male (69%), and African American (72%). All qualitative participants chose methadone or buprenorphine (67% vs. 33%) at enrollment, and 21 (58%) were in the control condition. Interviews revealed initiation and retention patterns by study arm and medication. All telemedicine buprenorphine participants started and later stopped the medication. Thirty-three percent of control buprenorphine participants followed this same pattern, while the rest never started the medication. For methadone participants, differences by arm were less pronounced in terms of never starting (33% telemedicine vs. 27% control), continuing (44% telemedicine vs. 40% control), and starting and stopping (22% telemedicine vs. 33% control). (Similar information for the entire sample is in supplemental figures 2 and 3.)

Table 5 displays qualitative themes with example quotes. Experiences were mostly driven by the medication received vs. arm assignment. Barriers discussed for both medications included difficulties contacting the provider, insurance, and transportation. Only methadone participants discussed insurance coverage challenges. However, there was an overlap in insurance and transportation themes for both medications, with some participants mentioning difficulties in insurance covering transportation despite their benefit eligibility. Lack of vehicle access and distance were also transportation barriers for both medications.

-insert Table 5 about here-

Specific to methadone, daily in-person dosing (often required during the titration period or due to housing instability) and transportation challenges were primary engagement and retention hurdles. Methadone participants also cited experiencing trauma and having competing priorities (mostly family responsibilities) as making daily dosing difficult. Negative interactions with treatment staff also led some participants to consider discontinuing methadone.

Regarding medication-related issues, participants who started methadone mentioned discontinuing because their doses were too low. In contrast, those who started buprenorphine mostly stopped during induction because they disliked the taste or felt sick. Also, some participants who regularly used fentanyl believed the buprenorphine home induction protocol did not work because it took too long to enter the necessary withdrawal period. The only differing theme between arms was that telemedicine buprenorphine participants expressed confusion about post-induction steps. This affected their ability to connect with in-person treatment, as some were unaware of the need for an in-person appointment or how to contact their provider.

Participants cited several reasons for continued illicit opioid use, regardless of medication. Unsurprisingly, those who never started or stopped treatment continued use. Others used MOUDs prescribed through the study or obtained illicitly to avoid withdrawal symptoms when unable to purchase heroin/fentanyl or when methadone dosing was inadequate. Finally, some used MOUDs to taper off heroin/fentanyl gradually, with the goal of eventually stopping illicit and prescribed opioid (i.e., MOUD) use.

Supplemental Table 2 synthesizes quantitative and qualitative findings, thus displaying metainferences that guide the following discussion.

Overall, 30% of participants were linked with treatment. Telemedicine was less effective than standard referral, but 48-hour in-person appointment scheduling had a greater impact on linkage success than arm assignment. This finding aligns with studies indicating shorter appointment wait times improve linkage [44, 45]. The interaction term in our models indicates telemedicine participants benefited more from 48-hour scheduling. Qualitative findings suggest this was due to telemedicine participants opting for buprenorphine who had to wait 5–7 days for their in-person appointment, while control participants could be scheduled within two business days. Ten telemedicine participants with medical complications were scheduled to see a provider within 24 hours. The positive effect of 48-hour scheduling translated into better treatment and engagement outcomes. Despite lower linkage, the telemedicine arm had higher engagement at 34 days (57% vs. 44%; see Supplemental Fig. 2), suggesting positive post-linkage treatment experiences. Qualitative data revealed the induction period presented barriers, leading most buprenorphine patients to discontinue medication or feel confused about next steps, consistent with research showing high post-induction discontinuation [46]. This outcome could likely have been avoided if telemedicine participants were provided virtual support during induction.

Prior studies have reported negative buprenorphine experiences such as an unpleasant taste and precipitated withdrawal symptoms [47, 48]. Home induction protocols, developed primarily for heroin use, may increase the risk of precipitated withdrawal for those with higher opioid tolerance due to fentanyl use [49]. Participant confusion regarding post-induction steps is not well-documented. Non-randomized studies of telemedicine buprenorphine linkage in SSP settings have focused on buprenorphine and included components likely to improve treatment outcomes, such as not requiring an in-person follow-up for continued prescribing and providing patient navigation [50–52]. Not requiring an in-person appointment for medication continuation aligns with US emergency rules that allowed pandemic telemedicine buprenorphine prescribing. These prescribing flexibilities should be considered as policymakers contemplate making these rules, currently set to expire in 2024 [53], more permanent. However, our findings suggest additional policies to support navigation and support services would help overcome barriers after initial telemedicine engagement [54]. These services can likely be delivered virtually with minimal in-person appointments [50–52].

Aligning with previous research [55], our findings suggest participants preferring methadone experienced greater success, as they were more likely to be linked and engaged. This is noteworthy given the many complaints about methadone treatment shared in qualitative interviews, similar to those reported in other studies [56, 57]. However, our findings suggest the explicit and fixed methadone requirements may provide a clear treatment roadmap, despite daily dosing hassles. Positive outcomes may also be partially related to emergency pandemic regulations that made methadone more flexible for some patients [58]. In contrast, the more flexible and complex buprenorphine models appeared to confuse treatment expectations and course.

Better linkage and engagement outcomes for SSP clients, compared to community-recruited participants, may be due to their pre-established relationship with the organization [4, 59–62], as assumed in the intervention design [29]. SSP clients were also more likely to self-report non-MOUD opioid use, possibly due to trust built through previous interactions. Having insurance improved linkage, a known MOUD treatment facilitator [54, 63], particularly for those with earnings above Medicaid thresholds. While FQHCs and methadone programs offer sliding scale fees, they can still pose difficulties for patients. Housing instability disrupts MOUD treatment continuity [54]. Latinos face significant addiction treatment access disparities [64], impacting linkage and engagement. Younger individuals are at higher risk of MOUD discontinuation [65], aligning with our finding that older participants were more likely to be retained.

Differences in self-reported non-MOUD opioid use compared to urine detection could be due to drug use variation between those available for on-site urine tests and those not. However, the more likely reason is the self-report window (30 days) encompassed more days than the urine detection window (3 days). Regardless, the key takeaway is most participants continued non-prescribed opioid use. Due to high discordance and difficulties verifying self-reported use, we focus the rest of the discussion on urine detection outcomes.

Employed individuals' lower urine detection likelihood may be due to employment being linked with better treatment outcomes and motivation to maintain sobriety [66]. The reason for the small association between higher withdrawal symptoms and lower 90-day urine detection is uncertain. It may be because starting buprenorphine home induction requires withdrawal symptoms, leading some with higher symptoms to have better experiences and continued treatment. Although older individuals had slightly higher retention odds, this did not translate into lower urine detection. The literature on MOUD treatment among older individuals is limited; their continued non-prescribed opioid use could stem from challenges breaking long-term habits or higher chronic pain levels [67]. Continued non-prescribed substance use is common in MOUD treatment, as our qualitative interviews and prior research show. While prior work suggests using non-prescribed MOUDs as self-medication [48, 68], our qualitative findings reveal reasons for using prescribed MOUDs alongside non-prescribed opioids, including more personal control over recovery, managing withdrawal, and compensating for inadequate methadone dosing—a noted issue in other studies [69]. Understanding these reasons can help treatment programs assist patients better.

Limitations

This study's pragmatic approach limited our control over the intervention and our FQHC provider’s treatment protocols, affecting internal validity [70]. However, operating under real-world conditions improves generalizability [70]. Specific factors associated with our FQHC partner, particularly on-site methadone treatment, may not be found elsewhere. Providing participants with a choice among all three MOUD options introduced heterogeneity, aligning better with patient choice and shared decision-making models in recent SSP telemedicine linkage studies [50–52]. We did not reach the original target sample size due to pandemic-related challenges. However, sample needs were determined expecting 20% attrition. Through administrative data and multiple imputations, we assessed our primary outcome including all 274 participants, the number needed based on prior power analyses. Our sample was larger than similar studies (n = 274 vs. <120) [50–52].

The STAMINA study is the only randomized trial of a telemedicine-based MOUD linkage intervention completed of which we are aware. Our findings clearly indicate that initial telemedicine linkage alone is insufficient to address the access barriers faced by high-risk individuals. Different MOUDs have distinct initiation processes that must be considered in intervention design. Specifically, buprenorphine’s flexibility likely requires additional virtual navigation or wraparound services during linkage to achieve desired outcomes. Recognizing the distinct medication and provider-based differences in MOUD initiation is crucial to ensuring the purported benefits of telemedicine are fully realized. These insights were facilitated by the integration of qualitative data into our statistical models and final interpretations, demonstrating the value of a mixed-methods approach in uncovering comprehensive and actionable findings.

COVID-19

Coronavirus disease 2019

FQHC

Federally Qualified Health Center

GEE

Generalized estimating equation

MNAR

Missing not at random

MOUD

Medications for opioid use disorder

OUD

Opioid use disorder

SSP

Syringe service program

STAMINA

Syringe Service Telemedicine Access for Medication-assisted Intervention through Navigation

Ethics approval and consent to participate

All procedures involving human subjects were approved by the Institutional Review Board of Chestnut Health Systems (#1138-0420), and a reliance agreement was used to establish a single IRB process with the University of Illinois Chicago. Written informed consent was obtained from all participants prior to trial enrollment.

Consent for publication

Not applicable

Availability of data and materials

Due to the sensitivity of information related to substance use and treatment, the dataset is available only upon request and with proper data use agreements in place. All materials developed for the study are available upon request.

Competing interests

None

Funding

This study was funded by Arnold Ventures (grant number not applicable). Opinions are those of the authors and do not necessarily reflect those of the funder.

Authors’ contributions

Watson: Conceptualization, Methodology, Formal Analysis, Investigation, Writing - Original Draft, Funding Acquisition; Franceschini: Formal Analysis, Investigation, Writing - Original Draft; Swartz: Conceptualization, Methodology, Formal Analysis, Writing - Review & Editing, Funding Acquisition; Dewey: Formal Analysis, Investigation; Mackesy-Amiti: Methodology, Formal Analysis, Writing - Review & Editing; Taylor: Data Curation, Project Administration; Staton: Formal Analysis, Investigation; Zhao: Formal Analysis; Messmer: Writing - Review & Editing; Jimenez: Writing - Review & Editing; Gastala: Conceptualization, Data Curation, Writing - Review & Editing

Acknowledgements

Kim Erwin and Adrian Politzer assisted with the design of study protocols and participant-facing materials. The Cook County Medical Examiner’s Office provided mortality data for this study. Mona Stivers provided detailed pre-submission editing services. Copies of de-identified data are available from the first author upon request.

Institute for Health Metrics and Evaluation. Global Burden of Disease Study 2019 Data Resources: GBD Results Tools [Internet]. 2024 [cited 2024 Mar 13]. Available from: https://www.healthdata.org/data-tools-practices
National Center for Health Statistics. Products - Vital Statistics Rapid Release - Provisional Drug Overdose Data [Internet]. 2024 [cited 2024 Jun 21]. Available from: https://www.cdc.gov/nchs/nvss/vsrr/drug-overdose-data.htm
Renner JA, Crawford MB. Medication for the treatment of opioid use disorder. In: Kelly JF, Wakeman SE, editors. Treating Opioid Addiction [Internet]. New York: Springer International Publishing; 2019 [cited 2021 Jul 16]. p. 89–107. Available from: https://doi.org/10.1007/978-3-030-16257-3_5
Treloar C, Rance J, Yates K, Mao L. Trust and people who inject drugs: The perspectives of clients and staff of Needle Syringe Programs. Int J Drug Policy. 2016;27:138–45.
Hawk K, D’Onofrio G. Emergency department screening and interventions for substance use disorders. Addict Sci Clin Pract. 2018;13:18.
Cernasev A, Hohmeier KC, Frederick K, Jasmin H, Gatwood J. A systematic literature review of patient perspectives of barriers and facilitators to access, adherence, stigma, and persistence to treatment for substance use disorder. Explor Res Clin Soc Pharm. 2021;2:100029.
Ashford RD, Curtis B, Brown AM. Peer-delivered harm reduction and recovery support services: initial evaluation from a hybrid recovery community drop-in center and syringe exchange program. Harm Reduct J. 2018;15:52.
Hagan H, McGough JP, Thiede H, Hopkins S, Duchin J, Alexander ER. Reduced injection frequency and increased entry and retention in drug treatment associated with needle-exchange participation in Seattle drug injectors. J Subst Abuse Treat. 2000;19:247–52.
Platt L, Minozzi S, Reed J, Vickerman P, Hagan H, French C, et al. Needle and syringe programmes and opioid substitution therapy for preventing HCV transmission among people who inject drugs: findings from a Cochrane Review and meta-analysis. Addiction. 2018;113:545–63.
Kidorf M, King VL, Pierce J, Kolodner K, Brooner RK. Benefits of concurrent syringe exchange and substance abuse treatment participation. J Subst Abuse Treat. 2011;40:265–71.
Kidorf M, King VL. Expanding the public health benefits of syringe exchange programs. Can J Psychiatry. 2008;53:487–95.
Heimer R. Can syringe exchange serve as a conduit to substance abuse treatment? J Subst Abuse Treat. 1998;15:183–91.
Fox AD, Sohler NL, Frost T, Lopez C, Cunningham CO. Development and evaluation of a community-based buprenorphine treatment intervention. Harm Reduct J. 2017;14:23.
Bråbäck M, Nilsson S, Isendahl P, Troberg K, Brådvik L, Håkansson A. Malmö Treatment Referral and Intervention Study (MATRIS)—effective referral from syringe exchange to treatment for heroin dependence: a pilot randomized controlled trial. Addiction. 2016;111:866–73.
Henderson LA, Vlahov D, Celentano DD, Strathdee SA. Readiness for cessation of drug use among recent attenders and nonattenders of a needle exchange program. JAIDS Journal of Acquired Immune Deficiency Syndromes. 2003;32:229–37.
Kidorf M, King VL, Neufeld K, Peirce J, Kolodner K, Brooner RK. Improving substance abuse treatment enrollment in community syringe exchangers. Addiction. 2009;104:786–95.
Kidorf M, Disney E, King V, Kolodner K, Beilenson P, Brooner RK. Challenges in motivating treatment enrollment in community syringe exchange participants. Journal of Urban Health: Bulletin of the New York Academy of Medicine. 2005;82:456–67.
Kidorf M, King VL, Gandotra N, Kolodner K, Brooner RK. Improving treatment enrollment and re-enrollment rates of syringe exchangers: 12-Month outcomes. Drug Alcohol Depend. 2012;124:162–6.
Kuo I, Brady J, Butler C, Schwartz R, Brooner R, Vlahov D, et al. Feasibility of referring drug users from a needle exchange program into an addiction treatment program: Experience with a mobile treatment van and LAAM maintenance. J Subst Abuse Treat. 2003;24:67–74.
Lloyd JJ, Ricketts EP, Strathdee SA, Cornelius LJ, Bishai D, Huettner S, et al. Social contextual factors associated with entry into opiate agonist treatment among injection drug users. Am J Drug Alcohol Abuse. 2005;31:555–70.
Riley ED, Safaeian M, Strathdee SA, Brooner RK, Beilenson P, Vlahov D. Drug user treatment referrals and entry among participants of a needle exchange program. Substance Use & Misuse. 2002;37:1869–86.
Strathdee SA, Ricketts EP, Huettner S, Cornelius L, Bishai D, Havens JR, et al. Facilitating entry into drug treatment among injection drug users referred from a needle exchange program: Results from a community-based behavioral intervention trial. Drug Alcohol Depend. 2006;83:225–32.
Havens JR, Lofwall MR, Frost SDW, Oser CB, Leukefeld CG, Crosby RA. Individual and network factors associated with prevalent hepatitis C infection among rural Appalachian injection drug users. Am J Public Health. 2013;103:e44–52.
Havens JR, Latkin CA, Pu M, Cornelius LJ, Bishai D, Huettner S, et al. Predictors of opiate agonist treatment retention among injection drug users referred from a needle exchange program. J Subst Abuse Treat. 2009;36:306–12.
Kidorf M, Brooner RK, Leoutsakos J-M, Peirce J. Treatment initiation strategies for syringe exchange referrals to methadone maintenance: A randomized clinical trial. Drug Alcohol Depend. 2018;187:343–50.
Lambdin BH, Bluthenthal RN, Tookes HE, Wenger L, Morris T, LaKosky P, et al. Buprenorphine implementation at syringe service programs following waiver of the Ryan Haight Act in the United States. Drug and Alcohol Dependence. 2022;237:109504.
Chan B, Freeman M, Ayers C, Korthuis PT, Paynter R, Kondo K, et al. A systematic review and meta-analysis of medications for stimulant use disorders in patients with co-occurring opioid use disorders. Drug Alcohol Depend. 2020;216:108193.
Guillen AG, Reddy M, Saadat S, Chakravarthy B. Utilization of Telehealth Solutions for Patients with Opioid Use Disorder Using Buprenorphine: A Scoping Review. Telemedicine and e-Health [Internet]. 2021 [cited 2022 Mar 8]; Available from: https://www.liebertpub.com/doi/full/10.1089/tmj.2021.0308
Watson DP, Swartz JA, Robison-Taylor L, Mackesy-Amiti ME, Erwin K, Gastala N, et al. Syringe service program-based telemedicine linkage to opioid use disorder treatment: protocol for the STAMINA randomized control trial. BMC Public Health. 2021;21:630.
Edmunds JA, Gicheva D, Thrift B, Hull M. Using Mixed Methods to Explore Variations in Impact Within RCTs: The Case of Project COMPASS. J Mix Methods Res. 2022;16:478–99.
Johnson RB, Russo F, Schoonenboom J. Causation in Mixed Methods Research: The Meeting of Philosophy, Science, and Practice. J Mix Methods Res. 2019;13:143–62.
National Committee on Quality Assurance. Initiation and Engagement of Substance Use Disorder Treatment [Internet]. NCQA. 2024 [cited 2024 Feb 21]. Available from: https://www.ncqa.org/hedis/measures/initiation-and-engagement-of-substance-use-disorder-treatment/
Kessler RC, Green JG, Gruber MJ, Sampson NA, Bromet E, Cuitan M, et al. Screening for serious mental illness in the general population with the K6 screening scale: results from the WHO World Mental Health (WMH) survey initiative. Int J Methods Psychiatr Res. 2010;19 Suppl 1:4–22.
American Psychiatric Association. Diagnostic and statistical manual of mental disorders [Internet]. Fifth Edition. American Psychiatric Association; 2013 [cited 2021 Jul 16]. Available from: http://psychiatryonline.org/doi/book/10.1176/appi.books.9780890425596
Wesson D, Ling W. The clinical opiate withdrawal scale (COWS). Journal of psychoactive drugs. 2003;35:253–9.
StataCorp. Stata Statistical Software: Release 17. College Station, Tx: StataCorp LLC; 2021.
Zorn CJW. Generalized Estimating Equation Models for Correlated Data: A Review with Applications. Am J Political Sci. 2001;45:470–90.
NIH Pragmatic Trials Collaboratory. Small-Sample Robust Variance Correction for Generalized Estimating Equations for Use in Cluster Randomized Clinical Trials [Internet]. 2015. Available from: https://dcricollab.dcri.duke.edu/sites/NIHKR/KR/Variance-correction-for-GEE_V1.R0.pdf
Zorn C. Comparing GEE and robust standard errors for conditionally dependent data. Political Research Quarterly. 2006;59:329–41.
White IR, Royston P, Wood AM. Multiple imputation using chained equations: Issues and guidance for practice. Stat Med. 2011;30:377–99.
Proudfoot K. Inductive/Deductive Hybrid Thematic Analysis in Mixed Methods Research. J Mix Methods Res. 2022;15586898221126816.
Saunders B, Sim J, Kingstone T, Baker S, Waterfield J, Bartlam B, et al. Saturation in qualitative research: exploring its conceptualization and operationalization. Qual Quant. 2018;52:1893–907.
MAXQDA: Qualitative Data Analysis Software | Windows & Mac [Internet]. MAXQDA - The Art of Data Analysis. [cited 2018 Apr 2]. Available from: https://www.maxqda.com/
Roy PJ, Price R, Choi S, Weinstein ZM, Bernstein E, Cunningham CO, et al. Shorter outpatient wait-times for buprenorphine are associated with linkage to care post-hospital discharge. Drug Alcohol Depend. 2021;224:108703.
Roy PJ, Choi S, Bernstein E, Walley AY. Appointment wait-times and arrival for patients at a low-barrier access addiction clinic. J Subs Abuse Treat. 2020;114:108011.
Zhang P, Tossone K, Ashmead R, Bickert T, Bailey E, Doogan NJ, et al. Examining differences in retention on medication for opioid use disorder: An analysis of Ohio Medicaid data. J Subs Abuse Treat. 2022;136:108686.
Muthulingam D, Bia J, Madden LM, Farnum SO, Barry DT, Altice FL. Using nominal group technique to identify barriers, facilitators, and preferences among patients seeking treatment for opioid use disorder: A needs assessment for decision making support. J Subs Abuse Treat. 2019;100:18–28.
Sud A, Salamanca-Buentello F, Buchman DZ, Sabioni P, Majid U. Beyond harm-producing versus harm-reducing: A qualitative meta-synthesis of people who use drugs’ perspectives of and experiences with the extramedical use and diversion of buprenorphine. J Subs Abuse Treat. 2022;135:108651.
Srivastava A, Kahan M, Leece P, McAndrew A. Buprenorphine unobserved “home” induction: a survey of Ontario’s addiction physicians. Addict Sci Clin Pract. 2019;14:1–8.
Lambdin BH, Kan D, Kral AH. Improving equity and access to buprenorphine treatment through telemedicine at syringe services programs. Subst Abuse Treat Prev Policy. 2022;17:51.
Jakubowski A, Norton BL, Hayes BT, Gibson BE, Fitzsimmons C, Stern LS, et al. Low-threshold buprenorphine treatment in a syringe services program: program description and outcomes. J Addict Med. 2022;16:447–53.
Suarez E, Bartholomew TS, Plesons M, Ciraldo K, Ostrer L, Serota DP, et al. Adaptation of the Tele-Harm Reduction intervention to promote initiation and retention in buprenorphine treatment among people who inject drugs: a retrospective cohort study. Annals of Medicine. 2023;55:733–43.
Knopf A. DEA extends telemedicine for controlled substances until end of 2024. Alcohol Drug Abuse Week. 2023;35:1–4.
Simpson SA, Hordes M, Blum J, Rinehart D, Al-Tayyib A. Barriers to Engagement in Opioid Use Disorder Treatment After Buprenorphine Induction. J Addict Med. 2022;16:479–82.
Degenhardt L, Clark B, Macpherson G, Leppan O, Nielsen S, Zahra E, et al. Buprenorphine versus methadone for the treatment of opioid dependence: a systematic review and meta-analysis of randomised and observational studies. The Lancet Psychiatry. 2023;10:386–402.
Frank D, Mateu-Gelabert P, Perlman DC, Walters SM, Curran L, Guarino H. “It’s like ‘liquid handcuffs”: The effects of take-home dosing policies on Methadone Maintenance Treatment (MMT) patients’ lives. Harm Reduction Journal. 2021;18:88.
Pasman E, Kollin R, Broman M, Lee G, Agius E, Lister JJ, et al. Cumulative barriers to retention in methadone treatment among adults from rural and small urban communities. Addiction Science & Clinical Practice. 2022;17:35.
Nunes EV, Levin FR, Reilly MP, El-Bassel N. Medication treatment for opioid use disorder in the age of COVID-19: Can new regulations modify the opioid cascade? J Subs Abuse Treat. 2021;122:108196.
Allen ST, Grieb SM, O’Rourke A, Yoder R, Planchet E, White RH, et al. Understanding the public health consequences of suspending a rural syringe services program: a qualitative study of the experiences of people who inject drugs. Harm Reduction Journal. 2019;16:1–10.
Treloar C, Newland J, Maher L. A qualitative study trialling the acceptability of new hepatitis C prevention messages for people who inject drugs: Symbiotic messages, pleasure and conditional interpretations. Harm Reduction Journal. 2015;12:1–9.
Macneil J, Pauly B. Needle exchange as a safe haven in an unsafe world. Drug and Alcohol Review. 2011;30:26–32.
I. Cesnales N, Witt M, K. Miles B. Syringe quantity limitations: Syringe exchange program participant health behavior and satisfaction. Journal of Social Work Practice in the Addictions. 2019;19:400–13.
Sharma A, Kelly SM, Mitchell SG, Gryczynski J, O’Grady KE, Schwartz RP. Update on Barriers to Pharmacotherapy for Opioid Use Disorders. Curr Psychiatry Rep. 2017;19:35.
Stahler GJ, Mennis J, Baron DA. Racial/ethnic disparities in the use of medications for opioid use disorder (MOUD) and their effects on residential drug treatment outcomes in the US. Drug Alcohol Depend. 2021;226:108849.
Krawczyk N, Williams AR, Saloner B, Cerdá M. Who stays in medication treatment for opioid use disorder? A national study of outpatient specialty treatment settings. J Subst Abuse Treat. 2021;126:108329.
Sahker E, Ali SR, Arndt S. Employment recovery capital in the treatment of substance use disorders: Six-month follow-up observations. Drug Alcohol Depend. 2019;205:107624.
Dufort A, Samaan Z. Problematic Opioid Use Among Older Adults: Epidemiology, Adverse Outcomes and Treatment Considerations. Drugs Aging. 2021;38:1043–53.
Hayes BT, Jakubowski A, Fitzsimmons C, Garcia B, Ramirez F, Fox AD. “The Doctor Says You Cannot Have [Buprenorphine]” Autonomy and Use of Prescribed or Non-Prescribed Buprenorphine. Substance Use & Misuse. 2021;56:1137–43.
Reimer J, Boniakowski E, Bachner C, Weber B, Tietje W, Verthein U, et al. When higher doses in opioid replacement treatment are still inadequate – association to multidimensional illness severity: a cohort study. Subst Abuse Treat Prev Policy. 2014;9:13.
Ford I, Norrie J. Pragmatic Trials. New England Journal of Medicine. 2016;375:454–63.

Tables 1 to 5 are available in the Supplementary Files section

No competing interests reported.

Download PDF

Reviews received at journal
30 Sep, 2024
Reviewers agreed at journal
27 Aug, 2024
Reviewers invited by journal
21 Aug, 2024
Editor assigned by journal
19 Aug, 2024
Submission checks completed at journal
16 Aug, 2024
First submitted to journal
16 Jul, 2024

You are reading this latest preprint version

Syringe service program-based telemedicine linkage to opioid use disorder treatment: the randomized mixed method STAMINA trial

Status:

Version 1

Abstract

Figures

BACKGROUND

Methods

Design

Main trial

Participants

Procedures

Randomization

Interventions

Measures

Statistical methods/analytics

Explanation of protocol modifications

Embedded qualitative component

Results

Discussion

Limitations

CONCLUSIONS

Abbreviations

Declarations

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1