Opioid Treatment Programs, Telemedicine and COVID-19: A Scoping Review

Abstract

Background:

Methadone and buprenorphine are effective medications for opioid use disorder (MOUD) that are highly regulated in the United States. The on-going opioid crisis, and more recently COVID-19, has prompted reconsideration of these restrictions in order to sustain and improve treatment access, with renewed interest in telemedicine. We reviewed the evidence on use of telemedicine interventions and applicability to MOUD policy changes in the post-COVID-19 treatment landscape.

Methods:

Ovid MEDLINE and the Cochrane Database of Systematic Reviews databases were searched from inception to April 2021 and reference lists were reviewed to identify additional studies. Studies were eligible if they examined telemedicine interventions and reported outcomes (e.g., treatment initiation, retention in care). Randomized trials and controlled observational studies were prioritized; other studies were included when stronger evidence was unavailable. One investigator abstracted key information and a second investigator verified data. We described the results qualitatively.

Results:

We identified nine studies: three controlled trials (two randomized), and six observational studies. Three studies evaluated patients treated with methadone and six studies with buprenorphine, including one study of pregnant women with OUD. All studies showed telemedicine approaches associated with similar outcomes (treatment retention, positive urine toxicology) compared to treatment as usual. Trials were limited by small samples sizes, lack of reporting harms, and most were conducted prior to the COVID-19 pandemic; observational studies were limited by failure to control for confounding.

Conclusions:

Limited evidence suggests that telemedicine may enhance access to MOUD with similar effectiveness compared with face-to-face treatment. Few studies have been published since COVID-19, and it is unclear the potential impact of these interventions on the existing racial/ethnic disparities in treatment. The COVID-19 pandemic and need for social distancing led to temporary policy changes for prescribing of MOUD that could inform additional research in this area to support comprehensive policy reforms.

INTRODUCTION:

Methadone and buprenorphine are the two Food and Drug Administration (FDA) opioid agonist medications approved for treatment of opioid use disorders (OUD). The use of these medications for the treatment of OUD in the United States (U.S.) is highly regulated by states, the Substance Abuse and Mental Health Services Administration (SAMHSA), and the Drug Enforcement Administration (DEA).¹ Based on current regulations, methadone can only be administered and dispensed in federally certified Opioid Treatment Programs (OTPs). Buprenorphine, in contrast, can be prescribed in outpatient office-based treatment settings, though prescribing requires that providers obtain a DEA waiver and complete periodic face-to-face visits with their patients.

There is increasing concern that U.S. MOUD treatment policies contribute to difficulties in accessing life-saving medications and retention in treatment. Individuals living in rural areas tend to have lengthy commutes to OTPs located primarily in metropolitan communities.² The requirement of face-to-face treatment visits for the receipt of MOUD has been cited as a reason for low uptake of MOUD for patients with OUD.³ In addition, the SARS-CoV-2 pandemic (COVID-19) and the need for social distancing measures further complicated OUD treatment due to infection risk during face-to-face treatment in congregate or office-based settings. In response to COVID-19, SAMHSA issued a declaration addressing changes in use of telemedicine to deliver services building off prior regulations from the Ryan Haight Act of 2008.⁴ Under COVID-19 related emergency declarations SAMHSA permitted the use of telemedicine technology for delivery of treatment and psycho-social services (e.g., counseling and education) at OTP settings, as well as office based settings for people under care of a DATA 2000 (Drug Abuse Treatment Act of 2000) waivered prescribers.⁵ However, unlike buprenorphine prescribing, patients starting methadone at an OTP have to go in-person for their first visit. The policy changes related to MOUD delivery during the COVID-19 pandemic present an opportunity to review the evidence for telemedicine interventions that can be considered for integration into OTP or other settings to maintain and increase access to OUD treatment. Reviewing the state of the evidence can inform policy makers and treatment programs on ways these interventions have been used in research to inform evidenced-based adaptations during COVID-19, and inform sustained policy and service delivery improvements post COVID-19, where we will continue to face rising opioid overdose death rates.⁶ Therefore, we conducted a scoping evidence review^{7, 8} on treatment of OUD patients with methadone or buprenorphine using telemedicine interventions.

REVIEW:

Methods

Search Strategies.

We searched Ovid MEDLINE and the Cochrane Database of Systematic Reviews databases from inception to July 2020 for studies of telemedicine interventions using methadone and buprenorphine for treatment of OUD. See Supplement Table 1 for a description of the search strategies. We also reviewed reference lists of relevant articles to identify additional studies. This analysis was based on a larger scoping review of policy research on treatment for OUD that covered additional research areas, including office-based methadone, interim methadone, and mobile methadone services.⁹ An updated search for newly published studies through April 2021 identified four additional studies that met inclusion criteria.

Study Selection.

We utilized a scoping review approach in which studies were selected using a hierarchical approach that prioritized the most relevant and highest quality evidence.^{7, 8} One investigator reviewed abstracts and full-text articles for inclusion. Studies were eligible if they addressed telemedicine delivery of MOUD treatment or services in OTP or office-based settings and reported outcomes (e.g., treatment initiation, treatment receipt, retention in care, drug use, or other health outcomes including mortality rates or quality of life), and were conducted in the U.S. or very high development index countries. We prioritized randomized trials and controlled observational studies; U.S.-based uncontrolled and descriptive studies were included when stronger evidence was unavailable. One individual abstracted key information (i.e., clinical setting, country, sample size, intervention/comparison, and main findings) into a table and a second person verified data. Given the limited evidence and scoping review approach, we summarized the literature descriptively and did not formally synthesize or grade the quality of the evidence, though methodological limitations are noted. For this review we defined telemedicine to include telehealth, telepsychiatry, tele-prescribing interventions. The literature flow diagram is available in Figure 1.

Figure 1.

Scoping Review Search Tree

Results

We identified nine studies of telemedicine and provision of methadone or buprenorphine for OUD. There were two randomized controlled trials, one non-randomized controlled trial, and six observational studies. One of the observational studies reported updated results¹⁰ of a previously published study;¹¹ updated study results were included in the analysis. Three studies investigated telemedicine interventions for methadone and six trials for buprenorphine.^10–18 Apart from a large retrospective cohort study conducted across 85 provincial clinics in Ontario, Canada (N=3,733),¹³ all studies were conducted in the U.S. and were single site studies ranging from 12 to 443 people. The RCTs, which compared telemedicine versus face-to-face counseling in methadone treated patients were limited by small sample sizes (n=37 and 85) and high attrition rates (n=13 and 26).^{15, 16} The non-randomized controlled trial (n=98) enrolled pregnant women with OUD seeking obstetric care and offered buprenorphine MOUD.¹⁴ Two of the observational studies had comparison groups, consisting of patients treated in-person at an urban addiction treatment center compared to patients in rural areas who had telemedicine visits with the MOUD prescriber.^{13, 18} The remaining four observational studies did not have comparison groups; two reported results from a retrospective observational study (n=177, n=443),¹¹ one was a pre-post study with single arm (n=12),¹² and one was a retrospective cohort (n=78) of a telemedicine intervention that operated during the early part of the COVID-19 pandemic.¹⁷

The two RCTs and non-randomized trial found provision of MOUD along with telemedicine counseling interventions associated with similar effectiveness as face-to-face usual care.^14–16 One pilot randomized trial (n=37) tested an internet-based group therapy using manual guided relapse prevention versus routine in-person based group therapy of the same content for urban OTP patients who required treatment intensification following insufficient response to a lower level of counseling.¹⁶ Mean age was 40 years, 62% of participants were women, and 44% were a racial minority. Internet-based counseling was associated with slightly higher rate of counseling adherence that was not statistically significant (89% vs 74%, p=0.07), and there were no differences in drug use (likelihood of positive urine drug screen at 6 weeks (37% vs 42%, p=not significant [NS]), or likelihood of two weeks of negative urine drug screen (70% vs. 71%, p=NS)).¹⁶ A slightly larger trial of the same intervention but enrolling stable abstinent methadone maintained patients (n=59) with average age 41 years, 56% female, 36% racial minority also found no difference between telemedicine counseling versus face-to-face (usual) care at 12 weeks in counseling attendance (7.3 vs. 6.1 visits, p=.17), drug use (11% vs. 9%, p=.83), and treatment satisfaction using the client satisfaction questionnaire (CSQ, 3.8 vs 3.6 on a 1 to 4 Likert-scale, p=NS).¹⁵ Attrition was high due to difficulties with internet access, and these trials were conducted prior to widespread broadband internet access and smart-phone technologies.

A non-randomized controlled trial of 98 pregnant women with OUD (mean age 30.2 (SD 5.1) offered telemedicine (n=44) versus in-person (n=54) treatment; visits in both groups were weekly for four-weeks, bi-weekly for four weeks, and monthly thereafter.¹⁴ Patients were referred to the telemedicine intervention and had similar demographics compared to those who received face-to-face care (average age 30 years, 40% white, and similar daily morphine-equivalent dose 140–145 MME), though the telemedicine group had higher rates of being married, engaged, or cohabitation compared to the in-person group (76% vs 54%, P=0.02) Telemedicine and in-person treatment were associated with similar treatment retention outcomes 6–8 weeks post-partum in likelihood of treatment retention (80.4% telemedicine vs 92.7% in-person; treatment effect −12.2%; 95% CI, −32.3% to −4.4), likelihood of neonatal abstinence syndrome (45.4% telemedicine vs 63.2% in-person, P=0.12), or likelihood of positive UDS at 6–8 weeks post-partum (16.0% vs 20.3%, P=0.74), after propensity score adjustment.

Two observational studies also found telemedicine approaches associated with similar treatment outcomes compared to face-to-face approaches.^{13, 18} One was a large (n=3,733) retrospective cohort analysis conducted among provincial clinics in Ontario, Canada that used telemedicine for treatment visits. In Canada, methadone can be prescribed by primary care physicians, and stabilized patients received treatment in a family physician’s office or community pharmacy where observed dosing occured.¹³ This study evaluated a telemedicine approach in which patients at a physical site under nurse supervision with similar procedures to in-person approaches received telehealth services from a physician in underserved rural and northern areas. The study found improved one-year retention with either methadone or buprenorphine among individuals who primarily used telemedicine for physician check-ups (75% or more of the encounters; n=1,570) compared to mostly in-person check-ups (25% or fewer video encounters; n=1,745) (adjusted odds ratio [aOR] 1.27; 95% CI 1.14 to 1.41).¹³ A two year retrospective chart review of buprenorphine treated patients at a mental health service center in rural West Virginia compared delivery of group therapy and medication management through telemedicine (n=46) or in-person visits (n=54) and found no differences in abstinence at 90 days (49% telemedicine vs 37% in-person, P=0.31), or retention at 365 days (42% vs 36%, P=NS), all unadjusted.¹⁸

Four uncontrolled retrospective studies found telemedicine interventions associated with moderate or high retention at two to 12 months.^{10–12, 17} All were small single site evaluations that sought to use telemedicine to expand access into underserved rural areas. A study of the first 177 patients enrolled in a telemedicine program in a rural Maryland clinic partnered with an urban academic medical center’s addiction medicine program reported 57% retention at 3-months (with 86% providing opioid-free urine screens).¹¹ A follow-up report of the same study of 468 patients (443 prescribed buprenorphine) found similar results (50% retention at 3-months, with 93% providing opioid-free urine screens).¹⁰ A third was a quality improvement evaluation of a telemedicine intervention for Veterans consisting of a hub (centralized prescribers) and spoke (rural clinic) model with implementation of a “how-to” Toolkit with facilitators that enrolled 12 patients from rural areas;¹² at 6-months, 9 (75%) were still receiving telemedicine prescribed buprenorphine. The only study we found that was conducted during COVID-19 reported outcomes of a telemedicine intervention in a urban health system that pivoted to using videoconferencing (or phone visits) in place of in-person initial and follow-up visits, along with between visit outreach by medical students and clinic staff.¹⁷ Among the 78 patients served, 68% were non-Hispanic white participants. At 2-months, 53.8% remained in the program and 27% transitioned to another community treatment program. This study was the only one to report adverse events, with eight reports of persisting withdrawal symptoms (4.3%), and one reported non-fatal overdose (0.5%).¹⁷

Discussion

The COVID-19 pandemic occurred in the context of an already existing and ongoing opioid epidemic. The U.S. response to the COVID-19 pandemic has led to a re-examination of numerous aspects of the healthcare delivery system, including treatment for OUD. Policy changes in response to COVID-19 to maintain access and delivery of OUD treatment could have the potential to expand the role of telemedicine in addiction treatment delivery moving forward. We found few studies of telemedicine interventions either in the delivery of MOUD or in the delivery of psychosocial counseling interventions in conjunction with MOUD. The available evidence reported consistent findings that telemedicine interventions may be associated with similar outcomes (drug use, adherence, and retention) compared to in-person approaches. This suggests that changes to OUD treatment policies in response to the pandemic that permitted use of telemedicine should be continued post-COVID-19, though additional research is needed.

Practice & Policy Implications.

There is increased attention on the downstream effects of COVID-19 on the drug-related overdose crisis and exacerbation of health inequities. Low access and uptake of medications for opioid use disorder (MOUD) is well described in the literature prior to COVID-19.^{3, 19} The reasons for this are multi-factorial, including lack of availability of treatment, and lack of patient centered approaches to treatments that place high burdens on patients.³ While the liberalization of MOUD prescribing policies is meant to decrease the spread of COVID-19, with minimal disruption in treatment, the changes also present opportunities to expand access to treatment in the post-pandemic era. Our findings suggest that modifications to contemporary policies to allow for telemedicine interventions could enhance access to care, which could be of particular importance in rural communities that live farther away from addictions treatment providers or OTPs compared to urban populations. An estimated 56% of rural communities do not have access to buprenorphine waivered providers;²⁰ and people in rural areas experience longer drive times to OTPs.^{2, 20, 21} Telemedicine is used for MOUD in other countries (e.g., Canada), and the rapid upscaling of telemedicine in the U.S. during COVID-19 demonstrates that implementation is feasible. A potential barrier to implementation is that many persons in rural areas or with lower socioeconomic status do not have adequate internet access.²²

It is also important that expansion of opioid treatment services, including use of telemedicine interventions to aid delivery, aim to decrease rather than increase health inequities and address persistent racial disparities in care, which are observed in the current design of the system.^23–25 It is known that Black patients are less likely to receive buprenorphine (office-based care).²⁶ Therefore, we must be careful not to unintentionally worsen existing racial disparities through the implementation of telemedicine interventions. It has already been reported in studies of health services utilization during COVID-19 where preliminary data suggests racial/ethnic disparities in receipt of virtual primary care visits.^27–29 Thus, if regulations regarding the use of telemedicine differ by MOUD and setting, there is potential for worsening existing racial/ethnic disparities in OUD treatment if the underlying reasons for differences in MOUD treatment selection are not addressed. In the studies we reviewed, only three studies included significant minoritized populations,^15–17 and no studies reported differences in outcomes for underserved populations.

Future Research.

Our review highlights several directions for future study. First, additional studies are needed to determine whether telemedicine approaches are more effective than usual care in-person approaches through completion of adequately powered inferiority trials. These trials should not only assess benefit, but also problems associated with telemedicine interventions. A prior review on telemedicine for substance use disorders including OUD had similar findings;³⁰ while our review includes two more recent studies, we expected to find additional published trials since COVID-19 but only identified one. There are emerging published reports of clinician and patient experience with telemedicine during COVID-19—most cite positive aspects including opportunity to improve quality of patient interactions and access, but also bring up concerns for future study including potential harms due to less structure and accountability, as well as technological challenges.^31–35 However, only one study reported potential harms in the studies we reviewed.¹⁷

Studies should evaluate diverse populations and evaluate impacts on access and outcomes in minoritized groups based on race/ethnicity, special populations (adolescents, pregnant populations), socioeconomic status, and rural status. Ideally, these studies would be randomized controlled trials but, as the Ontario study demonstrates,¹³ well-conducted observational studies may be informative. Whether telemedicine interventions are effective for long-term stable patients, newly stabilized patients, and/or unstable patients and during what part of the treatment process (e.g., intake, counseling, medication management in acute and maintenance phase of recovery) is unknown.

Limitations.

There were few studies of telemedicine interventions for MOUD and the studies we reviewed had design limitations limiting ability to draw broad conclusions. There was heterogeneity in how telemedicine was deployed: the two RCTs trialed internet-based group therapy in a methadone OTP setting, whereas the other studies used telemedicine for prescribing and treatment services in outpatient settings. There were differences in population (e.g., pregnant versus non-pregnant), and recovery stage (e.g., newly initiated versus MOUD maintained patients versus those unstable). All the included studies were pre COVID-19, and we did not identify any published studies of telemedicine since the pandemic. Given the rise in drug overdoses since the start of the pandemic, additional studies across these populations and settings would be important to study now.³⁶

While we did not conduct formal quality rating, important limitations were noted including non-randomization, lack of control groups, or did not provide details of allocation/concealment and outcomes assessment procedures. The Ontario study, where office-based methadone and pharmacy prescribing is permitted, may have decreased applicability to the U.S.

Conclusion.

Limited evidence suggests that telemedicine could enhance access to medications for opioid use disorder safely and with similar effectiveness compared with face-to-face treatment. Continuation of federal policies supporting the use of telehealth interventions enacted due to COVID-19 pandemic merit consideration and further study, even after resolution of the pandemic.

Table 1.

Overview of Studies for Telemedicine in Opioid Treatment Programs

Author, Year, MOUD	Design N	Setting Country	Intervention	Main Findings
Methadone
King, 200916	Pilot RCT N=50 (37 randomized)	OTP single site Baltimore, Maryland, U.S.	Analysis of non-responders to lower levels of treatment A. Internet based group therapy using manual guided relapse control B. In-person based group therapy of same manual guided relapse control	6-week outcomes: Counseling adherence: 89% versus 74%, p=0.07 Drug use (% positive urine at 6 weeks): 37% versus 42%, p=NS Step completion (2-weeks negative UDS and 100% adherence): 70% versus 71%, p=NS
King, 201415	RCT N=85 (59 randomized; 26 withdrew)	OTP single site Baltimore, Maryland, U.S.	Analysis of abstinent and counseling adherent patients A. Internet based group therapy using manual guided relapse control for 12 weeks B. In-person based group therapy of same manual guided relapse control for 12 weeks	Counseling attendance at 12 weeks: 7.3 (SD=2.49) versus 6.1 (SD=3.30), p=0.17 Drug use (% urine positive): 11% (SD=0.27) versus 9% (SD=0.16), p=0.83 Treatment satisfaction (CSQ): 3.8 (SD=0.31) versus 3.6 (SD=0.52)
Eibl, 201713	Retrospective cohort analysis N=3,733 across 58 centers	Outpatient Multi-site Ontario, Canada	Analysis of retention in care for patients newly initiating methadone A. Predominantly in-person (<25% encounters occurred by telemedicine) B. Predominantly telemedicine (>75% encounters occurred by telemedicine) C. Mixed (≥25% and ≥75% via telemedicine)	Retention rates at 30 days, 180 days, 365 days: A. 82% versus B. 88% versus C. 88%; 52% versus 63% versus 61%; 61% versus 50% versus 53% Days retained: median (IQR): A. 207 (50, 75) versus B. 366 (97, 556) versus C. 317 (72, 576) Multivariate Cox proportional hazard (adjusted for age, sex, clinic region, rurality, methadone peak dose) showed group B patients (predominantly treated with telemedicine had greater likelihood of being retained in treatment for 1 year versus those predominantly in-person, (aOR 1.27 95% CI 1.14 to 1.41; p<0.001)
Buprenorphine
Guille, 202014	Prospective non-RCT N= 98 women receiving perinatal treatment	Outpatient women’s health program in Charleston, South Carolina, U.S.	Pregnant women with OUD presenting to obstetrician’s office were seen in person first and evaluated by psychiatrist with perinatal and addiction training, induced on buprenorphine, and then seen weekly either in-person or via telemedicine	At 6–8 weeks post-partum: 85.4% (35/41) of telemedicine vs 91.7% (44/48) in-person were retained in treatment (P=0.50); after propensity score weighting, no significant differences were found (80.4% vs 92.7%, treatment effect −12.2%; 95% CI, −32.3% to −4.4%) Positive UDS at 6–8 weeks: 9.8% vs 20.8%, unadjusted P=0.24; adjusted – no difference) Incidence of neonatal abstinence syndrome (NAS): 43.6% vs 62.2%, unadjusted P=0.12; propensity score adj 45.4% vs 63.2%; −17.8%; 95%CI, −41.0% to 8.9%
Zheng, 201718	Retrospective Cohort N=100 at 1 site	OTP single site Morgantown, West Virginia, U.S.	Analysis of treatment outcomes at a comprehensive opioid addiction treatment with group-based medication management and focused group therapy (CBT model) weekly initially + 4 12-step meetings/week A. Telepsychiatry group: psychiatrist delivers addiction assessment and care in group setting via videoconferencing offered to 2 rural counties 225 miles away. Group therapy and counseling received locally B. In person group: assessment and care delivered in face-to-face group setting	90-day consecutive abstinence: 49% vs 37%, P=0.31 Time to 30-day abstinence (median, range): 30 (30,70) vs 30 (30,112), P=0.09; Time to 90-day abstinence (median, range): 90 (90,194) vs 94 (90,236), P=0.22 Rates of additional substance use were similar between groups Retention rates at 90-days: 12 (50%) vs 14(44.9%), P=0.99 Retention rates at 365-days: 10 (41.7%) vs 11 (35.5%), P=0.99
Weintraub, 2017, 202110, 11	Retrospective cohort (N=468, 443 prescribed buprenorphine)	Intensive outpatient OTP rural Maryland, U.S.	Rural treatment center with intensive outpatient treatment program, transitional housing, self-referral or criminal justice, medical, psychiatric, or inpatient substance use treatment centers partnered with academic center to provide buprenorphine care via telemedicine	By end of 3-months of treatment: 50% (220/443) remained engaged in treatment 93% (350/387), 91% (324/359), and 93% (209/225) of those retained maintained opioid negative UDS at 1-week, 1-month, and 3-month follow-up
Brunet, 202012	Implementation evaluation with pre-post evaluation N=12	VA outpatient clinics in rural areas, Connecticut, U.S.	Prescribers at “hub” sites were able to treat Veterans with OUD at 3 distal “spokes” that did not have local mental health staff. “Hub” prescriber responsible for evaluating patient for appropriateness of tele-buprenorphine, obtaining consent, prescribing, and follow-up. Spoke sites responsible for scheduling tele-buprenorphine session, conducting toxicology testing, and obtaining laboratory tests, etc.	At 6-months, 9/12 were still receiving tele-buprenorphine. 1 discontinued due to nausea adverse event, another transferred to in-patient treatment, and the third discontinued buprenorphine after 2 weeks
Tofighi, 202117	Retrospective cohort study N=78	NYC Health+Hospitals Virtual Clinic, New York, U.S., conducted during COVID-19	Initial visit of standard 30–45 minute evaluation using video conference or phone if patients lacked smartphone where providers counseled on home-induction of buprenorphine, emailed instructions, offered referrals, and prescribed buprenorphine, followed by between-visit patient calls by medical students and clinic coordinators. Follow-up visits every 1–2 weeks.	By end of 2-months: 42/78 (53.8%) remained in program 21/78 (26.9%) transitioned to community treatment program, 15/78 (19.2% lost to follow-up) 29/78 (39.7%) prescribed naloxone during initial visit Adverse events: 8 (4.3%) persistent withdrawal; 1 reported non-fatal overdose (0.5%)

Abbreviations: MOUD = medications for opioid use disorder; CBT = cognitive behavioral therapy; aOR = adjusted odds ratio; NS = not significant; OTP = opioid treatment program; RCT = randomized controlled trial; SD = standard deviation; UDS = urine drug screen; U.S. = United States.