Abstract
Background:
Opioid use disorder (OUD) drives high morbidity and mortality, but access to opioid agonist therapy (OAT) is limited in low- and middle-income countries. Integrating OAT into primary care may expand access and improve comorbidity management, though provider discomfort remains a barrier.
Objective
To compare healthcare utilization among individuals with OUD receiving methadone in specialty clinics versus primary care centers in Ukraine (01/2018–12/2023).
Design:
Two-arm randomized controlled trial with 2:1 allocation to intervention and control.
Setting:
Thirteen cities in Ukraine: Cherkasy, Dnipro, Kramatorsk, Kropyvnytskyi, Kryvyi Rih, Kyiv, Lviv, Mariupol, Mykolaiv, Odesa, Rivne, Sloviansk, and Zhytomyr.
Participants:
A total of 1,459 adults with OUD (950 intervention, 509 control) initiating or receiving methadone.
Interventions:
Methadone delivered in primary care aided with tele-mentoring, an Extension for Community Healthcare Outcomes (ECHO)-like model that is adapted to the Ukraine context, versus standard specialty clinic care.
Measurements:
Primary outcome: difference in composite Quality Health Indicator (QHI) scores between arms at 24 months, representing access to 17 guideline-concordant services (9 primary care and 8 specialty care) received, assessed through surveys and ranging from 0 to 100 percentage points. Secondary outcomes: domain-specific QHI scores and methadone treatment indicators.
Results:
Participants in primary care settings achieved higher composite QHI scores than those in specialty clinics, with a mean difference of 9.1 percentage points (95%CI:6.9–11.2) at 24 months. Results were similar for primary care QHI (12.3; 95%CI:9.0–15.6) and specialty care QHI (5.2; 95%CI:0.2–10.3). Methadone retention among new patients at 24 months was 67.2% in primary care versus 64.7% in specialty clinics.
Limitations:
QHIs reflect healthcare utilization rather than health outcomes. QHIs were equally weighted despite differing clinical significance.
Conclusion:
Integrating methadone treatment into primary care settings improves adherence to guideline-concordant healthcare without compromising methadone retention and treatment quality.
Primary Funding Source:
National Institutes on Drug Abuse.
Introduction
The global burden of opioid use disorder (OUD) remains substantial, with high morbidity and mortality due to overdose, infectious diseases, and non-communicable conditions (1, 2, 3). Opioid agonist therapies (OAT), specifically methadone and buprenorphine, are the most effective treatments for OUD, reducing overdose risk, illicit drug use, human immunodeficiency virus (HIV) and hepatitis C virus (HCV) transmission risk, while improving health and social outcomes (4, 5). Although the World Health Organization (WHO) recommends integrating OAT into primary care to expand access and address comorbidities, implementation has been uneven, particularly in low- and middle-income countries (6).
Integration of OAT into primary care not only expands access to treatment but facilitates preventive screenings and management of comorbidities within routine care (7, 8). This approach directly addresses the excess premature mortality observed among people with substance use disorders, much of which is attributable to conditions that are both preventable and treatable in primary care (9, 10). A three-site pilot study we conducted in 2016 found comparable OAT retention between primary care centers and specialty clinics, though primary care providers reported discomfort with managing patients with OUD and prescribing methadone (11). In Ukraine, such reluctance stems from stigma, limited training, and concerns about the complexity of care, along with lack of motivation due to low clinician salaries (11, 12).
Implementation strategies like the Extension for Community Healthcare Outcomes (Project ECHO) model, a tele-mentoring model designed to extend specialty expertise to primary care providers, offer an approach to strengthen provider confidence and build capacity (13). To further address lack of motivation among providers, pay-for-performance strategies have shown promise in improving provider engagement and adherence to guideline-concordant care, particularly in resource-constrained settings (14, 15).
Studies of OAT in primary care settings in high-income countries, mostly with buprenorphine, have shown retention in treatment comparable to specialty settings, as well as reductions in illicit drug use (16, 17, 18, 19, 20, 21). Little is known, however, about the impact of methadone integration into primary care on general health outcomes of individuals with OUD in low- and middle-income countries, where methadone remains the more accessible option (22).
To address this gap, we conducted a randomized controlled trial in Ukraine comparing healthcare utilization among individuals with OUD receiving methadone in specialty addiction treatment clinics (hereafter “specialty clinics”) versus primary care centers. Ukraine is a middle-income country with high OUD prevalence. At the start of the trial in 2018, national OAT coverage was at 3.6%, well below the WHO’s recommended 40% (23, 24). Opioid use in Ukraine is predominantly via injection, and OUD frequently co-occurs with HIV, HCV, and tuberculosis, with over one-fifth of those with OUD living with HIV (25, 26). To support primary care integration, the study incorporated physician training through a tele-mentoring program similar to the ECHO model adapted to Ukraine context and financial incentives through a pay-for-performance strategy.
Methods
Design and study population
The rationale and details of the trial design have been previously published (27). Patients were individually randomized into two arms: (a) control, in which they received methadone in specialty clinics, and (b) intervention, in which they received methadone in primary care centers. The study was conducted in 36 clinics across 12 cities in Ukraine. Within each city, we selected one specialty clinic and two primary care centers. Sites were chosen based on willingness to participate, and specialty clinics were included if they served more than 50 patients receiving methadone therapy. Enrollment began in January 2018 and continued through December 2023, while data collection extended through June 2024. The study sites in Mariupol were closed following Russia’s invasion of Ukraine in 2022. Subsequently, we added three sites in Lviv, bringing the total number of participating clinics to 39 (Supplementary Table 1).
Study participants were recruited from patients at participating specialty clinics. Recruitment was conducted by trained research staff who worked in coordination with clinic personnel. Approximately half of the recruited patients were selected among individuals newly initiating methadone treatment. These patients were approached at the time of their intake visit, after meeting with clinic staff. The other half of participants were randomly selected from clinic rosters of patients who were receiving methadone for over three months, and were invited to participate during routine clinic visits. Research staff introduced the study, confirmed eligibility, and obtained written informed consent. Participants were eligible if they were 18 years or older, were diagnosed with OUD (International Statistical Classification of Diseases 10th revision [ICD-10], F11), and were residing in site city. Exclusion criteria included being under criminal investigation, planning to move to another city, or being unable to provide informed consent. Participants randomized to the primary care arm had their care transferred to the primary care clinic on the same day to avoid delays in methadone initiation. When same-day transfer was not possible, participants began methadone treatment at the specialty clinics in which they were recruited and then transferred to the primary care clinic within 7 days without interruption of methadone care. Participants were excluded from the study if transfer was not possible within 7 days. Additionally, methadone treatment and study participation were independent. Patients could continue receiving methadone regardless of their study participation status. Institutional review boards at Yale University and the Ukrainian Institute on Public Health Policy approved the study, which is registered with ClinicalTrials.gov (NCT04927091).
Randomization
Randomization was conducted centrally through REDCap using permuted blocks of varying sizes to allocate participants in a 2:1 ratio to primary care centers versus specialty clinics. Randomization was stratified by OAT duration (newly initiating methadone vs continuously on methadone for more than 3 months) and city (28). Participants randomized to specialty clinics continued treatment there, while those randomized to primary care selected one of two primary care centers based on geographic preference. Clinical providers at one primary care center in each city received financial incentives if their patients reached pre-defined clinical goals (pay-for-performance), while those in the other primary care center did not receive any incentives. Participants were blinded to the financial incentive status of the primary care centers. Due to the nature of the intervention, blinding of participants with respect to the study intervention was not possible.
Implementation Strategies
In specialty clinics, patient management and clinical operations remained unchanged. In primary care centers, we used the integrated Promoting Action on Research Implementation in Health Services (i-PARIHS) framework to guide our implementation efforts (29). The model emphasizes four key constructs for real-world implementation of evidence-based practices in healthcare settings: innovation, context, recipients, and facilitation.
To address provider discomfort with OUD management, we implemented an ECHO-like continuous tele-mentoring program adapted to Ukraine context that connects specialists with primary care providers through videoconferencing in the intervention arm (13). Providers at primary care centers completed a three-day training, followed by weekly 60-minute collaborative tele-mentoring sessions led by national experts in addiction and infectious diseases. Each session featured a brief didactic lecture on infectious diseases content focused on management of HIV and tuberculosis, followed by case presentations by primary care providers, fostering interactive discussions and collaborative learning. Additionally, monthly quality improvement sessions during tele-mentoring sessions guided providers in setting goals, establishing measures, designing tests of change, and building sustainable capacity for integrating methadone into primary care settings. Attendance was encouraged but not mandatory, and educational credits were provided for attendance. Recorded sessions were available for asynchronous viewing.
A pay-for-performance strategy was implemented in one of the two primary care centers in each city (30). Providers were informed in advance about predefined quality health indicator goals required to receive incentives, as well as the payment structure and frequency. Both physicians and nurses were eligible for monthly payments based on patient achievements: if a patient achieved all pre-defined monthly indicators and was retained in OAT for 12 months, physicians received 100 Ukrainian hryvnia (UAH; ~$3.60) per patient, while nurses received 60 UAH (~$2.16). Each clinic aimed to enroll 40 participants, allowing for a maximum potential monthly payment of 4,000 UAH (~$111) for physicians and 2,400 UAH (~$67) for nurses. The average monthly salaries for primary care physicians and nurses were $250 and $150, respectively.
Data Collection
Data were collected through structured interviews at baseline and every 6 months for 24 months. Surveys were administered on-site by study staff during participants’ regular clinic visits. Because methadone treatment requires regular, often daily, clinic visits, study staff were able to adhere closely to the pre-specified schedule.
The baseline survey captured demographics, socioeconomic factors (e.g., income, employment, housing, education), and comorbidities (HIV, HCV, tuberculosis). All interviews assessed healthcare utilization using quality health indicators (QHI) adapted for Ukraine. QHI included nine primary and eight specialty care service indicators aligned with national guidelines and developed through a Delphi method with local experts (31) Primary care indicators included physical examination, complete blood count, urinalysis, electrocardiogram, cervical, breast, and prostate cancer screenings, and hepatitis B and C testing. Specialty care services included HIV and tuberculosis testing, CD4 count/viral load monitoring and antiretroviral therapy for patients with HIV, and methadone-related indicators such as retention on methadone, receipt of an optimal methadone dose (>85 mg) (32, 33), and unsupervised take-home methadone dosing. Development of the QHI inventory and relevant national guidelines are presented in the Appendix 2.1. Participants in both arms received 250 UAH (~$9) per survey.
A composite QHI score was calculated as the percentage of applicable services accessed by each participant (range 0–100). Since not all patients were eligible for all 17 measures (e.g., breast cancer screening applied only to women aged 50 years or older), QHI scores were calculated as the percentage of accessed services and screenings out of the total applicable for each participant. For annual screenings, a report at any six-month follow-up was counted for that visit and the next visit within the same 12-month period. Cervical cancer screening was considered completed for all subsequent follow-ups after first being reported (see Appendix 2.2).
We abstracted the clinics’ electronic medical records (EMR) to obtain information on health care utilization of services performed on-site or verified off-site through medical documentation. These data were used to calculate EMR-based QHI scores used in secondary analyses. Appendix 2 describes the calculation of survey- and EMR-based QHI scores.
All participants received standard methadone treatment according to national guidelines and no additional trial-related harms were anticipated.
Outcomes
The primary outcome was the difference between study arms in the composite QHI scores, based on survey questionnaires at 24 months. Secondary outcomes included differences in self-reported composite QHI score at 6, 12, and 18 months; differences in self-reported primary care and specialty care QHI scores at all time points; differences in EMR-based QHI indicators; self-reported individual healthcare utilization measures; and methadone retention ascertained through EMR as patients could continue receiving methadone regardless of their dropout from the study. An interim analysis of 818 patients recruited through November 2020 and followed for at least 12 months has been published (34).
Statistical analysis
The study was initially planned to randomize participants into three arms: specialty clinics, primary care centers with pay-for-performance, and primary care centers without pay-for-performance. Sample size calculations required 405 participants per arm to detect a ≥0.10 difference in composite QHI score at 24 months with 90% power and a two-sided α of 0.05. To account for attrition, the target sample size was increased by 10% to 1,350 participants. Calculations were performed using PASS 15 software. Before recruitment began, community representatives recommended allowing participants randomized to primary care to choose between the two available centers in each city. The design was modified to a two-arm trial with 2:1 allocation to primary care versus specialty clinics. All participants were randomized under this two-arm design, and the total sample size target remained unchanged.
We used a pre-specified intention-to-treat approach, which preserves the benefits of randomization and provides an unbiased estimate of intervention effectiveness under real-world clinical conditions (35). Data were assumed to be missing at random. Categorical variables were summarized as frequencies and percentages; continuous variables as means with standard deviations (SD). Linear mixed models with random intercepts estimated average differences in QHI scores between arms, adjusting for the stratification variable (methadone use longer than 3 months), age, sex, city, and predictors of missingness (HIV and HCV). As visit times clustered closely around the scheduled time points, we modeled time as a categorical variable. Sensitivity analyses applied pattern mixture models to impute missing data at 1, 2, and 3 SDs from the mean self-reported QHI scores (36). EMR-based QHI scores were analyzed using linear regression adjusted for the same covariates.
Secondary analyses examined differences in QHI scores at other timepoints and between primary care centers with and without pay-for-performance incentives (non-randomized comparison). Completion of individual health services was assessed for all eligible patients at each time point based on self-report and EMR data were used to calculate methadone retention rates in the total sample over 24 months.
We calculated a single P value for the primary outcome. The type I error rate was set at 5% (two-sided). For secondary outcomes, we report nominal 95% confidence intervals (95%CI). R version 4.4.1 was used for all analyses (37).
Results
Participant Characteristics
From January 3, 2018, to December 31, 2023, we screened 3,429 individuals for eligibility and enrolled 1,459 across 13 cities and 39 clinics (509 in specialty clinics and 950 in primary care centers; Figure 1). Baseline characteristics were comparable between the two groups (Table 1). The median age was 39 years (SD 7.7), 83% were male and prevalence of HIV (42%) and HCV (57%) was high. Of all participants, 45% had been receiving methadone treatment for more than 3 months. Study retention rates were similar between the two arms at each time point: 90.8% vs 92.3% at 6 months, 84.4% vs 86.1% at 12 months, 78.7% vs 81.5% at 18 months, and 73.3% vs 74.5% at 24 months. Detailed reasons for study dropout are provided in Supplementary Figure 1. Clinical and demographic characteristics of participants at pay-for-performance and non- pay-for-performance primary care centers are presented in Supplementary Table 2 and were overall comparable between the two groups.
Figure 1.
Flow of participants through the randomized controlled trial of integrating methadone into primary care centers in Ukraine
Table 1.
Baseline demographic and clinical characteristics of study participants.
| Specialty Addiction Clinics (N=509) |
Primary Care Centers (N=950) |
Total (N=1,459) |
|
|---|---|---|---|
|
| |||
| Age, cont. / mean (SD) | 39 (7.8) | 39 (7.6) | 39 (7.7) |
| Male | 418 (82%) | 793 (83%) | 1211 (83%) |
| Unemployed | 273 (54%) | 503 (53%) | 776 (53%) |
| Income below poverty 1 | 183 (36%) | 309 (33%) | 492 (34%) |
| Married or cohabitating | 166 (33%) | 331 (35%) | 497 (34%) |
| Housing | |||
| Lives in own house/apartment | 155 (30%) | 328 (37%) | 483 (33%) |
| Lives with family/partner/friends | 337 (67%) | 592 (60%) | 929 (64%) |
| Other2 | 17 (3%) | 30 (3%) | 47 (3%) |
| Above secondary education | 319 (63%) | 577 (64%) | 896 (61%) |
| Time on OAT | |||
| More than 3 months on OAT | 288 (57%) | 508 (53%) | 796 (55%) |
| Newly initiating OAT | 221 (43%) | 442 (47%) | 663 (45%) |
| HIV status, self-report | |||
| Negative | 261 (51%) | 497 (52%) | 758 (52%) |
| Positive | 215 (42%) | 396 (42%) | 611 (42%) |
| Unknown | 33 (7%) | 57 (6%) | 90 (6%) |
| HCV status, self-report | |||
| Negative | 135 (27%) | 238 (25%) | 373 (26%) |
| Positive | 284 (56%) | 548 (58%) | 832 (57%) |
| Unknown | 90 (18%) | 164 (17%) | 254 (17%) |
| Ever diagnosed with tuberculosis, self-report | 83 (16%) | 158 (17%) | 241 (17%) |
| City | |||
| Cherkasy | 37 (7%) | 76 (8%) | 113 (8%) |
| Dnipro | 39 (8%) | 83 (9%) | 122 (8%) |
| Kramatorsk | 53 (10%) | 103 (11%) | 156 (11%) |
| Kropyvnytskyi | 38 (8%) | 76 (8%) | 114 (8%) |
| Kryvyi Rih | 65 (13%) | 95 (10%) | 160 (11%) |
| Kyiv | 11 (2%) | 22 (2%) | 33 (2%) |
| Lviv | 30 (6%) | 55 (6%) | 85 (6%) |
| Mariupol | 36 (7%) | 69 (7%) | 105 (7%) |
| Mykolaiv | 50 (10%) | 89 (9%) | 139 (10%) |
| Odesa | 37 (7%) | 65 (7%) | 102 (7%) |
| Rivne | 41 (8%) | 77 (8%) | 118 (8%) |
| Sloviansk | 30 (6%) | 59 (6%) | 89 (6%) |
| Zhytomyr | 42 (8%) | 81 (9%) | 123 (8%) |
<1630 UAH (~$45)/month;
category other includes dormitory, hotel, temporary housing, and homeless.
Abbreviations: SD standard deviation; OAT opioid agonist therapies; HIV human immunodeficiency virus; HCV hepatitis C virus.
Retention on Methadone and Healthcare Utilization Metrics
Retention on methadone at 24 months was similar between specialty clinics and primary care centers. Based on EMR data, 356 of 509 patients (69.9%) in specialty clinics were retained on methadone, compared to 669 of 950 patients (70.4%) in primary care centers (difference in percentages 0.5, 95% CI −5.4 to 4.4). Among patients newly initiating methadone treatment, 143 of 221 (64.7%) in specialty clinics and 297 of 442 (67.2%) in primary care centers were retained at 24 months (difference in percentages 2.5, 95% CI −10.2 to 5.2). Kaplan-Meier curves comparing retention between the two study arms in the total sample and new patients only are presented in Supplementary Figure 2.
Based on self-reported surveys, retention on methadone at 24 months was 64.8% in the total sample (57.0% among new patients) in specialty clinics compared to 68.2% (62.7% among new patients) at primary care centers. The proportion of participants on an optimal methadone dose (>85 mg) increased from 33% to 67% over 24 months, while take-home methadone rose from 14% to 50% in primary care centers. In specialty clinics, optimal dosing increased from 31% to 58% and take-home dosing from 14% to 40% (Table 2).
Table 2.
Completion of selected general, OUD-related, and HIV-related health services at baseline and 24 months based on self-report.
| Specialty Addiction Clinics | Primary Care Centers | |||
|---|---|---|---|---|
|
| ||||
| Baseline | 24 months | Baseline | 24 months | |
|
| ||||
|
Screenings recommended for general population
| ||||
| Physical examination | 328/509 (64%) | 270/379 (71%) | 626/950 (66%) | 535/696 (77%) |
| Mammogram | 1/17 (6%) | 0/14 (0%) | 2/26 (8%) | 9/26 (35%) |
| Cervical cancer screening | 15/91 (16%) | 25/68 (37%) | 18/157 (11%) | 51/119 (43%) |
| Prostate cancer screening | 2/43 (5%) | 1/44 (2%) | 0/82 (0%) | 8/75 (11%) |
|
Screening recommended for individuals with OUD | ||||
| HBV screening | 156/509 (31%) | 105/379 (28%) | 350/950 (37%) | 324/696 (47%) |
| HCV screening | 102/225 (45%) | 91/314 (29%) | 186/402 (46%) | 199/549 (36%) |
| HIV screening | 220/294 (75%) | 144/201 (72%) | 410/554 (74%) | 297/377 (79%) |
| Tuberculosis screening | 66/503 (13%) | 34/375 (9%) | 109/931 (12%) | 67/692 (10%) |
|
Services recommended for individuals with HIV | ||||
| CD4 count or viral load | 182/215 (85%) | 136/178 (76%) | 338/396 (85%) | 253/319 (79%) |
| Antiretroviral therapy | 185/215 (86%) | 148/178 (83%) | 333/396 (84%) | 276/319 (87%) |
|
Methadone service indicators | ||||
| Take-home methadone dispersion | 69/509 (14%) | 150/379 (40%) | 136/950 (14%) | 348/696 (50%) |
| Methadone dose >85 mg | 156/509 (31%) | 220/379 (58%) | 313/950 (33%) | 468/696 (67%) |
|
Retention on methadone | ||||
| Total sample | - | 330/509 (64.8%) * | - | 648/950 (68.2%) * |
| New patients | - | 126/221 (57.0%) * | - | 277/442 (62.7%) * |
For the self-reported retention on methadone, the total sample at baseline was used as the denominator. Electronic medical records-based retention at 24 months was 69.9% in specialty clinics and 70.4% in primary care centers; values differ from self-reported estimates because some patients continued methadone treatment after study dropout.
Abbreviations: OUD opioid use disorder; HBV hepatitis B virus; HCV hepatitis C virus; HIV human immunodeficiency virus
Table 2 presents the percentage of eligible patients who reported completion of health service indicators grouped by general health screenings, screenings for individuals with OUD, and those with HIV. By 24 months, completion of physical examinations was high in both arms (71% in specialty clinics vs 77% in primary care centers). Primary care centers had higher percentages of cancers, HIV, HCV, and HBV screenings at 24 months compared to specialty clinics. Among patients with HIV, CD4 count/viral load monitoring and antiretroviral therapy remained high (76–86%) throughout follow-up across both arms. The complete set of QHI results across all timepoints is reported in Supplementary Table 3.
Primary and Secondary Outcomes
At baseline, composite QHI scores were similar between study arms (Figure 2 and Supplementary Table 4). Participants in primary care centers achieved significantly higher composite QHI scores at each subsequent time point, including a mean difference of 9.1 percentage points (95% CI: 6.9 to 11.2, P<0.001) at 24 months.
Figure 2. Adjusted expected values of the self-reported quality health indicator scores stratified by study arms.
The figure presents adjusted expected values for composite, primary, and specialty quality health indicator (QHI) scores (range 0–100). Models were adjusted for participant sex, age, HIV and HCV status, study city, and time on methadone (categorized as newly initiating methadone vs. more than 3 months on methadone). The primary endpoint was the difference over time in composite QHI scores between primary care centers and specialty addiction clinics at 24 months (p < 0.001). Primary care QHIs included physical examination, complete blood count, urinalysis, electrocardiogram, cervical and prostate cancer screenings, and hepatitis B and C testing. Specialty care indicators included HIV and tuberculosis screening, CD4 count/viral load monitoring, antiretroviral therapy for patients with HIV, and methadone-related measures: retention on methadone, receipt of an optimal dose (>85 mg), and unsupervised take-home methadone dosing. Composite QHIs included all indicators.
A similar pattern was observed in primary care QHI scores (mean difference at 24 months 12.3 percentage points, 95% CI 9.0 to 15.6) and specialty QHI scores (mean difference 5.2 percentage points, 95% CI 0.2 to 10.3). Across all time points, observed QHI scores spanned nearly the full possible range (0–100).
At 24 months, the average difference in EMR-based composite QHI scores between primary care and specialty arms was 27.1 percentage points (95%CI 24.7 to 29.4). Most of this difference was accounted for by the primary care QHI scores (mean difference 45.8 percentage points; 95%CI: 43.0 to 48.6%), with a smaller difference in specialty QHI scores (mean difference 9.8 percentage points; 95%CI: 7.2 to 12.5%) (Supplementary Table 5 and Supplementary Figure 3).
Participant characteristics receiving methadone at primary care centers with and without pay-for-performance are presented in Supplementary Table 6. The differences in QHI scores between these two groups were observed at only 18 and 24 months, with a difference of 5.2 (95%CI: 6.2 to 7.8) in composite QHI at 24 months. Based on EMR data, pay-for-performance clinics had modest gains in primary care QHI scores (Supplementary Table 7).
Sensitivity Analyses
Sensitivity analyses using pattern mixture models showed consistent results. Although in some extreme scenarios the association was no longer statistically significant, the direction of the effect remained consistent.
Discussion
This randomized controlled trial tested whether integrating methadone treatment into primary care, supported by ECHO-like tele-mentoring improves access to guideline-concordant healthcare. Patients receiving methadone in primary care settings reported accessing more recommended healthcare services than those managed in specialty clinics, with most of the gains driven by primary care screenings. The 9.1-point increase in composite QHI scores at 24 months is clinically meaningful: across up to 17 services, this corresponds to participants receiving 1–2 additional services on average. For primary care QHIs (eight services), the improvement similarly reflects at least one additional service. Methadone retention and other OUD indicators were similar across arms at 24 months, indicating that relocating care to primary care did not affect OAT quality while expanding access to preventive and chronic care services (34).
As one of the major barriers to methadone integration in primary care is primary care provider discomfort with prescribing methadone and managing patients with OUD, HIV, and tuberculosis, our trial offers a roadmap for a tele-mentoring-facilitated primary care model for OAT that can address this barrier. Participation in tele-mentoring provided case-based, expert guidance and near-real-time troubleshooting for primary care teams, which may have increased capability and confidence. Consistent with this mechanism, OUD-specific outcomes moved in favorable directions in primary care: treatment retention on methadone was similar to specialty clinics, a larger share of patients reached optimal dosing (>85 mg/day) (32, 33), and transition to take-home dispensing increased among eligible patients. While our design does not isolate the independent effect of tele-mentoring apart from co-location and workflow changes inherent to integrated care, our results support its role in reducing provider discomfort and enabling safe, guideline-concordant methadone management in primary care. Additionally, the finding that more patients in primary care transitioned to take-home dosing may have contributed to increased retention, a finding observed elsewhere (38).
Pay-for-performance financial incentives produced only modest, late-emerging effects. This may reflect suboptimal incentive or infrequent payments, competing workload, or indicators that required off-site referral. Because the pay-for-performance comparison was not randomized, causal inference is limited. Future work should test alternative behavioral-economics designs, including simpler payout schedules or clinic-level incentives tied to complexity-adjusted indicators.
Several contextual mechanisms likely contributed to the observed differences in service use. First, integrated care enabled co-location of routine OUD, HIV, tuberculosis and primary care services, increasing convenience for patients who interact frequently with the health system providing methadone. This is consistent with our findings of higher rates of routine examinations, blood pressure screening, laboratory analyses (e.g., complete blood count), and urinalysis in primary care, services readily available on-site. In contrast, screenings requiring referral off-site, for example, tuberculosis and cancer screenings, were low in both arms.
Second, providers in primary care centers may be more familiar with guideline-concordant screenings and medical services recommended for patients according to their age, gender, and comorbid conditions, and are likely to recommend preventive care to patients. Finally, stigma may contribute to differences in healthcare utilization. People who use drugs often face stigma in clinical settings, including in specialty clinics focused on addiction or HIV, which can lead to healthcare avoidance (39). Providing OUD-related care in primary care centers, where addiction is not the primary focus, may increase these patients’ comfort and trust in the healthcare environment and support their engagement in health-seeking activities (11). Having patients regularly receive methadone in primary care centers may also decrease negative stereotypes held by clinical staff (40), including those not providing direct care, creating a more supportive environment for patients (11, 41).
Methadone retention was similar between study arms. The 24-month retention rate observed aligns with previous findings in Ukraine and exceeds the global averages reported in a meta-analysis (61.0% at 6 months and 49.9% at 12 months) highlighting the value of ongoing tele-mentoring in maintaining clinical standards for managing complex conditions and underscoring that primary care-integrated delivery of methadone does not compromise OAT quality (32, 42). Notably, retention on methadone was higher in primary care than in specialty clinics among patients newly initiating methadone, suggesting that patients may benefit from the accessibility and continuity of care offered in primary care settings. This finding highlights the potential advantages of integrating methadone treatment into primary care, where patients can receive comprehensive, longitudinal care that supports treatment adherence and overall health management.
Existing research on integrated OAT has primarily focused on high-income settings, often evaluating buprenorphine-based models despite its higher cost relative to methadone (17). While these studies have demonstrated benefits for OUD-related outcomes such as retention in care and reductions in illicit drug use, less is known about the broader healthcare access implications of integrated OAT, particularly in LMICs where methadone remains the more accessible option. Other studies have evaluated co-location of services for OUD and HIV in low-income settings, but have not focused on comprehensive health outcomes (43, 44). A U.S.-based cohort study using QHI among patients prescribed buprenorphine found high screening rates for hypertension, HCV, and HIV, but lacked a comparison group and generalizability to LMICs (45). Our study is unique in its randomized controlled design, inclusion of a comparison arm, use of two implementation strategies (collaborative learning via tele-mentoring and pay-for-performance), and setting within a middle-income country with a high burden of OUD and HIV. Additionally, we followed participants for 24 months, providing a longer-term perspective on integrated care outcomes.
Our trial has important public health implications as it provides real-world evidence from a middle-income setting with high burden of comorbid OUD, HIV, and other infectious and non-communicable diseases. As the Eastern Europe and Central Asia region grapples with overlapping epidemics of OUD, HIV, and tuberculosis, our findings demonstrate that methadone can be effectively delivered in primary care settings with support from tele-mentoring, offering a scalable strategy for countries in the region where specialty clinics are limited and stigma, provider discomfort, and fragmented care remain barriers. Integrating OAT into primary care provides an opportunity to simultaneously improve access to life-saving therapies like methadone and link patients to preventive and chronic disease care, improving health outcomes for individuals with OUD across the region.
Limitations of the study should be noted. First, the COVID-19 pandemic and the full-scale Russian invasion in 2022 impacted the progress of the study. We continued the study during the pandemic after switching to telephone interviews. The Russian invasion temporarily disrupted study activities in Donbas and permanently in Mariupol. Recruitment was ultimately completed by adding sites in Lviv. Second, QHIs measured healthcare utilization but not health outcomes. Future research should examine whether increased utilization translates into improved clinical outcomes. Third, all QHI components were equally weighted in calculating the composite primary endpoint. Some components, however, may differentially affect health depending on individual needs (e.g., HIV services for participants with HIV). Metrics weighted by individual health priorities may better capture subgroup benefits from integrated services. Fourth, since specialty clinics required participants to complete primary care QHI services off-site, verification of receipt of these services was more difficult and may have led to underestimation of health services utilization in the control group in EMR data. This limitation applied only to EMR-derived measures and not to the self-reported QHI score, the study’s primary outcome. Finally, completion of self-reported surveys was 74% at 24 months, potentially introducing attrition bias. Our primary statistical analyses, based on linear mixed-effects models, assume a missing-at-random mechanism, but sensitivity analyses with pattern-mixture models suggested that attrition bias was unlikely to explain the observed differences.
Conclusion
Integrating methadone treatment into primary care settings improves adherence to guideline-concordant healthcare without compromising retention on methadone. Tele-mentoring–facilitated integration may provide an effective strategy for OUD management in high-burden settings. These findings are especially relevant for Eastern Europe, Central Asia, and similar regions where OUD is highly prevalent and frequently co-occurs with HIV, HCV, and tuberculosis, underscoring the potential of primary care-based models to address complex health needs.
Supplementary Material
Role of the Funding Source
This study was supported by the National Institutes on Drug Abuse (R01 DA043125). The funding source had no role in design, data collection, analysis, interpretation, or manuscript preparation.
Footnotes
Conflict of Interest
The authors declare that they have no conflicts of interest related to this work
Reproducible Research Statements
Protocol: posted as data supplement at annals website or available as a published manuscript at https://doi.org/10.1016/j.cct.2024.107690.
Statistical Code: Available to interested readers by contacting Dr. Machavariani at eteri.machavariani@yale.edu.
Data: Available to qualified researchers for the purpose of conducting additional analyses, subject to ethical approval and adherence to data use agreements. Requests can be directed to Eteri Machavariani (eteri.machavariani@yale.edu) and Kostyantyn Dumchev (dumchev@uiphp.org.ua).
References:
- 1.Degenhardt L, Grebely J, Stone J, Hickman M, Vickerman P, Marshall BDL, et al. Global patterns of opioid use and dependence: harms to populations, interventions, and future action. Lancet. 2019;394(10208):1560–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Degenhardt L, Peacock A, Colledge S, Leung J, Grebely J, Vickerman P, et al. Global prevalence of injecting drug use and sociodemographic characteristics and prevalence of HIV, HBV, and HCV in people who inject drugs: a multistage systematic review. Lancet Glob Health. 2017;5(12):e1192–e207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Larney S, Tran LT, Leung J, Santo T, Jr., Santomauro D, Hickman M, et al. All-Cause and Cause-Specific Mortality Among People Using Extramedical Opioids: A Systematic Review and Meta-analysis. JAMA Psychiatry. 2020;77(5):493–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Santo T, Jr., Clark B, Hickman M, Grebely J, Campbell G, Sordo L et al. Association of Opioid Agonist Treatment With All-Cause Mortality and Specific Causes of Death Among People With Opioid Dependence: A Systematic Review and Meta-analysis. JAMA Psychiatry. 2021;78(9):979–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.MacArthur GJ, van Velzen E, Palmateer N, Kimber J, Pharris A, Hope V, et al. Interventions to prevent HIV and Hepatitis C in people who inject drugs: a review of reviews to assess evidence of effectiveness. Int J Drug Policy. 2014;25(1):34–52. [DOI] [PubMed] [Google Scholar]
- 6.World Health Organization. Mental Health Action Plan 2013–2020. 2017. [Google Scholar]
- 7.Lagisetty P, Klasa K, Bush C, Heisler M, Chopra V, Bohnert A. Primary care models for treating opioid use disorders: What actually works? A systematic review. PloS one. 2017;12(10):e0186315. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Haddad MS, Zelenev A, Altice FL. Buprenorphine maintenance treatment retention improves nationally recommended preventive primary care screenings when integrated into urban federally qualified health centers. J Urban Health. 2015;92(1):193–213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Formánek T, Krupchanka D, Mladá K, Winkler P, Jones PB. Mortality and life-years lost following subsequent physical comorbidity in people with pre-existing substance use disorders: a national registry-based retrospective cohort study of hospitalised individuals in Czechia. The Lancet Psychiatry. 2022;9(12):957–68. [DOI] [PubMed] [Google Scholar]
- 10.Iturralde E, Slama N, Kline-Simon AH, Young-Wolff KC, Mordecai D, Sterling SA. Premature mortality associated with severe mental illness or substance use disorder in an integrated health care system. Gen Hosp Psychiatry. 2021;68:1–6. [DOI] [PubMed] [Google Scholar]
- 11.Morozova O, Dvoriak S, Pykalo I, Altice FL. Primary healthcare-based integrated care with opioid agonist treatment: First experience from Ukraine. Drug Alcohol Depend. 2017;173:132–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Louie DL, Assefa MT, McGovern MP. Attitudes of primary care physicians toward prescribing buprenorphine: a narrative review. BMC Fam Pract. 2019;20(1):157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Arora S, Kalishman S, Thornton K, Komaromy M, Katzman J, Struminger B, et al. Project ECHO (Project Extension for Community Healthcare Outcomes): A National and Global Model for Continuing Professional Development. J Contin Educ Health Prof. 2016;36 Suppl 1:S48–9. [DOI] [PubMed] [Google Scholar]
- 14.Cromwell J, Trisolini MG, Pope GC, Mitchell JB, Greenwald LM. Pay for performance in health care: Methods and approaches: RTI Press; 2011. [Google Scholar]
- 15.Mendelson A, Kondo K, Damberg C, Low A, Motuapuaka M, Freeman M, et al. The Effects of Pay-for-Performance Programs on Health, Health Care Use, and Processes of Care: A Systematic Review. Ann Intern Med. 2017;166(5):341–53. [DOI] [PubMed] [Google Scholar]
- 16.Fiellin DA, Weiss L, Botsko M, Egan JE, Altice FL, Bazerman LB, et al. Drug treatment outcomes among HIV-infected opioid-dependent patients receiving buprenorphine/naloxone. J Acquir Immune Defic Syndr. 2011;56 Suppl 1(0 1):S33–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Cooper RL, Edgerton RD, Watson J, Conley N, Agee WA, Wilus DM, et al. Meta-analysis of primary care delivered buprenorphine treatment retention outcomes. The American Journal of Drug and Alcohol Abuse. 2023;49(6):756–65. [DOI] [PubMed] [Google Scholar]
- 18.Ivasiy R, Madden LM, Johnson KA, Machavariani E, Ahmad B, Oliveros D, et al. Retention and dropout from sublingual and extended-release buprenorphine treatment: A comparative analysis of data from a nationally representative sample of commercially-insured people with opiod use disorder in the United States. International Journal of Drug Policy. 2025;138:104748. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ivasiy R, Madden LM, DiDomizio E, Johnson KA, Machavariani E, Ahmad B, et al. The cascade of care for commercially-insured persons with opioid use disorder and comorbid HIV and HCV infections. Drug Alcohol Depend. 2024;263:112410. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Carrieri PM, Michel L, Lions C, Cohen J, Vray M, Mora M, et al. Methadone induction in primary care for opioid dependence: a pragmatic randomized trial (ANRS Methaville). PLoS One. 2014;9(11):e112328. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Altice FL, Bruce RD, Lucas GM, Lum PJ, Korthuis PT, Flanigan TP, et al. HIV treatment outcomes among HIV-infected, opioid-dependent patients receiving buprenorphine/naloxone treatment within HIV clinical care settings: results from a multisite study. JAIDS Journal of Acquired Immune Deficiency Syndromes. 2011;56:S22–S32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Johnson AE, Tate MM, Filippovych M, Bromberg DJ, Kiriazova T, Madden LM, et al. Pre-Implementation Planning for Combined Rapid-Start Initiation of Antiretroviral and Opioid Agonist Therapies in People Who Inject Drugs with HIV in Ukraine: A Rights-Based Approach. Journal of the Association of Nurses in AIDS Care. 2025: In Press. [Google Scholar]
- 23.Public Health Center of the Ministry of Health of Ukraine. Opioid Agonist Therapies Statistics, 2018. [Available from: https://phc.org.ua/kontrol-zakhvoryuvan/zalezhnist-vid-psikhoaktivnikh-rechovin/zamisna-pidtrimuvalna-terapiya-zpt/statistika-zpt. [Google Scholar]
- 24.World Health Organization. Technical guide for countries to set targets for universal access to HIV prevention, treatment and care for injecting drug users 2012. [Google Scholar]
- 25.Public Health Center of the MoH of Ukraine. Information Bulletin “HIV Infection in Ukraine”. 2023. Contract No.: 54. [Google Scholar]
- 26.Sazonova Y, Duchenko G, Kovtun O, Kuzin I. Assessment of the number of key groups in Ukraine, Report. Kyiv2019 [Available from: https://aph.org.ua/wp-content/uploads/2019/06/Otsinka-chiselnosti_32200.pdf [Google Scholar]
- 27.Machavariani E, Dumchev K, Pykalo I, Filippovych M, Ivasiy R, Esserman D, et al. Design and implementation of a Type-2 hybrid, prospective randomized trial of opioid agonist therapies integration into primary care clinics in Ukraine. Contemp Clin Trials. 2024:107690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Harvey G, Kitson A. PARIHS revisited: from heuristic to integrated framework for the successful implementation of knowledge into practice. Implementation Science. 2016;11(1):33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Cromwell J, Trisolini MG, Pope GC, Mitchell JB, Greenwald LM. Pay for Performance in Health Care: Methods and Approaches. 2021. [Google Scholar]
- 31.Hasson F, Keeney S, McKenna H. Research guidelines for the Delphi survey technique. J Adv Nurs. 2000;32(4):1008–15. [PubMed] [Google Scholar]
- 32.Farnum SO, Makarenko I, Madden L, Mazhnaya A, Marcus R, Prokhorova T, et al. The real-world impact of dosing of methadone and buprenorphine in retention on opioid agonist therapies in Ukraine. Addiction. 2021;116(1):83–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.The ASAM National Practice Guideline for the Treatment of Opioid Use Disorder: 2020 Focused Update. J Addict Med. 2020;14(2S):1–91. [DOI] [PubMed] [Google Scholar]
- 34.Pashchenko O, Bromberg DJ, Dumchev K, LaMonaca K, Pykalo I, Filippovych M, et al. Preliminary analysis of self-reported quality health indicators of patients on opioid agonist therapy at specialty and primary care clinics in Ukraine: A randomized control trial. PLOS Global Public Health. 2022;2(11):e0000344. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.McCoy CE. Understanding the Intention-to-treat Principle in Randomized Controlled Trials. West J Emerg Med. 2017;18(6):1075–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Hedeker D, Gibbons RD. Application of random-effects pattern-mixture models for missing data in longitudinal studies. Psychol Methods. 1997;2(1):64–78. [Google Scholar]
- 37.R Core Team (2022). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. [Google Scholar]
- 38.Ivasiy R, Madden LM, Meteliuk A, Machavariani E, Ahmad B, Zelenev A, et al. The impact of emergency guidance to the COVID-19 pandemic on treatment entry, retention and mortality among patients on methadone in Ukraine. Addiction. 2024;119(9):1585–96. [DOI] [PubMed] [Google Scholar]
- 39.Biancarelli DL, Biello KB, Childs E, Drainoni M, Salhaney P, Edeza A, et al. Strategies used by people who inject drugs to avoid stigma in healthcare settings. Drug Alcohol Depend. 2019;198:80–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Machavariani E, Bromberg DJ, Dumchev K, Esserman D, Earnshaw VA, Pykalo I, et al. Decrease in provider stigma associated with improved Quality Health Indicators of individuals receiving methadone maintenance in primary healthcare settings in Ukraine. International Journal of Drug Policy (In Press; ). 2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Bromberg DJ, Machavariani E, Madden LM, Dumchev K, LaMonaca K, Earnshaw VA, et al. Integrating methadone into primary care settings in Ukraine: effects on provider stigma and knowledge. Journal of the International AIDS Society. 2024;27(2):e26202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Klimas J, Hamilton M-A, Gorfinkel L, Adam A, Cullen W, Wood E. Retention in opioid agonist treatment: a rapid review and meta-analysis comparing observational studies and randomized controlled trials. Systematic Reviews. 2021;10(1):216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Solomon SS, Solomon S, McFall AM, Srikrishnan AK, Anand S, Verma V, et al. Integrated HIV testing, prevention, and treatment intervention for key populations in India: a cluster-randomised trial. Lancet HIV. 2019;6(5):e283–e96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Hung V, Nguyen ST, Tieu VT, Nguyen TT, Duong TH, Lyss S, et al. Evaluation of the integrated clinic model for HIV/AIDS services in Ho Chi Minh City, Viet Nam, 2013–2014. Public Health Action. 2016;6(4):255–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.UNAIDS. Global AIDS monitoring 2019: Ukraine. UNAIDS; Geneva; 2019. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.