Damas et al MAT Clinic comparison Manuscript 2019

TWELVE -MONTH TREATMENT RETENTION AND ASSOCIATED FACTORS; A COMPARISON OF TWO METHADONE ASSISTED THERAPY C LINICS IN DAR ES SALAAM TANZANIA Damas Andrea 1, Jessie Mbwambo 1, Sheryl McC urdy 2, Pamela Kaduri 1, Frank Masao 1 1 Department Psychiatry and Mental Health, Muhimbili University of Health and Allied Sciences (MUHAS) , P.O.BOx 65001 Dar es Salaam Tanzania 2 University of Texas Health Science Center at Houston, School of Public Health, 7 000 Fannin St, Houston, TX, USA ABSTRACT Background and Aims Heroin addiction is a global health problem and Opioid use disorders account ed for nearly 12 million (70%) of the total 17 million disability -adjusted life years ( DALYs) attributed to drug use disorders in 2015. Methadone maintenance treatment (MMT) is the most effective pharmacotherapy for heroin addiction. Duration in treatment is critical to achieving better treatment outcomes. The specific aims of this research were to compare two Medically Assisted Treatment (MAT) clinics in Dar e s Salaam Tanzania with respect to client ch aracteristics, client outcome and factors that predict twelve -month treatment retention. Materials and Methods This Retrospective Registry Based Cohort Study utilized the data collected for routine clinical and pr ogram monitoring at the two sites. Cumulative retention in treatment was calculated using Life tables. The analysis of treatment retention predictor variables used both Kaplan -Meier and Cox proportional hazard analyses. Results Three hundred and sixty -two (181 clients in each clinic) clients were studied with respect to the client and program related characteristics. Twelve -month treatment retention was higher (73%) at Mwananyamala MAT clinic compared to Muhimbili MAT clinic (6 4%) with no statistically sign ificant difference (p> 0.05)Having an employment and average short distance (60mg), significantly predicted retention in treatment. Conclusion s In both clinics high methadone dose was associated with longer retention in treatment. We found, though not in both clinics, that employment, short distance travelled from home to clinic, and short duration of heroin use are associated with longer tenure in treatment. These findings have potential implications for clinical practice, research and scaling up Medically Ass isted Treatment services. Key words: Treatment retention, Medically Assisted Treatment, Factors, Tanzania . 1. INTRODUCTION It is estimated that a total of 250 millio n people, or 5% of the adult population between the ages of 15 and 64 years, used an illicit drug in 201 5[1]. The global prevalence of the use of opioids is 0.7 per cent of the world’s adult population , or 35 million users [1] . It is estimated that 533,000 opiate users live in East Africa [2]. Since the late 1990s, injection drug use, primarily of heroin has become widespread in Dar es Salaam and is spreading throughout the country [3]. It is estimated that, there are 50,000 people who in ject drugs (PWID) in Tanzania [4]. PWID experience a range of health, socioeconomic and legal challenges, often with poor outcomes, not least of which is the elevated risk of death compared with the general popu lation ([2]. Opioids including heroin remain major drugs in terms of health cons equences, with nearly 12 million of the total 17 million DALYs attributed to drug use disorders in 2015 being attributed to opioid use disorders [5 ]. Methadone maintenance treatment (MMT) is the most effective pharmacotherapy for heroin addiction, and is n ow available worldwide [6]. Duration in maintenance treatment is critical to achieving better treatment outcomes particularly sustained behavioral change among opioid users [7, 8, 9, 10 ]. Studies in Western countr ies have shown MMT retain 30% to 60% of clients over one year [11 ]. Twelve -month retention rates of 74.4% and 61.6% have been found in Tel -Aviv and Las Vegas clinics respectively [12 ]. Twelve – month treatment retention rates of 48 % in 674 patients from Massachusetts and Connecticut MMT clinics; 52% in 338 MMT patients from New York , 60 % in 1,487 MMT patients from Spain; and 38.4% in 477 from Australia have been reported [12]. Overall, 12 -month retention rate of 65.8 % was found in thirteen Medication -assisted treatment sites in Ukraine [13 ].One year treatment retention rate of 72.8 % was found in Jiangsu province in China [14 ]. A sy stematic review [15] concludes that Medication -assisted treatment programs in low – and middle -income countries achieve an average 50% retention rate after 12 months .A study in Tanzania indicated twelve -month retention rate of 57% [16 ]. Although most successful MMT clinics introduce the same basic treatment principles, they are still highly diverse from one another both within and between countries [12 ]. Studies from around the world reporting about MMT have been published; however, there are few retr ospective studies which have been conducted to compare factors that predict treatment retention and treatment outcome s between clinics. In this study , we compared two Medically Assisted Treatment ( MAT ) clinics, namely Muhimbili and Mwananyamala Clinics, locat ed in two different settings in Dar es Salaam Tanzania. Muhimbili National Hospital MAT clinic is located in a specialist institutional setting while Mwananyamala MAT clinic is located in a primary health care setting. The sp ecific aims of this research were to compare the two clinics with respect to client characteristics, client outcome (twelve -month treatment retention) and factors that predict twelv e-month treatment retention. 2. METHODOLOGY We followed the methods of Peles et al. , 2008 . 2.1 Study area Medically Assisted Treatment for people who use drugs was established in February 2011 at Muhimbili National Hospital in Dar es Salaam, the first of its kind in Tanzania. This site was chosen to serve as a pilot site for MAT. In September 2012; a second MAT clinic in Tanzania was opened at Mwananyamala Regional Hospital. This site was chosen based on the number of people injecting drugs in that locality. The establishment of MAT clinics was supported by the Tanzanian government, the U.S. Centers for Disease Control and Prevention (CDC), Muhimbili Univers ity of Health and Allied Sciences, and Pangaea Global AIDS (Pangaea) .The MAT clinics provide methadone to patients and connect them to other care and treatment as part of a comprehensive package of services for PWIDs. Clients in both clinics may enter the clinic as referral from the local Community Based Organizations (CBO) . All clinics collaborate with the CB Os in recruiting clients. Client participation is voluntary. The admission process includes screening and asses sment by the nurses, social workers and clinicians. Informed consent and information about the MAT Clinic rules and regulations is provided at entry to the program by receptionists. Only violent aggressive behavior or selling drugs within the clinic may re sult in immediate involuntary discontinuatio n of treatment in both clinics. In both clinics, all patients undergo blood tests (i.e., hepatitis C antibody, human immunodeficiency virus (HIV) antibodies, hepatitis B antigens, or the venereal disease research laboratory test) for infectious diseases and chest X -rays for those with clinica l symptoms of TB on admission. Clients do not pay for the services provided. In both clinics, each patient drinks the individual methadone dosage dispensed by pharmacists ever y day in the clinic and no take home dosages of medication are provided. Patients on ART and anti TB also take their medications in the clinic as Directly Observed Therapy (DOT) or take home therapy or both. 2.2 Study population Study participants include d clients enrolled in MA T from 1 st January 2014 to 31 st December 2014 at Muhimbili National Hospital and Mwananya mala Regional Referral Hospital . The study follow up duration was from 1 st January 2014 to 31 st December 2015. Exclusion criteria inclu ded: Patient data that had missing demogra phic/psychosocial intake data; patients transferred to another MAT program; or patients who died during the period . Patients in both clinics were all former heroin -dependent adults (none younger than 18 years).In both clinics clients met criteria for entering methadone treatment (i.e. Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM -IV -TR, 2000) criteria for heroin dependence and multiple daily self -administrations of heroin for 1 year or more. 2.3 Sample size estimates Assuming one clinic had twelve -month treatment retention of 72% and the other clinic 57% twelve -month treatment retention [16 ]; we calculated a sample of 155 clients that would enable us to fin d a 15 % difference in twelve -month treatment retention between the two clinics with 80% power and ά of 5%. Since the g reater the sample size, the smaller the sampling error and the more representative the sample, all 181 clients enrolled in Muhimbili MAT clinic between January 2014 and December 2014 inclusive, were included in the study. Using a systematic random sampling method 181 clients were sampled fr om 399 c lients enrolled in Mwananyamala MAT clinic in the same period. 2.4 Data Source The study utilized the data collected for routine clinical and program monitoring at the two Methadone Assisted Therapy clinics. All data were collected prospectiv ely by a social worker, nurse and clini cian and saved both in patients’ charts, file and electronically on a Microsoft Excel spreadsheet in both clinics. The patient’s file included baseline variables including age, sex, marital status, children, education, du ration of opiate abuse, status of hepatitis C, HIV,TB, and hepatitis B tests, and urinalyses positive or negative for each drug of potential abuse as well as review notes and assessments. From the selected clinical records for 181 clients on methadone f or each clinic, usin g the data collection tool we extracted data from each of the clinical records, clients’ address, date of birth, gender, date commenced on methadone, employment status, education, marital status, date of last exit, and the reason for e xit, use ofmore than one drug and HIV sero -status. The client average distance from home to clinic was estimated using Google Map Distance Calculator. An average methadone dose was constructed and used for further analysis. 2.5 Measures Exposure variables Client characteristics included demographic risk factors age, sex, marital status, education level, duration of heroin use, employment status, HIV sero -status. Client’s age was treated as a continuous variable. Duration of heroin use was treated as a cont inuous variable based on total duration of heroin use measured in years Program related characteristics included methadone dose, accessibility (client average distance from the clinic). All numerical variables namely age , methadone dose, and average client distance to the clinic were transformed into categorical variables. Outcome Measures The outcome variables were treatment retention and attrition. Treatment retention was defined as clients receiving methadone treatment through the end of the study period. Attrition or dropping out of the methadone program was defined as absence of daily methadone dosage for a minimum of 30 consecutive days. 2.6 Statistical Methods All data were normally distributed; therefore parametric descriptive statistics and significance tests were used to des cribe the study population as appropriate . Proportional differences between clinics were analyzed by chi-square tests. The methadone dosage and other continuous variables were analyzed for significant differences usi ng analysis of variance.Client duration in treatment (in days) from first admission until the patient stopped treatment or until the end of the follow -up (12 months) was taken to calculate cumulative retention in treatment using survival analyses life ta bles and Kaplan -Meier Methods . To determine whether there were statistically significant differences in the survival distributions between the two clinics, the log rank test of the two curves was run where p value