Systematic review and meta-analysis of retention in treatment using medications for opioid use disorder by medication, race/ethnicity, and gender in the United States

Highlights

• Gender and race/ethnicity are not associated with differences in outcomes between three medications for opioid use disorder.

• African Americans who were recruited for buprenorphine treatment were retained longer than those not recruited.

• Methadone programs that require psychosocial counseling are associated with lower treatment retention when compared to medication alone.

Abstract

Background and Aims

There is mounting evidence that opioid use disorder is experienced differently by people of different genders and race/ethnicity groups. Similarly, in the US access to specific medications for opioid use is limited by gender and race/ethnicity. This study aims to evaluate if gender or race/ethnicity is associated with different rates of treatment retention in the US, for each of three medications used to treat opioid use disorder.

Methods

A systematic search was conducted using PubMed, CINHAL, and PsychINFO, databases. All studies that provided a ratio of those retained in treatment at a specified time in terms of gender and/or race/ethnicity and medication were included. Variables were created to assess the effects of time in treatment, recruited sample, required attendance at concurrent psychosocial treatment, and adherence to strict rules of conduct for continuation in treatment on retention. Meta-analytical and meta-regression methods were used to compare studies on the ratio of those who completed a specific time in treatment by race/ethnicity group and by gender.

Results

Nineteen articles that provided the outcome variable of interest were found (11 buprenorphine, six methadone, and two naltrexone). Meta-analyses found that treatment retention was similar for all gender and racial/ethnic groups for all three medications. Meta-regression found that those of the African American group who were recruited into buprenorphine treatment were retained significantly longer than African Americans in buprenorphine treatment who were studied retrospectively. Also, both genders had significantly lower retention in methadone treatment when there was the additional requirement of psychosocial therapy.

Introduction

Opioid use disorder (OUD) is a major health problem in the United States, effecting not only the 1.62 million people who qualify for this diagnosis (Substance Abuse and Mental Health Services Administration, 2019) but also society as a whole, through OUD morbidity and mortality, disease transmission, crime, and health care utilization (Stotts, Dodrill, & Kosten, 2009). Medications for the treatment of OUD (MOUD) have become the gold standard mode of treatment. MOUD either takes the form of agonist therapy or the administration of a medication that completely blocks the effect of opioids (antagonist) (Mancher & Leshner, 2019). Three MOUDs are approved by the Food and Drug Administration in the United States: (Longo & Schuckit, 2016) buprenorphine, methadone, (agonists) and naltrexone (an antagonist). All three are associated with lower all-cause and opioid related mortality in a recent meta-analysis (Sordo et al., 2017). The three MOUDs have different physiological effects and create three distinct treatment experiences in the United States.

Methadone therapy typically requires, daily or quasi-daily, attendance at dedicated, governmentally certified substance use disorder treatment clinics to avoid diversion of the medication for illicit use (Mancher & Leshner, 2019). This structure stigmatizes those in treatment, limits their employment opportunities and requires parents of small children to arrange for childcare when they are attending treatment. In short, methadone is effective but places clinic visits central to the identity and lifestyle of those in treatment (Doukas, 2011).

Buprenorphine has been demonstrated to have positive outcomes similar to methadone (Mattick, Breen, Kimber, & Davoli, 2014). However, the reduced incentive for diversion permits primary care physicians to treat patients once a month, allowing those in treatment to pursue full-time employment and does not require daily childcare to participate. Buprenorphine’s opioid blocking action only lasts for 12–24 h which means that a person will be able to return to illicit opioid use by skipping a dose of their medication (Blanco-Gandía & Rodríguez-Arias, 2018). This possibility is exacerbated by the lack of structured support provided by more frequent contact with the healthcare system that exists in methadone treatment.

Naltrexone treatment relies on blocking opioid receptors and eliminating the rewarding experience of their use (Mancher & Leshner, 2019). It has two acknowledged limitations: a requirement for detoxification (with possible withdrawal symptoms) that deters engagement (Blanco-Gandía & Rodríguez-Arias, 2018) and that oral naltrexone has been associated with unpleasant physical side-effects (Carroll et al., 2018). These limitations have been overcome by the creation of extended release naltrexone, an injection, which is effective for one month, with minimal side effects (Blanco-Gandía & Rodríguez-Arias, 2018).

People who belong to different gender and/or racial/ethnic groups experience OUD differently. Gender and race/ethnicity also influence a persons' roles and responsibilities within society. These differences in conjunction with the distinct experience of each of the MOUDs which require different biopsychosocial supports for success may imply that specific medications may be more effective for certain people than for others.

Previous research has found that the experience of females with OUD is different than males in several ways including the increased prevalence of a history of trauma (McHugh, Votaw, Sugarman, & Greenfield, 2018), anxiety, and depression (Cleveland & Bonugli, 2014). Becker, McClellan, and Reed (2017) also found that women attempting to gain access to treatment face greater barriers than men. These differences may be due to the legacy of sexism and the embedded patriarchal society which exists in the United States. It may be hypothesized that these same forces also create conditions in which one type of treatment may be more effective for females than males.

Similarly, one MOUD may be more compatible for an individual's lifestyle based on their gender or race/ethnicity group. For example, Mancher and Leshner (2019) proposed that since women are often the primary caregivers for younger children, childcare issues may deter female treatment participation. This may be extended to the possibility that reducing the need for childcare to attend treatment may influence higher success rates for women in buprenorphine or naltrexone treatment since they require fewer clinic visits than methadone.

Previous research has found that there are differences in the experience of African Americans who use opioids compared to their White counterparts. These differences include a higher prevalence of heroin use (as opposed to prescription opioids) (Pouget, Fong, & Rosenblum, 2018) and differences in the history of the current opioid epidemic. At the start of the century, African American mortality from opioid overdose remained relatively stable even during the period that opioid mortality increased rapidly among White Americans (Alexander et al., 2018, Furr‐Holden et al., 2021). This has changed recently with the introduction of synthetic opioids, especially fentanyl, into the illicit opioid market, and now African American opioid death rates have surpassed those of White Americans. The differences in these trends may influence differences in treatment needs that are more consistent with one MOUD compared to the others.

Additionally, many studies assume that there will be differences OUD treatment outcomes based on demographics primarily race/ethnicity (Manhapra et al., 2017, McHugh et al., 2013, Proctor et al., 2015), but not many explore the causes of these differences though James and Jordan (James & Jordan, 2018) hypothesize that some of the differences may be due to cultural insensitivity of existing programs that dissuade African Americans from engaging or remaining in treatment. We may also speculate that marginalized communities may develop coping mechanisms that are different than the dominant society that may be more compatible with a specific MOUD; such differences have been found in previous research in other minority ethnic communities (Bart, Wang, Hodges, Nolan, & Carlson, 2012).

This study chose to use treatment retention as the primary outcome measure since it is associated with reduced illicit drug use, criminal behavior, and risky sexual behaviors (Bart, 2012). As such, retention “is a key measure of success in opioid treatment programs and is considered a significant indicator of treatment effectiveness” (Amiri et al., 2018). Treatment retention is also a ubiquitous measure and is frequently reported by gender and race/ethnicity.

Though treatment retention is a valid outcome measure, a recent review found that not only were retention rates widely dispersed (3–90.7%), the time over which the studies measured retention varied considerably (from one to 12 months) (Timko, Schultz, Cucciare, Vittorio, & Garrison-Diehn, 2016). Additionally, there are factors intrinsic to the study design that may encourage or discourage retention. Studies that recruited participants or those who use volunteers may have participants who are more committed to treatment than those who were evaluated retrospectively. Conversely, studies that required participation in other psychosocial treatment may improve retention or may decrease retention due to the added burden of more frequent attendance. Studies that require adherence to a ridged code of behavioral conduct (such as abstinence from all illicit substance use) for continuation in the study may also decrease treatment retention. Furthermore, treatment retention as a proportion will change based on the length of treatment and every month of continued treatment allows for more dropout that may not continue as a linear progression.

The aim of this study is to conduct a meta-analysis of the rates of treatment retention for each of the three MOUDs within and between gender and race/ethnicity groups to determine if societal or cultural factors lead to a specific medication being more effective for a specific group of people. Then, meta-regression was used to assess the effect that length of study has on treatment retention as well as the hypothesis that a recruited sample, required participation in psychosocial counseling, and adherence to strict rules for retention will account for some of the heterogeneity in results found in the included studies.