Using Machine Learning to Predict Treatment Adherence in Patients on Medication for Opioid Use Disorder.

OBJECTIVE: Patients receiving medication for opioid use disorder (MOUD) may continue using nonprescribed drugs or have trouble with medication adherence, and it is difficult to predict which patients will continue to do so. In this study, we develop and validate an automated risk-modeling framework to predict opioid abstinence and medication adherence at a patient's next attended appointment and evaluate the predictive performance of machine-learning algorithms versus logistic regression. METHODS: Urine drug screen and attendance records from 40,005 appointments drawn from 2742 patients at a multilocation office-based MOUD program were used to train logistic regression, logistic ridge regression, and XGBoost models to predict a composite indicator of treatment adherence (opioid-negative and norbuprenorphine-positive urine, no evidence of urine adulteration) at next attended appointment. RESULTS: The XGBoost model had similar accuracy and discriminative ability (accuracy, 88%; area under the receiver operating curve, 0.87) to the two logistic regression models (accuracy, 88%; area under the receiver operating curve, 0.87). The XGBoost model had nearly perfect calibration in independent validation data; the logistic and ridge regression models slightly overestimated adherence likelihood. Historical treatment adherence, attendance rate, and fentanyl-positive urine at current appointment were the strongest contributors to treatment adherence at next attended appointment. DISCUSSION: There is a need for risk prediction tools to improve delivery of MOUD. This study presents an automated and portable risk-modeling framework to predict treatment adherence at each patient's next attended appointment. The XGBoost algorithm appears to provide similar classification accuracy to logistic regression models; however, XGBoost may offer improved calibration of risk estimates compared with logistic regression.

Patient Satisfaction With Medications for Opioid Use Disorder Treatment via Telemedicine: Brief Literature Review and Development of a New Assessment.

Telemedicine is increasingly being used to treat patients with opioid use disorder (OUD). It has particular value in rural areas of the United States impacted by the opioid crisis as these areas have a shortage of trained addiction medicine providers. Patient satisfaction significantly impacts positive clinical outcomes in OUD treatment and thus is of great clinical interest. Yet little is known regarding patient satisfaction with the increasingly important platform of telemedicine-delivered medications for opioid use disorder (tMOUD). The goal of this review is to provide a summary of the existing literature regarding patient satisfaction with tMOUD. We also submit a novel survey based on an existing framework designed to assess tMOUD satisfaction, and present pilot data (N = 14) acquired from patients engaged in rural tMOUD care. Telemedicine provides a feasible method for delivering MOUD in rural areas, and our survey provides a useful assessment to measure patient satisfaction with tMOUD. In light of the pressing need for innovative and technology-driven solutions to the opioid epidemic (especially in light of the COVID-19 pandemic), future research should focus on the development and refinement of tools to assess the important implementation goal of patient satisfaction.

Patient-centered methadone treatment: a randomized clinical trial.

BACKGROUND AND AIMS: Methadone patients who discontinue treatment are at high risk of relapse, yet a substantial proportion discontinue treatment within the first year. We investigated whether a patient-centered approach to methadone treatment improved participant outcomes at 12 months following admission, compared with methadone treatment-as-usual. DESIGN: Two-arm open-label randomized trial. SETTING: Two methadone treatment programs (MTPs) in Baltimore, MD, USA. PARTICIPANTS: Three hundred newly admitted MTP patients were enrolled between 13 September 2011 and 26 March 2014. Their mean age was 42.7 years [standard deviation (SD) = 10.1] and 59% were males. INTERVENTION: Newly admitted MTP patients were assigned randomly to either patient-centered methadone treatment (PCM; n = 149), which modified the MTP's rules (e.g. counseling attendance was optional), and counselor roles (e.g. counselors were not responsible for enforcing clinic rules) or treatment-as-usual (TAU; n = 151). MEASUREMENTS: The primary outcome was opioid-positive urine test at 12-month follow-up. Other 12-month outcomes included days of heroin and cocaine use, cocaine-positive urine tests, meeting DSM-IV opioid and cocaine dependence diagnostic criteria, HIV risk behavior and quality of life and retention in treatment. FINDINGS: There was no significant difference between PCM and TAU conditions in opioid-positive urine screens at 12 months [adjusted odds ratio = 0.98; 95% confidence interval (CI) = 0.61, 1.56]. There were also no significant differences in any of the secondary outcome measures (all Ps > 0.05) except Quality of Life Global Score (P = 0.04; 95% CI = 0.01, 0.45). There were no significant differences between conditions in the number of individual or group counseling sessions attended. (Ps > 0.05). CONCLUSIONS: Patient-centered methadone treatment (with optional counseling and the counselor not serving as the treatment program disciplinarian) does not appear to be more effective than methadone treatment-as-usual.

Improvements in outcomes in methadone patients on probation/parole regardless of counseling early in treatment.

OBJECTIVE: This secondary data analysis examined the association between criminal justice (CJ) status and outcomes over 12 months of methadone maintenance treatment. METHODS: In the parent study, 230 newly admitted patients were randomly assigned to methadone either with or without counseling for 4 months followed by standard methadone with counseling. Participants completed the ASI and urine drug testing at baseline and 4- and 12-month follow-up and the Treatment Readiness (TR) scale at baseline. The relationship between baseline CJ status (whether participants were on probation or parole), CJ status by study counseling condition, and CJ status by TR with heroin and cocaine use, illegal activity, days in treatment and treatment retention, arrests, and the number of days incarcerated or hospitalized during follow-up was examined. RESULTS: Compared with participants not on probation/parole, probationers/parolees showed significant reductions in cocaine-positive tests from baseline to 12 months (P < 0.001). Probationers/parolees additionally reported significantly fewer days of illegal activity than nonprobationers/parolees at 12 months (P = 0.02). There was no relationship between CJ status and counseling condition for any outcomes. The relationship between CJ status and TR was significant only for cocaine-positive tests assessed over time (P = 0.017). CONCLUSIONS: Findings suggest that methadone participants on probation/parole showed improvements in outcomes in comparison with participants not on probation/parole, regardless of whether they received counseling during the first 4 months of treatment.

Adding integrative meditation with ear acupressure to outpatient treatment of cocaine addiction: a randomized controlled pilot study.

OBJECTIVES: Cocaine addiction continues to be a major public health problem in the United States. With no U.S. Food and Drug Administration-approved pharmaceutical therapy, treatment often relies on psychosocial interventions. This pilot therapy development study attempts to examine the feasibility and preliminary efficacy of adding breathing-based Integrative Meditation and Ear Acupressure (IMEA) to outpatient treatment of cocaine addiction. DESIGN: Fifty-six (56) cocaine-dependent patients were recruited from an outpatient addiction treatment facility in Baltimore, MD and randomized into either an IMEA or a treatment as usual (TAU) group for the 12 weeks of study, with weekly meetings to monitor treatment outcomes and to facilitate meditative therapy. OUTCOME MEASURES: The outcome measures consisted of treatment retention rates by week 8 and 12; abstinence rates measured by 6 continuous weeks of negative urinalysis for cocaine, and addiction-related symptoms such as anxiety, craving, depression, and withdrawal symptoms. RESULTS: With the assistance of simplified breath training and a portable MP4 device, 80% of IMEA participants self-reported practicing breathing or meditation 5+ days a week with acceptable compliance and showed strong interest in meditative techniques. Compared to TAU, IMEA participants reported significantly higher treatment completion rates by week 8 (89% versus 63%) and week 12 (81% versus 58%), higher abstinence rates (66% versus 34%), and significantly greater reduction in craving, anxiety, and other addiction-related symptoms. Some participants continued meditation after study completion. CONCLUSIONS: It is feasible to add breathing-based IMEA to outpatient treatment of cocaine addiction. Although a number of limitations exist for this pilot study, further large-scale clinical trials and therapy-development studies of IMEA for addiction are warranted.

Randomized trial of standard methadone treatment compared to initiating methadone without counseling: 12-month findings.

AIMS: This study aimed to determine the relative effectiveness of 12 months of interim methadone (IM; supervised methadone with emergency counseling only for the first 4 months of treatment), standard methadone treatment (SM; with routine counseling) and restored methadone treatment (RM: routine counseling with smaller case-loads). DESIGN: A randomized controlled trial was conducted comparing IM, SM and RM treatment. IM lasted for 4 months, after which participants were transferred to SM. SETTING: The study was conducted in two methadone treatment programs in Baltimore, MD, USA. PARTICIPANTS: The study included 230 adult methadone patients newly admitted through waiting-lists. MEASUREMENTS: We administered the Addiction Severity Index and a supplemental questionnaire at baseline, 4 and 12 months post- baseline. Measurements included retention in treatment, self-reported days of heroin and cocaine use, criminal behavior and arrests and urine tests for heroin and cocaine metabolites. FINDINGS: At 12 months, on an intent-to-treat basis, there were no significant differences in retention in treatment among the IM, SM and RM groups (60.6%, 54.8% and 37.0%, respectively). Positive urine tests for the three groups declined significantly from baseline (Ps < 0.001 and 0.003, for heroin and cocaine metabolites, respectively) but there were no significant group x time interactions for these measures. At least one arrest was reported by 30.6% of the sample during the year, but there were no significant between-group effects. CONCLUSIONS: Limited availability of drug counseling services should not be a barrier to providing supervised methadone to adults dependent on heroin--at least for the first 4 months of treatment.

Impact of methadone with versus without drug abuse counseling on HIV risk: 4- and 12-month findings from a clinical trial.

OBJECTIVE: Human immunodeficiency virus (HIV)-risk behaviors were examined at 4- and 12-month follow-up for 230 newly admitted methadone patients randomly assigned to receive either methadone only (n = 99) or methadone with drug abuse counseling (n = 131) in the first 4 months of treatment. METHODS: The AIDS Risk Assessment was administered at baseline (treatment entry) and at 4- and 12-month follow-up. Linear mixed model analysis examined changes in HIV drug- and sex-risk behaviors over the 12 months in the total sample, drug-risk behaviors in the subsample that reported injecting drugs at baseline (n = 110), and sex-risk behaviors in the subsample that reported engaging in unprotected sex at baseline (n = 130). RESULTS: Significant decreases over time were found in the frequencies of injecting, injecting with other injectors, and sharing cooker, cotton, or rinse water in the total sample and the injector subsample (P < 0.05). Decreases were also found in the frequencies of having sex without a condom either with someone who was not a spouse or primary partner or while high (P < 0.05) in the total sample and the frequencies of having sex without a condom and having sex without a condom while high in the unprotected-sex subsample (P < 0.05). No significant treatment group main effects or Treatment Group × Time interaction effects were found in any of the HIV-risk behaviors in the total sample or either subsample (P > 0.05). CONCLUSIONS: During the first 12 months of treatment, providing drug abuse counseling with methadone compared with providing methadone alone was not associated with significant changes in HIV-risk behaviors for methadone maintenance patients.

Benzodiazepine use and misuse among patients in a methadone program.

BACKGROUND: Benzodiazepines (BZD) misuse is a serious public health problem, especially among opiate-dependent patients with anxiety enrolled in methadone program because it puts patients at higher risk of life-threatening multiple drug overdoses. Both elevated anxiety and BZD misuse increase the risk for ex-addicts to relapse. However, there is no recent study to assess how serious the problem is and what factors are associated with BZD misuse. This study estimates the prevalence of BZD misuse in a methadone program, and provides information on the characteristics of BZD users compared to non-users. METHODS: An anonymous survey was carried out at a methadone program in Baltimore, MD, and all patients were invited to participate through group meetings and fliers around the clinic on a voluntary basis. Of the 205 returned questionnaires, 194 were complete and entered into final data analysis. Those who completed the questionnaire were offered a $5 gift card as an appreciation. RESULTS: 47% of the respondents had a history of BZD use, and 39.8% used BZD without a prescription. Half of the BZD users (54%) started using BZD after entering the methadone program, and 61% of previous BZD users reported increased or resumed use after entering methadone program. Compared to the non-users, BZD users were more likely to be White, have prescribed medication for mental problems, have preexistent anxiety problems before opiate use, and had anxiety problems before entering methadone program. They reported more mental health problems in the past month, and had higher scores in anxiety state, depression and perceived stress (p < .05). CONCLUSIONS: Important information on epidemiology of BZD misuse among methadone-maintenance patients suggests that most methadone programs do not address co-occurring anxiety problems, and methadone treatment may trigger onset or worsening of BZD misuse. Further study is needed to explore how to curb misuse and abuse of BZD in the addiction population, and provide effective treatments targeting simultaneously addiction symptoms, anxiety disorders and BZD misuse.

Interim methadone treatment compared to standard methadone treatment: 4-month findings.

Interim methadone (IM; with emergency counseling only) is an effective but highly restricted alternative to methadone treatment program (MTP) waiting lists. However, it is not known whether IM disadvantages patients as compared with standard methadone treatment (SM). In this clinical trial, conducted in two MTPs, 230 newly admitted patients were randomly assigned to IM, SM, and "restored" methadone treatment (SM with a counselor with a reduced caseload). Data were analyzed using generalized estimating equations and generalized linear modeling. There were no significant differences among conditions in days in treatment or of heroin or cocaine use and heroin- or cocaine-positive urine drug tests. The IM as compared to the SM group had significantly fewer self-reported days of criminal activity and lower amounts of money spent on drugs and illegal income. These findings suggest that when SM is unavailable, IM should be more widely used and less restricted.

Transitioning opioid-dependent patients from detoxification to long-term treatment: efficacy of intensive role induction.

Despite findings that opioid detoxification serves little more than a palliative function, few patients who enter detoxification subsequently transition to long-term treatment. The current study evaluated intensive role induction (IRI), a strategy adapted from a single-session intervention previously shown to facilitate engagement of substance-dependent patients in drug-free treatment. IRI was delivered either alone or combined with case management (IRI+CM) to determine the capacity of each condition to enhance transition and engagement in long-term treatment of detoxification patients. Study participants were 240 individuals admitted to a 30-day buprenorphine detoxification delivered at a publicly funded outpatient drug treatment clinic. Following clinic intake, participants were randomly assigned to IRI, IRI+CM, or standard clinic treatment (ST). Outcomes were assessed in terms of adherence and satisfaction with the detoxification program, detoxification completion, and transition and retention in treatment following detoxification. Participants who received IRI and IRI+CM attended more counseling sessions during detoxification than those who received ST (both ps<.001). IRI, but not IRI+CM participants, were more likely to complete detoxification (p=.017), rated their counselors more favorably (p=.01), and were retained in long-term treatment for more days following detoxification (p=.005), than ST participants. The current study demonstrated that an easily administered psychosocial intervention can be effective for enhancing patient involvement in detoxification and for enabling their engagement in long-term treatment following detoxification.

Brief vs. extended buprenorphine detoxification in a community treatment program: engagement and short-term outcomes.

BACKGROUND: Despite evidence supporting the efficacy of buprenorphine relative to established detoxification agents such as clonidine, little research has examined: 1) how best to implement buprenorphine detoxification in outpatient settings; and 2) whether extending the length of buprenorphine detoxification improves treatment engagement and outcomes. OBJECTIVES: The current study examined the impact on 1) successful detoxification completion; 2) transition to longer-term treatment; and 3) treatment engagement of two different length opioid detoxifications using buprenorphine. METHODS: The study compared data obtained from two consecutive studies of early treatment engagement strategies. In one study (n = 364), opioid-addicted participants entered treatment through a Brief (5-day) buprenorphine detoxification. In the other study (n = 146), participants entered treatment through an Extended (i.e., 30-day) buprenorphine detoxification. RESULTS: Results indicated a greater likelihood of successful completion and of transition among participants who received the Extended as compared to the Brief detoxification. Extended detoxification participants attended more counseling sessions and submitted fewer drug-positive urine specimens during the first 30 days of treatment, inclusive of detoxification, than did Brief detoxification participants. CONCLUSIONS: Results demonstrate that longer periods of detoxification improve participant engagement in treatment and early treatment outcomes. SCIENTIFIC SIGNIFICANCE: Current findings demonstrate the feasibility of implementing an extended buprenorphine detoxification within a community-based treatment clinic.

Young heroin users in Baltimore: a qualitative study.

This study describes the characteristics of 67 young heroin users, interviewed using a semistructured qualitative questionnaire (QQ) as part of a larger study of 18- to 25-year-old heroin users seeking detoxification with buprenorphine at a drug treatment center in Baltimore. This new generation of heroin users has a different demographic profile compared to older heroin users in this area. Our data, supported by data from another clinic and from the Maryland State Alcohol and Drug Abuse Administration, seem to indicate that the younger heroin users in treatment settings are predominantly White, with a high proportion of women, often living in the suburbs. Based on responses to the QQ, all subjects initiated heroin use intranasally, usually in a group setting; 75% had subsequently gone on to use intravenously. The typical young heroin user in Baltimore Metropolitan area appears to be a young White man or woman from a middle/working-class background, with exposure to drug use among close contacts while growing up, experimenting with gateway drugs with peers before proceeding first to intranasal, and then intravenous heroin use, engaging in criminal activities to support the habit, repeatedly seeking help with assistance from family, but failing to sustain abstinence due to continued exposure to drug using peers and a poorly implemented plan of aftercare. Further research should focus on efforts to engage peer groups and families in order to improve treatment outcomes in young heroin users.

Preliminary study of neurite outgrowth within polyimide microtubes.

The cone electrode first developed by P.R. Kennedy paved headway in the area of cortical prostheses. While effective, to date, no optimization of the materials, length, diameter or controlling neurotrophic effects have been extensively quantified for such systems. This paper describes an in-vitro model system for the study of neurite outgrowth using PC-12 cells and an array of polyimide microtubes. Our aim is to obtain preliminary design specifications for the eventual optimization of in-vivo neurotrophic electrodes. We performed preliminary characterization of the number and average lengths of PC-12 neurites that penetrated into the tubes mounted within a standard Petri dish. To describe system performance, we observed an increase in the average number of neurites that grew into the tubes over a period of days. We also observed an increase in the average length of the neurites (with a 95% confidence) between day 3 and day 4 of between 14.97 microm and 62.27 microm. In addition, we measured a length change (with a 95% confidence) between day 4 and day 6 of 93.51 microm and 145.45 microm. These results will soon be augmented by quantification of neurites using a photo lithographically patterned glass microgroove system.

A comparison of the patterns of drug use among patients with and without severe mental illness.

We compared the patterns of substance use among patients with severe mental illness (SMI) served by three community treatment teams, substance users with non-severe mental illness (NSMI), and substance abusers without mental illness (no mental illness: NMI). There were highly significant differences in substance use patterns among the groups: among patients with SMI, the order of drug preference was alcohol, cocaine, cannabis, and then heroin; in the NMI group, it was cocaine, alcohol, heroin, and then cannabis; while in the NSMI group, it was alcohol, cocaine, heroin, and then cannabis. The data suggest that the choice of drugs by individuals with SMI may be determined by factors related to their illness, in addition to the price and availability of the particular drug.

Short-term outcomes after brief ambulatory opioid detoxification with buprenorphine in young heroin users.

AIMS: This study examines the outcomes at 1, 3 and 6 months after a very brief outpatient detoxification with buprenorphine in 18-25-year-old heroin users. DESIGN: Prospective follow-up study. SETTING: Outpatient drug treatment clinic, providing brief detoxification in downtown Baltimore, Maryland, USA. PARTICIPANTS: One hundred and twenty-three subjects between 18 and 25 years old; 56% male; 95% Caucasian; seeking detoxification; living in Baltimore City and five surrounding counties. INTERVENTION: Detoxification with buprenorphine over 3 days. Follow-up at 1, 3 and 6 months. MEASUREMENTS: Drug use history, the Addiction Severity Index at baseline and follow-up, urine drug screens, evaluation of the detoxification experience. FINDINGS: By self-report, 37% of the total sample were not currently using heroin at 1 month, 32% at 3 months and 29% at 6 months, and 6.7%, 10.1% and 11.8% had an opioid negative urine test at 1, 3 and 6 months, respectively. There was a significant reduction from the baseline in mean Addiction Severity Index drug use composite score, as well as the mean number of days of heroin and cocaine use during past 30 days, that was sustained over the three follow-up points. Engagement in aftercare was generally poor. CONCLUSIONS: The findings show a reduced frequency and intensity of drug use, suggesting a possible role for brief outpatient detoxification in reducing the severity of dependence for some younger heroin users who may not yet be ready to engage in long-term abstinence-oriented or opioid substitution treatments.

Geriatric patients on a substance abuse consultation service.

The authors describe the demographic and substance use characteristics of older adults (N=310) evaluated by a hospital-based substance abuse consultation service during a 6-year period and compares them to younger adults evaluated by the service. Older adults were more likely to use alcohol and less likely to be injection drug users and heroin, cocaine, or polysubstance users. Elderly patients were also more likely to be admitted for cardiac or gastrointestinal conditions and less likely to be admitted for infections. We estimated that only 1% of older adults admitted to the hospital were referred for substance abuse consultations. Our results suggest that substance use disorders in elderly patients are underdiagnosed and undertreated in the hospital setting.