Skip to main content
Download PDF
Download ePub

Measuring time in buprenorphine treatment stages among people with HIV and opioid use disorder by retention definition and its association with cocaine and hazardous alcohol use

Addiction Science & Clinical Practice volume 18, Article number: 51 (2023) Cite this article

Abstract

Background

We use a novel, longitudinal approach to describe average time spent in opioid use disorder (OUD) cascade of care stages for people with HIV (PWH) and with OUD, incorporating four definitions of treatment retention. Using this approach, we describe the impact of cocaine or hazardous alcohol use on time spent retained on buprenorphine.

Methods

We followed PWH with OUD enrolled in the Johns Hopkins HIV Clinical Cohort from their first buprenorphine treatment episode between 2013 and 2020. We estimated 4-year restricted mean time spent on buprenorphine below buprenorphine retention threshold, on buprenorphine above retention threshold, off buprenorphine and in HIV care, loss to follow-up, and death. Retention definitions were based on retention threshold (180 vs 90 days) and allowable treatment gap (7 vs 30 days). Differences in 2-year restricted mean time spent retained on buprenorphine were estimated for patients with and without cocaine or hazardous alcohol use.

Results

The study sample (N = 179) was 63% male, 82% non-Hispanic Black, and mean age was 53 (SD 8) years. Patients spent on average 13.9 months (95% CI 11.4, 16.4) on buprenorphine over 4 years. There were differences in time spent retained on buprenorphine based on the retention definition, ranging from 6.5 months (95% CI 4.6, 8.5) to 9.6 months (95% CI 7.4, 11.8). Patients with cocaine use spent fewer months retained on buprenorphine. There were no differences for patients with hazardous alcohol use.

Conclusions

PWH with OUD spend relatively little time receiving buprenorphine in their HIV primary care clinic. Concurrent cocaine use at buprenorphine initiation negatively impact time on buprenorphine.

Introduction

Williams et al. [1, 2] described eight stages of the opioid use disorder (OUD) cascade of care, ranging from primary prevention (stage 0) to recovery (stage 7), to address population and individual patient needs. The stages relevant to people with OUD are (1) diagnosis of OUD, (2) linkage to care, (3) initiation of medications for OUD, (4) treatment retention, and (5) recovery. The OUD cascade of care continues to evolve. The need to refer to treatment (“link to care”) is becoming less relevant as buprenorphine, a medication for OUD, can now be prescribed by any clinician with a Drug Enforcement Agency (DEA) registration and is increasingly available on-demand in general medical and community settings.

OUD is a chronic condition often typified by periods of abstinence and relapse to opioid use. As such, patients frequently transition on and off medications for OUD, and consequently, between stages of the OUD cascade of care. This is similar to the HIV care continuum where people with HIV (PWH) can cycle on and off antiretroviral medications and between having detectable or nondetectable HIV viral loads. Lesko et al. [3] proposed a novel method to measure and visualize the HIV care continuum by accounting for patients’ transitions between stages over time [3]. This approach could be applicable to the OUD cascade of care by accounting for patients’ transitions between OUD treatment stages and describe how patients’ time is distributed across the various stages of the OUD cascade of care. Additionally, this method can be used to estimate the impact of other factors on total time in a given treatment stage. For example, studies consistently demonstrate that cocaine and methamphetamine use have a negative impact on the length of OUD treatment episodes, whereas studies on the impact of alcohol use are mixed [4, 5].

Retaining patients on medication for OUD is necessary to reduce the adverse effects of opioid use including overdose and death [ 6,7,8], but there are various definitions of retention. The American Society of Addiction Medicine’s (ASAM) OUD treatment guidelines state that patients should be on OUD treatment for > 90 days for positive long-term outcomes and the National Quality Forum (NQF) set a quality metric for patients who initiate OUD treatment of 180-day retention with no gaps > 7 days [9,10,11].

Herein, we applied this novel, longitudinal method of describing time in treatment stages to PWH and OUD. We focused on PWH since treating OUD is critical to reaching the “Ending the HIV Epidemic” goals [12, 13] and medications for OUD decreases opioid use, mortality, and HIV transmission in this population [14,15,16,17]. We aimed to describe overall time spent on buprenorphine, time spent on buprenorphine above the retention threshold, and how estimates of time spent retained on buprenorphine differ based on the definition of retention. Lastly, we described the association between co-occurring hazardous alcohol use or cocaine use at buprenorphine initiation and time spent retained on a medication for OUD.

Methods

Study sample

The Johns Hopkins HIV Clinical Cohort (JHHCC) consists of PWH ages 18 years and older who are receiving HIV primary care at the John G Bartlett Specialty Practice (“Bartlett clinic”) in Baltimore, Maryland, who consent to share their data [18]. In 2013, the Bartlett clinic started office-based buprenorphine treatment for patients with OUD where buprenorphine could be initiated on-demand. The current study is a retrospective cohort analysis of all patients in the JHHCC with an OUD diagnosis who received their first buprenorphine prescription in the Bartlett clinic between January 1st 2013 and December 31st 2020 (to ensure patients had at least 1-year of follow-up). Each patient was followed from the start of their first buprenorphine prescription in the clinic in the study period until the first occurrence of the following events: censoring at the end of the observation period on December 31st, 2021, after 4 years of follow-up, or death. We administratively censored time at 4 years to ensure there were enough patients under follow-up to provide stable inference. This study was approved by a Johns Hopkins Institutional Review Board.

Outcome measurement

We focused on the following buprenorphine treatment stages as shown in Fig. 1: on buprenorphine, which was further categorized into current treatment episode durations (1) below and (2) above retention thresholds; (3) off buprenorphine and lost to follow-up (LTFU); (4) off buprenorphine and retained in HIV care at the Bartlett Clinic; and (5) death. Time spent “retained” on buprenorphine only accrued after a particular treatment episode surpassed the retention threshold. For example, for a 90-day retention threshold, if a buprenorphine treatment episode was 98 days, the first 90 days would be counted in stage (1) and the last 8 days would be counted in stage (2) and we would say the patient spent 8 days retained on buprenorphine. For a 180-day retention threshold, all 98 days would be accounted in stage (1) and we would say the patient spent 0 days retained on buprenorphine. A participant’s “initial” buprenorphine prescription was their first prescription in the Bartlett clinic and not necessarily their first lifetime treatment attempt. Consistent with national quality metric measurements, patients who received an early refill were assumed to finish their existing supply before starting the new prescription [19]. Prescriptions received on the same day were assumed to be taken concurrently (e.g. buprenorphine 8mg X 14 days and buprenorphine 12mg X 14 days duration was counted as 14 covered days). We only counted buprenorphine prescriptions that were dispensed as sublingual formulations to restrict ourselves to buprenorphine for the treatment of OUD and not for pain management [9]. Data on buprenorphine prescriptions that originated outside of the Bartlett clinic and Johns Hopkins Health system were not available for research unless a patient reported the prescription to their HIV care team and the care team recorded the prescription in the patient’s medical record. Methadone is another medication for OUD and is only available in federally licensed opioid treatment programs. Federal regulations limit opioid treatment program data sharing and the JHHCC cannot capture methadone treatment administrative data including start dates, dosages, or end dates.

Fig. 1
figure 1

Depiction of buprenorphine treatment stages and direction of possible transitions. By definition, all participants start follow up when they initiate buprenorphine and are therefore in box 1. If the treatment episode duration is above the retention threshold (e.g., 90 or 180 days), they move to box 2. If they are in box 1 or 2 and stop buprenorphine and are lost to follow up, they move to box 3. If they are in box 1 or 2 and stop buprenorphine and are retained in HIV clinic they are in box 4. Participants can move between being lost to follow up (box 3) and retained in care (4). If after stopping buprenorphine (box 3 or 4) they restart buprenorphine they would return to box 1. Death (box 5) is a competing (terminal) event

Full size image

Lost to follow-up (LTFU) was defined as a ≥ 12-month gap between clinical visits or buprenorphine prescriptions. Patients were classified as returned to care after being LTFU if they attended a clinical visit or received a buprenorphine prescription from the clinic. Patient death dates were ascertained from clinic sources (e.g., emergency contacts or checks against the local health information exchange) and matches against the Social Security Death Index and National Death Index.

Defining treatment episodes and retention duration thresholds

A buprenorphine treatment episode was defined as the continuous period from an initial prescription start date to either a gap in buprenorphine prescriptions, or the end of the last prescription, whichever came first [10]. We evaluated four treatment retention definitions, ordered from the most strict to least strict criteria for retention. First, 180 days of continuous treatment without a > 7-day prescription gap based on the NQF definition [10, 19]. Second, 180 days of continuous treatment without a > 30-day prescription gap. The 30-day gap to define a new treatment episode has been used previously to evaluate buprenorphine treatment durations [20]. Third, 90 days of continuous treatment without a > 7-day gap. Fourth, 90 days of continuous treatment without a > 30-day gap. The 90-day threshold was based off the ASAM treatment guidelines which suggest that 90 days of continuous treatment is associated with improved outcomes [9]. The ASAM guidelines do not describe the gap in prescriptions that would indicate a new treatment episode.

Covariates

Recent hazardous alcohol use and cocaine use

We explored whether hazardous alcohol and cocaine use at the initial buprenorphine prescription was associated with time spent retained on buprenorphine for the 2 years after initiating buprenorphine. We focused on the 2-year period after initiation because hazardous alcohol or cocaine use at initiation is more likely to affect the outcomes closer to initial buprenorphine prescription. We focused on hazardous alcohol and cocaine use for this demonstration because of the high prevalence of use of these substances in the cohort and because they were measured with good accuracy in our cohort [21, 22]. Consistent with patterns of drug use in Baltimore, there was very low (< 1%) prevalence of methamphetamine use in our cohort, so we did not consider it in this analysis. Patients were classified as having recent cocaine or hazardous alcohol use if they self-reported substance use or had substance use indicated on chart abstraction in the period 9 months before to 3 months after the initial buprenorphine prescription in the clinic. Briefly, chart abstractions are conducted for every contiguous 6-month period that patients are in care, with trained staff reviewing clinical notes to identify evidence of recent substance use (e.g., notes in a problem list). Patients self-report substance use approximately every 6 months on tablet computers in conjunction with a clinic visit. Patients report past 3-month use (yes or no) of cocaine on the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) and past 12-month alcohol use on the Alcohol Use Disorders Identification Test-Consumption (AUDIT-C) [23, 24]. Hazardous alcohol use was defined as an AUDIT-C score of ≥ 3 for women and ≥ 4 for men, where sex at birth was used to determine the cut-off. Combing chart abstraction and patient self-report improved the detection of substance use compared to using either method in isolation [21].

Baseline demographic and clinical covariates

We described the cohort based on demographic and clinical characteristics at the start of analytic follow-up. Gender, race, ethnicity, birth year, and HIV acquisition risk factor were collected upon enrollment into the JHHCC. Patients self-reported the most likely source of their HIV infection at enrollment into HIV care. Risk factors include injection drug use, being a man who had sex with men, or heterosexual sex. Risk factors are not mutually exclusive and may not reflect ongoing risk behaviors.

Statistical analyses

Our analytic approach replicates approaches to visualize time in different HIV care continuum stages [3, 25, 26]. Briefly, to estimate the time in each stage, we first calculated the cumulative incidence of each instance of the following events, where the time origin is the initial buprenorphine prescription:

  1. A.

    Stopping buprenorphine when the treatment episode was below the retention duration threshold (i.e., 90 or 180 days)

  2. B.

    Buprenorphine treatment episode crosses the retention duration threshold

  3. C.

    Stopping buprenorphine when treatment episode was greater than or equal to the retention duration threshold

  4. D.

    Re-starting buprenorphine after stopping

  5. E.

    LTFU after buprenorphine initiation

  6. F.

    No longer LTFU after buprenorphine initiation

  7. G.

    Death after buprenorphine initiation

By design all participants entered the study on buprenorphine. Participants could experience multiple instances of each of the above events with the exception of death. We estimated cumulative incidence functions separately for each instance of each event (e.g., the 3rd instance of re-starting buprenorphine after stopping). We used the Aalen-Johansen estimator to account for death as a competing event when calculating risk of events A–F [27]. The proportion of the study population in each stage at each time t was calculated by adding and subtracting the cumulative incidence functions at each time t as follows:

  1. 1)

    Dead after buprenorphine = cumulative incidence of death after buprenorphine initiation

  2. 2)

    On buprenorphine with treatment duration below retention threshold = 1 minus the sum of cumulative incidence functions for time to stopping buprenorphine treatment when episode duration is below the retention duration threshold plus sum of cumulative incidence functions for re-starting buprenorphine.

  3. 3)

    On buprenorphine with treatment duration above retention threshold = sum of cumulative incidence functions for time to buprenorphine treatment episode when duration above the retention threshold duration minus the sum of cumulative incidence functions for time to stopping buprenorphine when treatment episode is equal to or above the retention duration threshold.

  4. 4)

    Off buprenorphine and LTFU = sum of cumulative incidence of time to LTFU after buprenorphine initiation minus the sum of cumulative incidence of no longer LTFU

  5. 5)

    Off buprenorphine and retained in HIV care = 1 minus the proportion in stages 1–4 in this list.

  6. 6)

    On buprenorphine of any duration = sum of on buprenorphine below the retention duration threshold (2) and on buprenorphine with treatment duration equal to or greater than the retention duration threshold (3).

The sum of the proportions of the population in stages 1–5 is 1 and proportions can be presented as a set of stacked curves representing the distribution of the cohort across the stages over time. The area under each curve graphed separately corresponds to the restricted mean time spent in each buprenorphine treatment stage over the 4 years of follow-up after an initial buprenorphine prescription.

In secondary analyses, we compared the time spent retained on buprenorphine restricted to the 2 years of follow-up after an initial buprenorphine prescription, stratified by recent hazardous alcohol use or cocaine use at the time of the initial buprenorphine treatment episode. Because our goal with this analysis is descriptive, we present unadjusted differences [28, 29]. To handle missing data on recent hazardous alcohol and cocaine use at the initial buprenorphine initiation, we assumed data were missing at random and used a method described by Schomaker & Heumann (2018) combining bootstrap estimation with multiple imputation [30,31,32]. Each bootstrap sample was used to create 40 complete imputed datasets, an optimal number of imputations based on preliminary analysis [33]. To estimate 95% confidence intervals (CI) for estimates of the restricted mean time and restricted mean time differences, we took the 2.5th and 97.5th percentiles of estimates from 2000 bootstrap resamples of the data. The outcome of interest was calculated for each of the imputed dataset and the mean was calculated to give a single estimate for each bootstrap sample. This process was repeated for each of the four retention definitions. All analyses were conducted in R [34, 35].

Results

Study sample

Between January 1st 2013 and December 31st 2020, there were 179 participants who initiated buprenorphine treatment. Among these participants, 18% (n = 33) died during follow-up. There were 63% (n = 112) participants who identified as male and 82% (n = 147) who identified as non-Hispanic Black. Mean age at the start of the initial buprenorphine prescription was 52.8 (SD 8.0) years (Table 1). Among the participants, 56% (n = 101) reported injection drug use and 68% (n = 122) reported heterosexual sex as their possible routes of HIV transmission. A > 7-day gap (retention definitions 1 and 3) in prescriptions denoting a new treatment episode resulted in 410 buprenorphine treatment episodes while a > 30-day gap (retention definitions 2 and 4) resulted in 325 buprenorphine treatment episodes. The percentage of episodes where the duration met or exceeded the retention definition was 24% (n = 99), 32% (n = 104), 39% (n = 160), and 48% (n = 157) for retention definitions 1 through 4, respectively. Table 2 shows the number of buprenorphine treatment episodes per patient, duration, and gap between the end of one treatment episode and the beginning of the next episode, if a subsequent episode was observed.

Table 1 Demographic and clinical characteristics of participants with HIV and opioid use disorder who initiated clinic-based buprenorphine in the Johns Hopkins HIV Clinical Cohort, 2013–2020
Full size table
Table 2 Buprenorphine treatment episode characteristics for >7-day and >30-day prescription gap thresholds indicating a new treatment episode
Full size table

Time spent in treatment stages

Over course of 4 years after initiation, the mean time on buprenorphine was 13.9 months (95% CI 11.4, 16.4), which represents 29% (95% CI 24%, 34%) of follow-up time. Participants spent 18.9 months (95% CI 15.6, 21.7), or 39% (95% CI 33%, 45%) of the time, off buprenorphine and retained in HIV care and 11.2 months (95% CI 8.5, 14.1), or 23% (95% CI 18%, 29%) of the time, off buprenorphine and lost to clinic follow-up. Participants who died contributed on average 4.2 months (95% CI 2.5, 6.3), or 9% (5%, 13%) of their follow-up time, in this state.

Table 3 presents the amount of time spent on buprenorphine above the retention thresholds. Retention definition 1 (≥ 180 days continuous treatment without any gaps > 7 days) was the most strict retention definition and resulted in an average of 6.5 months (95% CI 4.6, 8.5) spent retained on buprenorphine. Retention definition 4 (≥ 90 days of continuous treatment without any gaps > 30 days) was the least strict retention definition and resulted in an average of 9.6 months (95% CI 7.4, 11.8) spent retained on buprenorphine. Table 4 presents the pairwise differences in average time spent retained between retention definitions. Consider time spent retained for retention definitions 1 and 3, which both use a > 7-day gap to define the end of a treatment episode. The estimated difference in average time spent retained on buprenorphine between retention definition 1 and retention definition 3 is 1.9 (95% CI 1.2, 2.8) months. Another interpretation is that if all treatment episodes were > 180 days, the expected difference between time spent retained for definitions 1 and 3 would be 3 months (90 days). Since we observed less of a difference, it shows that patients who make it to the 90-day threshold, on average have an additional 1.9 months (95% CI 1.2, 2.8) on treatment before their treatment episode ends, or about 1.1 (95% CI 0.2, 1.8) months short of the 180-day threshold.

Table 3 Average number of months and percent of time spent on buprenorphine above and below retention thresholds based on retention definitions each opioid use disorder treatment stage restricted to 4 years after initiating buprenorphine treatment among patients with HIV and opioid use disorder by retention definition in the Johns Hopkins HIV Clinical Cohort, 2013–2020
Full size table
Table 4 Difference in estimated average number of months spent retained on buprenorphine between each of the four retention definitions
Full size table

Visualization of time spent in treatment stages

Figure 2 depicts these data for retention definition 1. The area between curves is the total amount of time that an individual spends in each stage on average. The interpretation at a given time point is the probability of a person being in each stage. For example, at 180 days after initiating buprenorphine treatment, the probability of a patient having been on buprenorphine for ≥ 180 days is 25% (dark purple). At 1.75 years, the probability of a patient being on buprenorphine of any treatment episode length decreases to 20% (light purple), which steadily increases to 25% by year 4. Additional file 1: Fig. S1 presents the stacked proportions for all retention definitions and shows, across all definitions of retention, in years 1–4, fewer than 25% of patients are on buprenorphine and retained.

Fig. 2
figure 2

Proportion of participants with HIV and opioid use disorder who initiated clinic-based buprenorphine in each opioid use disorder treatment stage in the Johns Hopkins HIV Clinical Cohort restricted to 4-year follow-up, with retention defined as ≥ 180 days of treatment without a > 7-day gap

Full size image

Time spent retained in treatment, stratified by cocaine or hazardous alcohol use

Out of 179 participants, 32% (n = 57) had recent cocaine use and 22% (n = 39) of participants had recent hazardous alcohol use (Table 1). Data were missing on recent cocaine or hazardous alcohol use for 10% (n = 17) and 8% (n = 14) of participants, respectively.

Table 5 shows the average time spent retained on buprenorphine over 2 years after initiation for each of the retention definitions overall and stratified by recent cocaine or hazardous alcohol use. Similar to the results discussed above, the average time retained on buprenorphine over 2 years after initiation was lowest when using the most strict retention definition 1 (≥ 180 days of treatment without > 7-day gap; 6.4 months [95% CI, 4.1, 8.9]) and highest for the least strict definition 4 (≥ 90 days of treatment without > 30-day gap; 10.4 months [95% CI 7.5, 13.7]).

Table 5 Average number of months spent retained on buprenorphine above the retention thresholds overall and by the presence of cocaine or hazardous alcohol use at initiation restricted to 2 years after initiating buprenorphine treatment among participants with HIV and opioid use disorder in the Johns Hopkins HIV Clinical Cohort, 2013-2020
Full size table

For retention definition 1 (≥ 180 days of treatment without a > 7-day gap), recent cocaine use was associated with 1.8 (95% CI, 0.1, 3.5) fewer months spent retained on buprenorphine over 2 years after first buprenorphine initiation. Hazardous alcohol use was associated with 0.9 (95% CI − 1.0, 2.8) fewer months spent retained on buprenorphine over 2 years after initiation. For all retention definitions, cocaine use was associated with a significant reduction in the average time spent on buprenorphine over 2 years after buprenorphine initiation. No statistically significant differences were observed for hazardous alcohol use across all definitions.

Discussion

Applying a novel method accounting for multiple treatment episodes and transitions between treatment stages in a retrospective study of 179 PWH initiating buprenorphine for the treatment of OUD in an office-based setting, integrated within a HIV specialty clinic, we found very low probability of being retained on buprenorphine across 4 years of follow-up. In addition, consistent with a prior study, we found that the definition of retention directed impacted the estimates of number of treatment episodes, treatment episode durations, and time spent on buprenorphine [36]. These results highlight the need to understand not only the barriers to retention in OUD care cascade but also how the definition of retention influences our conclusions.

Two factors constituted each retention definition. First, the threshold for what is considered retained and second, the gap between prescriptions that would constitute a new treatment episode. There is limited research on the optimal length of buprenorphine treatment. A simulation study based on a cohort of patients with OUD in the United States Veterans Health Administration suggests 2 months of buprenorphine treatment and 4 months of methadone was needed to reduce all-cause mortality [11]. While reducing mortality is a very important goal, recovery from OUD is the final stage of the OUD cascade of care and people might require longer treatment durations to attain recovery. The National Institute of Drug Abuse’s (NIDA) Principles of Drug Addiction [37] states that most people being treated for a substance use disorder need 3 months in treatment to reduce drug use and achieve optimal outcomes. For this reason, the ASAM treatment guidelines state that there is limited effectiveness of treatment durations less than 90 days, but state that there is no recommended time limit for pharmacological treatment [9]. Although the NQF committee identified a duration of 180 days for retention, it did not recommend this measure be used in pay-for-performance due to our limited understanding of optimal treatment durations. A NIDA Clinical Trails Network study “Optimizing Retention, Duration, and Discontinuation Strategies for Opioid Use Disorder Pharmacotherapy” (CTN-0100) is ongoing to explore the effect of treatment duration on post-discontinuation outcomes [38].

Second, retention depends on the gap in treatment that defines a new treatment episode. The NQF committee specifies a > 7-day gap in treatment in their performance measure, while the ASAM and NIDA reports do not specify a gap. From a pharmacologic perspective, buprenorphine can suppress cravings for up to 2 days after last dose [39] and patients would likely need to undergo re-induction after this period, particularly if they have been using other opioids like heroin or fentanyl. Practically, many patients use buprenorphine obtained from friends, family, the illicit market, or left over from previous prescriptions in a self-directed manner when there are gaps in their prescriptions [40,41,42]. The relatively large difference in number of treatment episodes based on a 7- or 30-day gap in prescriptions suggest that patients often have gaps in prescriptions. Numerous studies describe the multi-level barriers to OUD treatment for PWH [43,44,45] and patients are often report ambivalence about medications for OUD, largely due to the associated stigma and perceptions about what it means to be “in recovery” [46, 47]. It is important to note that the present analysis is based on buprenorphine prescriptions, not dispensed medication, and there could be delays in filling medications relative to the prescription date, which would lead us to underestimate the actual gap in treatment and overestimate treatment duration and retention.

These factors aside, we found that regardless of the retention definition, the average amount of time an individual participant spent on buprenorphine and retained in this clinic was low—13.9 months over the 4 years after buprenorphine initiation. Based on visual inspection of the graphs, the probability that a patient will be on buprenorphine at a given point in the 1.5 to 4 years after initiation is less than 25% across all retention definitions suggesting that long-term retention on buprenorphine originating from the clinic is low in this cohort. This finding is consistent with results from similar office-based settings. In a single-site, primary care office-based opioid treatment clinic in New York City from 2006 to 2013, 1-year buprenorphine treatment retention was about 20% [48]. Earlier studies of primary-care office based opioid treatment had 1-year retention rates as high as 50% [49, 50]. A large sample of 1237 patients at a safety-net office-based opioid treatment program found that 45% of patients were retained for ≥ 1 year. Our results continue to highlight the need to develop and test interventions to improve patient retention on buprenorphine in outpatient settings [51].

We sought to demonstrate a potential application of this method by describing the association between cocaine or hazardous alcohol use and time spent retained on buprenorphine. While results should be considered exploratory, we observed significantly less time retained on buprenorphine among participants with cocaine use, consistent with previous research [4, 5]. Notably, differences were found for all retention definitions suggesting this finding does not depend on the retention definition. Hazardous alcohol use at buprenorphine initiation did not result in statistically significant differences in a time spent retained on buprenorphine. Multiple substance use is a common clinical situation and these results show that different patterns of substance use likely impact buprenorphine retention in different ways [52, 53]. It is possible that patients with cocaine or hazardous alcohol use might be referred more often to higher levels of care (e.g., specialty outpatient or inpatient addiction treatment settings). Given that these treatment episodes are not documented in our data, patients referred to a higher level of care might continue buprenorphine treatment outside the clinic. This could lead to a misclassification in our analysis, implying they are off buprenorphine. Such misclassification could result in overestimating retention differences when compared to patients without cocaine or hazardous alcohol use, who might be more inclined to continue treatment within the Bartlett clinic.

This study is limited in that only buprenorphine prescriptions that were prescribed or recorded by clinicians in the Bartlett clinic were included. Patients could receive buprenorphine through other clinics or federally licensed opioid treatment programs. However, the Bartlett clinic has characteristics of low-threshold treatment, including treatment on demand, and patients who remain active in HIV care in the clinic would have the opportunity to receive buprenorphine from the clinic [54]. We suspect that patients are likely to use the Bartlett clinic as their primary source of OUD treatment with buprenorphine. Incorporating prescription drug monitoring program and opioid treatment program medication dispensing data would yield a more comprehensive picture of patient transitions between treatment settings and stages within the OUD cascade of care. Additionally, we only had data on buprenorphine prescriptions and not whether those prescriptions were dispensed or whether dispensation was delayed. Inclusion of pharmacy dispensing would improve the accuracy of our estimates. Finally, our findings are based on data from an office-based opioid treatment program in one HIV care clinic and may not be generalizable across all care settings. In particular, because Baltimore has a low prevalence of methamphetamine use, buprenorphine use patterns may be different in our clinic compared to locations with a higher prevalence of methamphetamine use.

Despite these limitations, we believe this novel approach of estimating time in OUD treatment stages could improve our understanding of OUD treatment and factors that impact retention on medications for OUD. Future applications of this approach could expand to include other clinically-relevant treatment stages including tapering, early discontinuation before meeting a retention threshold, or transitioning to a different medication for OUD (e.g., methadone). In addition, recent data suggests that Black and other minoritized communities have disproportionately high opioid overdose deaths and disparities in treatment with medications for OUD [55, 56]. Evaluating how other factors, including the local drug supply (e.g., fentanyl saturation), state policies, other common substance use (e.g., cannabis), and patient demographics impact time in treatment stages could help inform individual and public health interventions.

Conclusion

We demonstrate how a novel approach can account for patient transitions between buprenorphine treatment stages to estimate the average time spent in each stage over time. Participants spent less than 25% of their time on buprenorphine treatment in the same clinic in the 4 years following buprenorphine initiation. For participants with recent cocaine use at initiation, less time is spent retained on buprenorphine. Our results highlight the necessity of incorporating multiple data sources, including prescription drug monitoring program and opioid treatment program data, into clinical research. Future studies that incorporate diverse, rich data sources could provide a comprehensive assessment of patients’ time in treatment stages and be used to explore optimal treatment durations and factors associated with retention on medications for OUD.

Availability of data and materials

The datasets generated and/or analyzed during the current study are not publicly available due their containing information that could compromise the privacy of research participants as incorporated in the informed consent, but are available from the corresponding author on reasonable request.

References

  1. Williams AR, Nunes EV, Bisaga A, Levin FR, Olfson M. Development of a cascade of care for responding to the opioid epidemic. Am J Drug Alcohol Abuse. 2019;45:1–10.

    PubMed  PubMed Central  Google Scholar 

  2. Williams AR, Johnson KA, Thomas CP, Reif S, Socías ME, Henry BF, et al. Opioid use disorder cascade of care framework design: a roadmap. Subst Abuse. 2022;43:1207–14.

    Google Scholar 

  3. Lesko CR, Edwards JK, Moore RD, Lau B. A longitudinal, HIV care continuum: 10-year restricted mean time in each care continuum stage after enrollment in care, by history of IDU. AIDS. 2016;30:2227–34.

    PubMed  Google Scholar 

  4. Blondino CT, Gormley MA, Taylor DDH, Lowery E, Clifford JS, Burkart B, et al. The Influence of Co-occurring substance use on the effectiveness of opiate treatment programs according to intervention type. Epidemiol Rev. 2020;42:57–78.

    PubMed  Google Scholar 

  5. Bunting AM, Krawczyk N, Choo T-H, Pavlicova M, McNeely J, Tofighi B, et al. Polysubstance use before and during treatment with medication for opioid use disorder: Prevalence and association with treatment outcomes. J Subst Abuse Treat. 2022. https://0-doi-org.brum.beds.ac.uk/10.1016/j.jsat.2022.108830.

    Article  PubMed  Google Scholar 

  6. Nielsen S, Larance B, Degenhardt L, Gowing L, Kehler C, Lintzeris N. Opioid agonist treatment for pharmaceutical opioid dependent people. Cochrane Database Syst Rev. 2016. https://0-doi-org.brum.beds.ac.uk/10.1002/14651858.CD011117.pub2.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Sordo L, Barrio G, Bravo MJ, Indave BI, Degenhardt L, Wiessing L, et al. Mortality risk during and after opioid substitution treatment: systematic review and meta-analysis of cohort studies. BMJ. 2017;357:j1550.

    PubMed  PubMed Central  Google Scholar 

  8. Larochelle MR, Bernson D, Land T, Stopka TJ, Wang N, Xuan Z, et al. Medication for opioid use disorder after nonfatal opioid overdose and association with mortality: a cohort study. Ann Intern Med. 2018. https://0-doi-org.brum.beds.ac.uk/10.7326/M17-3107.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Cunningham C, Edlund MJ, Fishman M, Gordon AJ, Jones HE, Kampman KM, et al. The ASAM national practice guideline for the treatment of opioid use disorder: 2020 focused update. J Addict Med. 2020;14:1–91.

    Google Scholar 

  10. National Quality Forum. Behavioral health 2016–2017: technical report [Internet]. Washington DC; 2017. https://www.qualityforum.org/Publications/2017/08/Behavioral_Health_2016-2017_Final_Report.aspx.

  11. Ching JH, Owens DK, Trafton JA, Goldhaber-Fiebert JD, Salomon JA. Impact of treatment duration on mortality among Veterans with opioid use disorder in the United States Veterans Health Administration. Addict Abingdon Engl. 2021;116:3494–503.

    Google Scholar 

  12. Fauci AS, Redfield RR, Sigounas G, Weahkee MD, Giroir BP. Ending the HIV Epidemic: a plan for the United States. JAMA. 2019;321:844.

    PubMed  Google Scholar 

  13. Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the Use of Antiretroviral Agents in Adults and Adolescents with HIV. Department of Health and Human Services. https://clinicalinfo.hiv.gov/en/guidelines/adult-and-adolescent-arv

  14. MacArthur GJ, Minozzi S, Martin N, Vickerman P, Deren S, Bruneau J, et al. Opiate substitution treatment and HIV transmission in people who inject drugs: systematic review and meta-analysis. BMJ. 2012;345:e5945.

    PubMed  PubMed Central  Google Scholar 

  15. Fanucchi L, Springer SA, Korthuis PT. Medications for treatment of opioid use disorder among persons living with HIV. Curr HIV/AIDS Rep. 2019;16:1–6.

    PubMed  PubMed Central  Google Scholar 

  16. Kim J, Lesko CR, Fojo AT, Keruly JC, Moore RD, Chander G, et al. The effect of buprenorphine on human immunodeficiency virus viral suppression. Clin Infect Dis. 2021;73:1951–6.

    CAS  PubMed  PubMed Central  Google Scholar 

  17. Lucas GM, Chaudhry A, Hsu J, Woodson T, Lau B, Olsen Y, et al. Clinic-based treatment of opioid-dependent HIV-infected patients versus referral to an opioid treatment program. Ann Intern Med. 2010;152:704–11.

    PubMed  PubMed Central  Google Scholar 

  18. Moore RD. Understanding the clinical and economic outcomes of HIV therapy: the Johns Hopkins HIV clinical practice cohort. J Acquir Immune Defic Syndr Hum Retrovirology Off Publ Int Retrovirology Assoc. 1998;17(Suppl 1):S38-41.

    Google Scholar 

  19. National Quality Forum. Continuity of Pharmacotherapy for Opioid Use Disorder (OUD). 2017. Report No.: NQF 3175. https://qpp.cms.gov/docs/QPP_quality_measure_specifications/CQM-Measures/2019_Measure_468_MIPSCQM.pdf.

  20. Stein BD, Landis RK, Sheng F, Saloner B, Gordon AJ, Sorbero M, et al. Buprenorphine treatment episodes during the first year of COVID: a retrospective examination of treatment initiation and retention. J Gen Intern Med. 2023;38:733–7.

    PubMed  Google Scholar 

  21. Lesko CR, Keil AP, Moore RD, Chander G, Fojo AT, Lau B. Measurement of current substance use in a cohort of HIV-infected persons in continuity HIV Care, 2007–2015. Am J Epidemiol. 2018;187:1970–9.

    PubMed  PubMed Central  Google Scholar 

  22. Crane HM, Lober W, Webster E, Harrington RD, Crane PK, Davis TE, et al. Routine collection of patient-reported outcomes in an HIV clinic setting: the first 100 patients. Curr HIV Res. 2007;5:109–18.

    CAS  PubMed  Google Scholar 

  23. Bush K, Kivlahan DR, McDonell MB, Fihn SD, Bradley KA. The AUDIT alcohol consumption questions (AUDIT-C): an effective brief screening test for problem drinking. Arch Intern Med. 1998;158:1789–95.

    CAS  PubMed  Google Scholar 

  24. Higgins-Biddle JC, Babor TF. A review of the alcohol use disorders identification test (AUDIT), AUDIT-C, and USAUDIT for screening in the United States: past issues and future directions. Am J Drug Alcohol Abuse. 2018;44:578–86.

    PubMed  PubMed Central  Google Scholar 

  25. Edwards JK, Cole SR, Adimora A, Fine J, Martin J, Eron J. Illustration of a measure to combine viral suppression and viral rebound in studies of HIV therapy. JAIDS J Acquir Immune Defic Syndr. 2015;68:241–4.

    CAS  PubMed  Google Scholar 

  26. Gouskova NA, Cole SR, Eron JJ, Fine JP. Viral suppression in HIV studies: combining times to suppression and rebound. Biometrics. 2014;70:441–8.

    PubMed  PubMed Central  Google Scholar 

  27. Cole SR, Lau B, Eron JJ, Brookhart MA, Kitahata MM, Martin JN, et al. Estimation of the standardized risk difference and ratio in a competing risks framework: application to injection drug use and progression to AIDS after initiation of antiretroviral therapy. Am J Epidemiol. 2015;181:238–45.

    PubMed  Google Scholar 

  28. Lesko CR, Fox MP, Edwards JK. A framework for descriptive epidemiology. Am J Epidemiol. 2022. https://0-doi-org.brum.beds.ac.uk/10.1093/aje/kwac115.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Kaufman JS. Statistics, adjusted statistics, and maladjusted statistics. Am J Law Med. 2017;43:193–208.

    PubMed  Google Scholar 

  30. Schomaker M, Heumann C. Bootstrap inference when using multiple imputation. Stat Med. 2018;37:2252–66.

    PubMed  PubMed Central  Google Scholar 

  31. Efron B, Tibshirani R. An introduction to the bootstrap. New York: Chapman & Hall; 1993.

    Google Scholar 

  32. Rubin DB. Multiple imputation after 18+ years. J Am Stat Assoc. 1996;91:473–89.

    Google Scholar 

  33. von Hippel PT. How many imputations do you need? A two-stage calculation using a quadratic rule. Sociol Methods Res. 2020;49:699–718.

    Google Scholar 

  34. R Core Team. R: A language and environment for statistical computing. Vienna: Foundation for Statistical Computing; 2021.

    Google Scholar 

  35. Buuren S van, Groothuis-Oudshoorn K. mice : Multivariate Imputation by Chained Equations in R. J Stat Softw. 2011. http://www.jstatsoft.org/v45/i03/. Accessed 21 Jul 2022.

  36. Dong H, Stringfellow EJ, Russell WA, Bearnot B, Jalali MS. Impact of alternative ways to operationalize buprenorphine treatment duration on understanding continuity of care for opioid use disorder. Int J Ment Health Addict. 2022. https://0-doi-org.brum.beds.ac.uk/10.1007/s11469-022-00985-w.

    Article  Google Scholar 

  37. National Institutes of Health, National Institute on Drug Abuse. Principles of drug addiction treatment: a research-based guide (Third Edition). 2018. https://health.gov/healthypeople/tools-action/browse-evidence-based-resources/principles-drug-addiction-treatment-research-based-guide-third-edition. Accessed 5 Apr 2023.

  38. Shulman M, Weiss R, Rotrosen J, Novo P, Costello E, Nunes EV. Prior national drug abuse treatment clinical trials network (CTN) opioid use disorder trials as background and rationale for NIDA CTN-0100 “optimizing retention, duration and discontinuation strategies for opioid use disorder pharmacotherapy (RDD).” Addict Sci Clin Pract. 2021;16:15.

    PubMed  PubMed Central  Google Scholar 

  39. Greenwald M, Johanson C-E, Bueller J, Chang Y, Moody DE, Kilbourn M, et al. Buprenorphine duration of action: mu-opioid receptor availability and pharmacokinetic and behavioral indices. Biol Psychiatry. 2007;61:101–10.

    CAS  PubMed  Google Scholar 

  40. Kavanaugh PR, McLean K. Motivations for diverted buprenorphine use in a multisite qualitative study. J Drug Issues. 2020;50:550–65.

    Google Scholar 

  41. Gandhi P, Rouhani S, Park JN, Urquhart GJ, Allen ST, Morales KB, et al. Alternative use of buprenorphine among people who use opioids in three U.S. Cities. Subst Abuse. 2022;43:364–70.

    CAS  Google Scholar 

  42. McLean K, Kavanaugh PR. “They’re making it so hard for people to get help:” Motivations for non-prescribed buprenorphine use in a time of treatment expansion. Int J Drug Policy. 2019;71:118–24.

    PubMed  Google Scholar 

  43. Morgan JR, Quinn EK, Chaisson CE, Ciemins E, Stempniewicz N, White LF, et al. Potential barriers to filling buprenorphine and naltrexone prescriptions among a retrospective cohort of individuals with opioid use disorder. J Subst Abuse Treat. 2022;133:108540.

    CAS  PubMed  Google Scholar 

  44. Kleinman RA, Morris NP. Federal barriers to addressing the opioid epidemic. J Gen Intern Med. 2020;35:1304–6.

    PubMed  PubMed Central  Google Scholar 

  45. Kennedy-Hendricks A, Bandara S, Merritt S, Barry CL, Saloner B. Structural and organizational factors shaping access to medication treatment for opioid use disorder in community supervision. Drug Alcohol Depend. 2021;226:108881.

    CAS  PubMed  Google Scholar 

  46. Thakrar AP, Pytell JD, Stoller KB, Walters V, Weiss RD, Chander G. Transitioning off methadone: a qualitative study exploring why patients discontinue methadone treatment for opioid use disorder. J Subst Use Addict Treat. 2023;150:209055.

    CAS  PubMed  Google Scholar 

  47. Krawczyk N, Negron T, Nieto M, Agus D, Fingerhood MI. Overcoming medication stigma in peer recovery: a new paradigm. Subst Abuse. 2018;39:404–9.

    Google Scholar 

  48. Bhatraju EP, Grossman E, Tofighi B, McNeely J, DiRocco D, Flannery M, et al. Public sector low threshold office-based buprenorphine treatment: outcomes at year 7. Addict Sci Clin Pract. 2017;12:7.

    PubMed  PubMed Central  Google Scholar 

  49. Alford DP, LaBelle CT, Kretsch N, Bergeron A, Winter M, Botticelli M, et al. Collaborative care of opioid-addicted patients in primary care using buprenorphine. Arch Intern Med. 2011;171:425–31.

    PubMed  PubMed Central  Google Scholar 

  50. Soeffing JM, Martin LD, Fingerhood MI, Jasinski DR, Rastegar DA. Buprenorphine maintenance treatment in a primary care setting: outcomes at 1 year. J Subst Abuse Treat. 2009;37:426–30.

    PubMed  Google Scholar 

  51. Chan B, Gean E, Arkhipova-Jenkins I, Gilbert J, Hilgart J, Fiordalisi C, et al. Retention strategies for medications for opioid use disorder in adults: a rapid evidence review. J Addict Med. 2021;15:74–84.

    PubMed  Google Scholar 

  52. Substance Abuse and Mental Health Services Administration. Key substance use and mental health indicators in the United States: Results from the 2021 National Survey on Drug Use and Health. Rockville, MD: Center for Behavioral Health Statistics and Quality, Substance Abuse and Mental Health Services Administration; 2022 Dec. Report No.: (HHS Publication No. PEP22-07-01-005, NSDUH Series H-57. https://www.samhsa.gov/data/

  53. Lesko CR, Falade-Nwulia OO, Pytell JD, Hutton HE, Fojo AT, Keruly JC, et al. Joint effects of substance use disorders and recent substance use on HIV viral non-suppression among people engaged in HIV care in an urban clinic, 2014–2019. Addiction. 2023. https://0-doi-org.brum.beds.ac.uk/10.1111/add.16301

  54. Jakubowski A, Fox A. Defining low-threshold buprenorphine treatment. J Addict Med. 2020;14:95–8.

    PubMed  PubMed Central  Google Scholar 

  55. Kariisa M. Vital Signs: drug overdose deaths, by selected sociodemographic and social determinants of health characteristics—25 states and the district of Columbia, 2019–2020. MMWR Morb Mortal Wkly Rep. 2022. https://0-doi-org.brum.beds.ac.uk/10.15585/mmwr.mm7129e2.

    Article  PubMed  PubMed Central  Google Scholar 

  56. Andraka-Christou B. Addressing racial and ethnic disparities in the use of medications for opioid use disorder. Health Aff Proj Hope. 2021;40:920–7.

    Google Scholar 

Download references

Acknowledgements

Not applicable

Funding

This study was supported by grants from the National Institutes of Health (K24 AA027483, K01 AA028193, U01 DA036935, R01 90090700, R01 DA052297, P30 AI094189, and P01 AA029544).

Author information

Authors and Affiliations

  1. Division of General Internal Medicine, Department of Medicine, University of Colorado School of Medicine, Mail Stop B180, 12631 E. 17Th Ave, Aurora, CO, 80045, USA

    Jarratt D. Pytell

  2. Division of General Internal Medicine, Johns Hopkins School of Medicine, Baltimore, MD, USA

    Anthony T. Fojo

  3. Division of Infectious Diseases, Johns Hopkins University School of Medicine, Baltimore, MD, USA

    Jeanne C. Keruly, LaQuita N. Snow, Oluwaseun Falade-Nwulia & Richard D. Moore

  4. Division of General Internal Medicine, University of Washington School of Medicine, Seattle, WA, USA

    Geetanjali Chander

  5. Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA

    Catherine R. Lesko

Authors
  1. Jarratt D. Pytell
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Anthony T. Fojo
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Jeanne C. Keruly
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. LaQuita N. Snow
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Oluwaseun Falade-Nwulia
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Richard D. Moore
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Geetanjali Chander
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Catherine R. Lesko
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

JDP: conceptualization, formal analysis, writing-original draft. ATF: conceptualization, methodology, writing—review & editing. JCK: project administration, data curation, writing-review & editing. LNS: project administration, data curation, writing-review & editing. OFN: conceptualization, methodology, writing—review & editing. RDM: conceptualization, funding acquisition, writing—review & editing. GC: conceptualization, funding acquisition, writing—review & editing. CRL: conceptualization, methodology, writing—review & editing, supervision.

Corresponding author

Correspondence to Jarratt D. Pytell.

Ethics declarations

Ethics approval and consent to participate

Patient participants provide written consent to be included in the Johns Hopkins HIV Clinical Cohort. A Johns Hopkins Institute Review Board has approved the study.

Consent for publication

Not applicable.

Competing interests

The authors have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1:

Figure S1. Depiction of treatment buprenorphine treatment states for each of the retention definitions.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Pytell, J.D., Fojo, A.T., Keruly, J.C. et al. Measuring time in buprenorphine treatment stages among people with HIV and opioid use disorder by retention definition and its association with cocaine and hazardous alcohol use. Addict Sci Clin Pract 18, 51 (2023). https://0-doi-org.brum.beds.ac.uk/10.1186/s13722-023-00408-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://0-doi-org.brum.beds.ac.uk/10.1186/s13722-023-00408-8

Keywords

Addiction Science & Clinical Practice

ISSN: 1940-0640