Initiation of Low-Threshold Buprenorphine in Non-Treatment Seeking Patients with Opioid Use Disorder Engaged in HCV Treatment

Abstract

Objective:

The ANCHOR program offered buprenorphine treatment to people who inject drugs engaged in hepatitis C (HCV) treatment at a Washington, DC harm reduction organization. This analysis describes the program model and outcomes of the opioid care continuum at one year.

Methods:

Primary outcomes were initiation of buprenorphine and retention in care, defined by an active buprenorphine prescription at given time points. Secondary outcomes included treatment interruptions, reasons for treatment non-initiation and termination, buprenorphine and opiate use, and HIV risk behaviors. Buprenorphine and opiate use were measured by urine toxicology screens and HIV risk behavior was quantified using a validated survey.

Results:

Of 67 patients receiving HCV treatment not on opioid agonist therapy (OAT) at baseline, 96% (n=64) were interested and 73% (n=49) initiated buprenorphine. Retention was 82% (n=40), 65% (n=32), and 59% (n=29) at months 1, 6, and 12, respectively. Retention at 12 months was associated with self-reported engagement in routine medical care (p<0.01), but was not associated with gender, stable housing, past OAT, or past overdose. Among retained patients, urine screens positive for opioids were 73% (n=29), 56% (n=18), and 79% (n=23) at months 1, 6, and 12. There was a significant mean decrease in HIV risk taking behavior scores over the treatment period, primarily driven by reduced injection frequency.

Conclusions:

Patients engaged in HCV treatment at a harm reduction organization showed a high rate of initiation of buprenorphine treatment, with retention comparable to other treatment settings. While most patients continued using opioids on treatment, there was a reduced frequency of injection drug use, a significant driver of OUD-related risk. These data support the use of low-threshold buprenorphine access alongside HCV treatment to reduce morbidity and mortality in people with OUD.

Introduction

The opioid epidemic and increasing prevalence of synthetic opioids in the drug supply continue to cause significant morbidity and mortality in the United States.¹ According to the Centers for Disease Control (CDC), 69,615 individuals died of drug overdose in 2018.² Among the estimated 2.0 million individuals in the U.S. with an opioid use disorder, the majority do not receive treatment.^3,4,5

An effective response to the ongoing opioid epidemic will require engaging more people with opioid use disorder (OUD), including people who inject drugs (PWID), in addiction treatment. Buprenorphine, a partial opioid agonist, is one evidence-based treatment for OUD that has shown to be effective in retaining patients in treatment, decreasing illicit opioid use, and lowering risk of all-cause and overdose mortality.^6,7,8,9

Treatment models integrating buprenorphine into other medical care, including primary care, emergency department visits, and inpatient medicine, have expanded access to opioid agonist therapy (OAT) outside traditional addiction treatment settings.¹⁰ Beyond expanding access, offering buprenorphine to PWID in the setting of co-located medical care has also been shown to improve uptake and retention as compared to traditional standalone OAT programs.^11,12

However, engaging the greatest number of PWID will require interventions outside of a traditional health system-based approach. Although PWID often report interest in buprenorphine treatment, many do not know where to access buprenorphine and are less likely to engage in regular outpatient care than people who do not use drugs.^13,14 Stigma, insurance barriers, financial barriers, and lack of access to transportation and other resources, all inhibit PWID from seeking medical care.^15,16,17,18 Notably, PWID engaged in harm reduction services, particularly those who have experienced barriers to buprenorphine access in the past, prefer buprenorphine treatment delivered at a harm reduction organization.¹⁹

A core principle of harm reduction is meeting people where they are, at whatever stage they are in the recovery process. As such, expanding access to buprenorphine for PWID within this framework may allow for more successful engagement of PWID in treatment of OUD. Methadone maintenance programs first applied the harm reduction framework to OAT delivery by providing OAT regardless of ongoing opiate use, and demonstrated retention rates similar to regular threshold programs with significant reductions in mortality and HIV risk.^20,21,22 In New York City and Philadelphia, buprenorphine programs integrated into harm reduction organizations engaged individuals accessing other services and demonstrated retention comparable to traditional outpatient medical settings.^23,24 In recent years, low-threshold buprenorphine programs aimed to engage marginalized populations of PWID, including criminal system-involved and homeless individuals, have also been developed.^25,26 Features of these low-threshold programs include provision of medical care in locations where PWID already access services, elimination of urine toxicology screens as a barrier to treatment, lack of penalization for missed appointments, and acceptance of treatment goals that do not include abstinence.

This analysis explores the creation of a buprenorphine clinic co-located with an existing hepatitis C (HCV) treatment clinic, embedded in a harm reduction drop-in center. The clinic was established with the goal of combining the benefits of co-locating buprenorphine treatment and other medical care with the advantages of providing OAT in a low-threshold, harm reduction-focused setting. This evaluation describes the program model, patient population, and buprenorphine-related outcomes within the first year.

Methods

The Clinic

The ANCHOR HCV clinic was established in 2016 in collaboration with a harm reduction organization with longstanding ties to and advocacy for marginalized populations in Washington DC, including sex workers, people who inject drugs, and transgender individuals. The organization’s drop-in center provides services including syringe exchange, naloxone distribution, support groups, housing and legal case work, laundry, showers, and community lunch.

A room in the drop-in center was transformed into a medical suite and staffed with physicians, nurses, and research assistants from the study team, as well as community health workers (CHW) hired through the organization. In May 2017, patients were offered initation of buprenorphine co-located with HCV treatment as part of the ANCHOR investigation (NCT03221309)²⁷, a pilot study to evaluate a comprehensive model of HCV treatment in people who inject opioids.

Eligibility and Intake

The ANCHOR study enrolled patients with chronic HCV, OUD, and self-reported injection drug use (IDU) within three months. Patients interested in HCV treatment were recruited from drop-in center clientele and CHW-driven HCV screening in the community. All enrolled patients received direct-acting antiviral (DAA) therapy for HCV. Following initiation of DAA, patients not already on OAT were offered initiation of collocated buprenorphine treatment, which could occur at any point within the 24 week time period from DAA initiation to assessment of HCV cure. Patients who initiated buprenorphine treatment were then followed for a year to assess buprenorphine treatment outcomes.

Initiation of buprenorphine was first discussed with a CHW and physician at the ANCHOR screening visit and interest was continually reassessed throughout the patient’s HCV treatment course. If the patient was interested in starting buprenorphine, they completed an intake with the physician, which included evaluation of the clinical criteria for OUD,²⁸ review of the patient’s past history and treatment of OUD, and current goals and potential challenges. Initially, patients could not initiate buprenorphine treatment if they were found to have hazardous alcohol use by AUDIT-C or benzodiazepine use, however these exclusions were removed after September 2017 guidance by the Food and Drug Administration (FDA).²⁹

All patients not engaged in mental health care were referred to off-site mental health services prior to buprenorphine induction, with CHW assistance in scheduling the intake visit. Patients were also referred to an on-site Narcotics Anonymous (NA) meeting. Attendance at mental health or NA appointments was not required for buprenorphine induction or continuation.

Induction

After completing buprenorphine intake with a physician, patients were scheduled for induction one week later. This delay was required to allow time for completion of insurance prior authorizations, which were required at the time by DC Medicaid and Medicaid managed care organizations. Referral to mental health care did not impact timing of inductions. Initially all inductions took place in-office, however over time patients were allowed to initiate via home induction, per patient preference. A CHW would help coordinate transportation to in-office inductions if needed. Initial prescriptions were for one week, with a dose of 16mg daily.

Pharmacy and Insurance Coverage

Buprenorphine was paid for through patients’ insurance. This was facilitated by high rates of insurance coverage in Washington DC, regardless of income level. Most prescriptions were processed through the same local pharmacy, where a relationship and standard operating procedure was developed. Financial assistance was available for co-pays as needed by the study.

Follow-up

At each follow-up visit, the patient saw a physician to discuss their progress and provide a urine sample to screen for buprenorphine, opiates, amphetamines, cocaine, benzodiazepines, and phencyclidine. It was expected that buprenorphine was present in each urine screen, but patients were not penalized for urine drug screens positive for other drugs. Of note, presence of fentanyl was not captured in the urine drug screens. Decreased drug use was recognized as a successful outcome of OUD treatment, in alignment with the clinic’s harm reduction framework. Follow-up visits and prescriptions were initially weekly, but were decreased in frequency based on the physician’s clinical determination.

Probation and Discontinuation

If a patient had one negative urine buprenorphine test, the physician would discuss this with the patient at the next visit. In cases where the patient ran out of medication early because they were taking more doses than prescribed, this was addressed and the daily dose was adjusted as needed. If the patient could not offer a satisfactory explanation for the negative result, the patient was informed that a second negative result would trigger required directly observed therapy (DOT), both to ensure the patient was taking medication appropriately and to validate the laboratory results. Usually, DOT was required for three consecutive days. If the patient attended all days, they could return to regular follow-up intervals. Patients who did not attend DOT visits were discontinued from the program, with option to reengage in buprenorphine treatment at a later date.

Measures

At screening, all patients completed a comprehensive baseline epidemiologic survey. This survey captured patient characteristics, including housing status, education level, history of overdose, and drug use frequency and behaviors. Engagement in healthcare was assessed in this baseline survey with the question “Do you have a regular provider who you see at least once a year for routine medical care?” Patients also completed the AUDIT-C for alcohol use during screening, with scores of 3 or greater in women, and 4 or greater in men considered hazardous alcohol use.

Patients completed the the Darke HIV Risk-Taking Behavior Survey (HRBS) at Day 0 and weeks 4, 12, 24, 48 and 72 of HCV treatment.³⁰ This validated survey evaluates both injection drug use and sexual risk behaviors. Injection frequency and needle reuse were evaluated using responses to the drug use subsection. The HRBS surveys completed prior and closest to initiation of buprenorphine and end of treatment were used for pre- and post-treatment analysis. End of treatment was considered the timepoint of buprenorphine termination for individuals who discontinued buprenorphine prior to the end of the study period, or 12 months after initiation for those who remained on buprenorphine for the entire study period and beyond.

All surveys were verbally administered by a CHW or research assistant and documented in a secure online database in real time.

Outcomes

The primary outcomes were buprenorphine initiation and retention at 1 month, 6 months, and 12 months. Initiation was defined as completion of a buprenorphine induction. Retention at 1 month was defined as an active buprenorphine prescription at day 30. Retention at 6 and 12 months were defined as an active buprenorphine prescription within 7 days of those time points. Non-retention at an earlier time point did not preclude a patient from being categorized as retained at later time point if they re-engaged in treatment. Active buprenorphine prescriptions were validated utilizing the state’s Physician Drug Monitoring Program (PDMP).

Secondary outcomes included treatment interruptions and reasons for treatment non-initiation and termination. A treatment interruption was defined as a period of 7 days or more without an active buprenorphine prescription. Periods of less than seven days without an active buprenorphine prescription were not categorized as interruptions. There was no maximum treatment interruption that precluded patients from resuming treatment. Total treatment duration for each patient was defined as total days between buprenorphine initiation and buprenorphine end of treatment over a one year time period. Active treatment duration was total treatment duration minus cumulative days of treatment interruptions. Active treatment percentage (active treatment duration/total treatment duration) for each patient was calculated, and the median of the percentages for all patients is reported.

Reasons for treatment non-initiation and termination were determined from physician notes, or from public records in the cases of incarceration or death.

Each individual’s urine drug screen results were aggregated to a percentage of urine screen results positive for buprenorphine, opioids, and other drugs. Urine drug screens positive for buprenorphine but negative for norbuprenorphine were considered negative. The median and IQR of these summary percentages were determined. Percentage of urine drug screens positive for opioids at the visit closest to Day 30, month 6 and month 12 of buprenorphine treatment were also determined.

Analysis

Chi-squared tests were used to evaluate associations between categorical variables and 12-month retention. Paired t-tests were used to evaluate for significance of mean change in HRBS risk scores.

Results

Buprenorphine Initiation

Of 67 patients receiving HCV treatment not on OAT at baseline, 64 (96%) were interested in buprenorphine, and 49 (73%) started buprenorphine (Figure 1), 6 (12%) via home induction. The majority (36, 74%) of patients initiated buprenorphine during the first month of HCV treatment, while 3 (6%), 6 (12%), and 4 (8%) initiated treatment during the second month, third month, and after end of HCV treatment respectively. Among patients who were interested in buprenorphine but did not start (n=15), the leading barriers to initiation were hazardous alcohol use (7, 47%), benzodiazepine use (2, 13%), and lack of insurance (1, 7%). Three patients started methadone (20%) at another local clinic, 1 patient was no longer interested (7%), and 1 patient was lost to follow up (7%).

Figure 1.

Buprenorphine treatment cascade of care (n=67)

Patient Characteristics

The majority of the 49 patients who initiatied treatment identified as male (80%), black (98%), and had a median age of 57 years old. Most patients had previously been prescribed OAT (78%), and were injecting daily or more (74%) at baseline (Table 1).

Table 1.

Baseline Patient Characteristics

Characteristic
Sex, n (%)
Male	39 (80)
Female	10 (20)
Age in years, median (IQR)	57 (54, 62)
Race, n (%)
Black	48 (98)
More than one	1 (2)
Housing, n (%)
Stable	25 (51)
Unstable	24 (49)
Income source, n (%)
None	28 (57)
Government benefits	17 (35)
Job, pension, working under the table, other	4 (8)
Education, n (%)
8th grade or less	2 (4)
Some High School	16 (33)
High School/GED	20 (41)
Some College/2 year degree	9 (18)
College graduate	0 (0)
Master’s or higher	2 (4)
Ever incarcerated, n (%)
No	3 (6)
Yes	46 (94)
At least annual visits with a provider for routine medical care, n (%)
No	24 (49)
Yes	25 (51)
Prior prescribed OAT treatment, n (%)
No	11 (22)
Yes	38 (78)
Prior non-prescribed OAT use, n (%)
No	8 (16)
Yes	41 (84)
Age at first injection drug use, median (IQR)	20 (17.8, 30)
Injection frequency, n (%)
Daily or more	36 (74)
More than once a week, but less than daily	7 (14)
Once a week or less	6 (12)
Receptive sharing of injecting equipment 3 months prior to screening, n (%)
No	32 (65)
Yes	17 (35)
Enrolled in syringe exchange program, n (%)
No	7 (14)
Yes	42 (86)
Active medical insurance, n (%)	49 (100)

Retention

Of the 49 patients who initiated treatment, 40 (82%) were retained at 1 month, 32 (65%) were retained at 6 months, and 29 (59%) were retained at 1 year (Figure 2).

Figure 2.

Buprenorphine treatment retention, interruption, and termination

Of the 20 patients who terminated buprenorphine treatment prior to 1 year, 9 (45%) patients had negative urine buprenorphine screens and did not attend subsequent DOT appointments, 5 (25%) were incarcerated, 2 died (10%), 2 chose to discontinue (10%), 1 was terminated from the program after repeatedly filling opiate and benzodiazepine prescriptions (5%), and 1 was lost to follow-up (5%) (Figure 2).

There was a significant association between a report of seeing a provider at least once a year for routine medical care and being retained on buprenorphine treatment for 12 months (p<0.01). Retention at 12 months was not significantly associated with gender, stable housing, having a working phone, past OAT treatment, or past overdose. (Table 2).

Table 2:

Characteristics associated with buprenorphine retention at 12 months

Characteristic	Retained at 12 months (n,%)	Not Retained at 12 months (n,%)	p-value
Baseline Characteristics
Gender			0.44
Female	7 (70)	3 (30)
Male	22 (56)	17 (44)
Regular medical provider			0.0025
No	9 (37.5)	15 (62.5)
Yes	20 (80)	5 (20)
Stable Housing			0.64
No	15 (62.5)	9 (37.5)
Yes	14 (56)	11 (44)
Working phone			0.08
No	3 (33)	6 (67)
Yes	26 (65)	14 (35)
Past OAT treatment			0.73
No	22 (58)	16 (42)
Yes	7 (64)	4 (36)
Past Overdose			0.46
No	10 (53)	9 (47)
Yes	19 (63)	11 (37)
Overdose during study			0.56
No	25 (61)	16 (39)
Yes	4 (50)	4 (50)
Baseline opioid injection frequency			0.85
Less than daily	21 (60)	14 (40)
Daily or more	8 (57)	6 (43)
Incarcerated in 12 months prior to
baseline			0.65
No	22 (61)	14 (39)
Yes	7 (54)	6 (46)
Incarcerated during study			0.16
No	24 (65)	13 (35)
Yes	5 (42)	7 (58)

Treatment Interruptions

Nineteen (39%) patients had at least one interruption in treatment of 7 days or more. Among those patients, there was a median of 1 (IQR: 1, 2) interruption lasting a median of 54 days (IQR: 27, 119). Median active percentage of treatment was 85% (56, 93).

Urine Drug Screen Results

There was a median of 18 (IQR: 10, 20) urine drug screens collected per patient. The median percentage of urine drug screens containing buprenorphine, non-prescribed opioids, and other drugs per patient were 89% (IQR: 75–100), 71% (IQR: 50–100), and 43% (13,84) respectively. The median percentage of urine drug screens positive only for prescribed medications was 11% (0,33) per patient.

At the visit prior to buprenorphine induction, 39 (80%) patients had a urine screen containing non-prescribed opioids. At treatment months 1, 6, and 12, the number of patients with urine drug screens containing opioids was 29 (73%), 18 (56%), and 23 (79%), respectively. At the same time points, the number of patients with urine drug screens containing buprenorphine was 32 (80%), 28 (88%), and 23 (79%) (Figure 3). Of the 34 (70%) patients who ever had a urine screen that did not contain opioids, there was a median time of 32 days (IQR: 12, 75) until first negative screen.

Figure 3.

Urine Drug Screen Results

HIV Risk Behavior

Forty-seven (96%) patients completed the HRBS prior to buprenorphine induction and end of buprenorphine treatment. The average total risk score was 5.5 ± 3.1 (mean ± SD) prior to buprenorphine induction and 2.8 ± 3.1 (mean ± SD) prior to end of treatment, with a significant mean decrease of 2.7 points (95% CI: 1.4–4.0, p=0.0002). The average drug-risk sub-score was 4.4 ± 2.6 (mean ± SD) prior to buprenorphine induction and 1.7 ± 2.7 (mean ± SD) prior to end of treatment, with a significant mean decrease of 2.7 points (95% CI: 1.5–3.9, p=0.00003). Patients with a urine drug screen containing opioids at 12 months also demonstrated significant mean decrease of 2.45 points (95% CI: 0.78–4.1, p=0.006) in average total risk score and 2.95 points (95% CI: 1.4–4.5) in drug-risk sub-score over the same period.

Injection frequency was the risk behavior with the largest magnitude of change. Prior to initiation of treatment 32 (68%) patients were injecting daily or more. Prior to end of buprenorphine treatment, only 4 (9%) patients reported injecting daily or more. Needle re-use also decreased. Prior to buprenorphine initiation, 14 (30%) patients reported needle re-use, while 7 (15%) patients reported needle reuse prior to end of buprenorphine treatment. No patients reported using needles after other individuals prior to buprenorphine treatment initiation or at the end of buprenorphine treatment.

Discussion

The ANCHOR study demonstrates that a novel model of buprenorphine care co-located with HCV treatment in a harm-reduction setting was successful in engaging and retaining non-treatment seeking PWID in OAT. HCV treatment may be an opportune time to engage PWID with OUD into treatment, as PWID have high rates of HCV infection and high interest in HCV treatment.^31,32 Further, the prospect of HCV cure may be a motivating time to engage in behavioral change.³³ Among PWID undergoing HCV treatment in this study, interest in initiating buprenorphine was very high, with 96% of those not on OAT at baseline interested in initiating medication for addiction treatment. Of those interested, over three-fourths initiated buprenorphine treatment. These results reinforce that a substantial portion of PWID accessing other healthcare services can be successfully engaged into buprenorphine treatment,^34,35 and that engagement in HCV treatment could be leveraged to facilitate initiation of buprenorphine in a non-treatment seeking population.

The ANCHOR study also supports previous literature indicating that buprenorphine treatment provided in harm-reduction settings can show retention rates similar to care provided in more traditional settings. In our population, retention at 1, 6, 12 months was 82%, 65%, and 59% respectively. These retention rates are similar to past studies of buprenorphine treatment provided in harm reduction programs, including at Prevention Point in Philadelphia (65% at 6 months and 56% at 12 months)¹⁴ and the Harm Reduction Coalition in New York City (63% at 6 months and 42% at 12 months).¹⁵ Studies of buprenorphine retention in more traditional settings have shown a comparable range of retention at 6 (49%−57%) and 12 (61–62%) months.^36,37,38

Of note, there was no significant difference in retention between patients who had stable housing and those who did not. This challenges clinical guidelines, which suggest that the stability of a patient’s recovery environment should factor into the clinical determination of the appropriateness of buprenorphine treatment.³⁹

While our retention rates are similar to those described in more traditional models, it is important to note that our program’s retention at each time point includes patients who likely would have been removed from more traditional programs. Our model embraced a harm-reduction framework, where complete abstinence was not the goal of treatment when it did not align with the goals of the patient. Therefore, patients were not penalized or removed from the program for urine drug screens with opioids present, and 30% of patients never had a urine drug screen without opioids.

However, this does not suggest that retention of those patients is without benefit. While the majority of patients continued to have urine drug screens containing opioids, injection frequency and HIV-risk taking behaviors were significantly decreased between baseline and end of buprenorphine treatment, consistent with prior results of low-threshold methadone treatment.²² Significant reduction in HRBS score was preserved in the subset of patients with a urine screen containing opioids at month 12. As decreased injection and drug use frequency are associated with decreased risk of overdose, HIV and HCV acquisition, and development of endocarditis and skin and soft tissue infections,^40,41 reductions in these behaviors confer morbidity and mortality benefits that should not be discounted.

There are several limitations to this analysis. Our small sample size limits the ability to determine the association of several factors with retention, including incarceration. In addition, 39% of patients had interruptions in treatment with subsequent resumption of care. These patients may not be considered to be retained in other studies. Despite this, we felt that the ability to re-engage after interruption is noteworthy in such a marginalized population. Furthermore, the lack of a control arm limits our ability to show a causal relationship between engagement in HCV treatment and buprenorphine initiation and retention. The HRBS was administered at specified time points aligned with HCV rather than buprenorphine treatment, limiting our ability to track changes in the score over consistent time intervals. The HRBS was verbally administered by CHWs and research assistants, and while continued IDU was not stigmatized by the study team, reports still could have been subject to social desirability bias. Finally, injection frequency and HIV-risk behaviors are an indirect measure of risk for infectious disease complications of IDU. Larger studies that directly measure the impact of buprenorphine provided within a harm reduction framework on these outcomes are needed.

There is a disconnect between the scientific consensus that opioid use disorder is a chronic relapsing condition which can benefit from pharmacotherapy, and traditional treatment programs that silo OUD treatment, and hold complete remission as a necessary pre-requisite for continued treatment. As supported by 2018 FDA guidance, a reduction of disease severity, as represented by a reduction in the risk-taking behaviors responsible for many of the co-morbidities associated with opioid use disorder, is a desirable treatment outcome on its own merit.⁴² As evidenced by this study, providing OUD treatment within a harm reduction framework confers a substantial and sustained benefit to patients, even if complete remission is not achieved. Further, while treatment of OUD has historically occurred independent of primary and infectious disease medical care, integrated care models have increasingly allowed marginalized patients to access multiple services at once, often with improved retention and outcomes. This study reinforces this concept, demonstrating that offering co-location of OUD care into HCV care can result in high rates of buprenorphine initiation, sustained treatment engagement and reduced risk behavior among non-treatment seeking PWID with OUD. Critically, this co-located care was all offered in a low-threshold environment, outside of a traditional medical center. Combatting the opioid epidemic in the United States will require not just a shift in treatment goals towards a harm reduction framework, but will also require a shift in treatment delivery to improve comfort and engagement for non-treatment seeking people with OUD.