Methadone treatments in a Swiss Region, 2001–2008: a registry-based analysis

Opioid substitution treatments with methadone (OSTM) have markedly developed in Switzerland (about 7.66 million inhabitants) increasing from approximately 5000 patients on treatment in 1989 to more than 18 000 in 2000 [1]. In 2008, more than 16 500 persons were on OSTM in Switzerland [2], among an estimated number of opiate-dependent persons in Switzerland of 25 000 [3]. The decrease between 2000 and 2008 is estimated to be due to a decrease in the number of new heroin consumers. This evolution has also occurred in other European countries; for example, in Norway, the number of treatments rose from less than 500 in 1998 to 3000 in 2004 [4].

First conceived as a therapeutic response that was limited in time and ended with a withdrawal from methadone [5], OSTM progressively evolved towards maintenance treatments that are considered necessary over the long-term (methadone maintenance treatments, MMT) [6]. The current representation of drug dependence is that of a chronic disease with repeated phases of consumption, abstinence, and treatment [7]. Periods of abstinence are mostly short [8, 9] and are often followed by relapse [10]. In Switzerland, a therapeutic approach based on patients’ needs is currently recommended [11], and most OSTM are probably MMT.

In the Swiss canton of Vaud (about 730 000 inhabitants), the number of OSTM followed the same trend of increase and stabilisation: since 2000, approximately 1600 patients have benefitted from an OSTM. Treatments require an authorisation from the chief cantonal medical officer, and are administered mainly by general practitioners (GPs) in private practice who are allowed to begin and regularly follow MMT. The two specialized drug treatment centres follow only one-third of all patients. Access to treatment is unlimited and treatments are mainly oriented towards long-term MMT. Treatments have to be in accordance with the guidelines [12] laid down by the Cantonal Health Department regarding treatment induction, doses, mode of delivery, controls, security, rules for “take home” formulations, etc. Documents of entry to treatment or prolongation of treatment have to be filled in by the GP requiring a prescription authorization (see below under methods). Methadone in 1% solution (syrup) is the substance of reference, but buprenorphin or – rarely - oral slow release morphine may also be prescribed, according to the guidelines from the Swiss Society of Addiction Medicine. These last two products are very seldom used. Medical doctors requiring a treatment authorization from the chief cantonal medical officer for the first time have to engage in the continuous educational programme on methadone. Evaluation of abstinence from drug consumption is regularly conducted (e.g. with urine controls). Continued drug use has to be reported to the authority but is not a compulsory reason for removing the patient from treatment.

Many studies have analysed treatment retention in OSTM [4, 13‐20] or factors associated with retention [21, 22]. Waal’s evaluation of substitution treatments in Norway showed a high retention rate in spite of very restrictive treatment rules, with less convincing results regarding treatment outcome (rehabilitation) [4]. According to Magura et al. [22], factors associated with retention are mainly related to treatment characteristics: only two of 16 pre-treatment variables, compared with five of six during-treatment variables, had significant effects on retention. Others have analysed the periodicity of treatment(s) and relapse(s) over time: as reviewed by Magura et al. in 2001 [23] and Amato et al. in 2005 [24], a high rate of relapse after the end of methadone treatment and a rapid return to treatment have been described. In a cohort analysis in New South Wales, Burns et al. reported an estimate of 197 days for treatment duration and a mean of 2.5 treatment episodes per patient [14].

Few studies have analysed the patterns of treatment history - i.e. the occurrence of several consecutive treatment episodes with interruptions in-between - or the duration of treatment interruptions. In Alicante [25], a survival analysis conducted on patients on MMT demonstrated that 16% of patients had more than one treatment episode, with a median interruption duration of 13 months. Patients who re-enrolled had longer total time within MMT than patients retained in treatment and drop-outs. This is also apparent in the study of Nosyk et al. [26].

The objectives of this study are – by using registry-based data:

In the canton of Vaud, methadone treatment can be prescribed and prolongations obtained only with a compulsory official authorization from the Chief cantonal medical officer, and each authorization (“entry”) “prolongation” or “end of treatment” document is retained in the methadone treatment register of the canton of Vaud. The analysis includes completely anonymized data of all patients registered in the methadone treatment register database of the Public Health Service of the canton of Vaud between January 1, 2001 and June 30, 2008. As a result, no connection to any other database such as the death register is possible. The database only contains a patient number for each patient; the corresponding name is retained in the Public Health Service of the canton of Vaud and remained unknown to the researchers. This research received the approval of the Chief medical officer of the canton. It is a secondary and retrospective analysis of anonymous data and obtaining patient consent in this situation is not possible.

Each patient can have three types of documents recorded in the register under his/her patient number:

Each doctor following a patient on methadone must complete these documents; data are sent to the Public Health Service which records them in the registry and then delivers the treatment authorization.

In case of transfer to another doctor, a relatively long period may elapse before the new doctor sends the new “entry” document, even if the patient continues MMT. Because of this administrative delay up to 2 months before the registration of the new treatment document, we have considered in this analysis a period of more than two months to define a treatment interruption.

The following variables were included in the analysis: age, gender, living conditions (fixed abode, without fixed abode, or institution including prison), source of income (full-time or part-time employment, social insurance, social aid), had a previous methadone treatment (yes/no), currently injecting (yes/no), health (HIV test [yes/no-not known] and HIV status [positive/negative-unknown], HCV test [yes/no-not known] and HCV status [positive/negative-unknown], physical condition [good/bad], psychological condition [good/bad] and social situation [good/bad], as estimated by the doctor without standardized rating system), treatment administration mode (only 2 mentions proposed without more precision: methadone delivered each day or 1 to 4 times each week), and methadone dose (maintenance dose). We chose not to include the available variable “drugs use during the treatment (heroin, cocaine/daily, occasionally, no use during last month)” because of the bad quality of the data (many missing data). The “end of treatment” document includes in addition reasons for stopping: dropping out, methadone withdrawal, move to another canton, transfer to another doctor, entry into prison, death, other. It also includes the daily dose of methadone at the time of interruption.

This study confirms that OSTM are actually long-term treatments in the studied region of Switzerland. A majority of patients followed their treatment continuously: among the patients registered between 2001 and 2008, 77.5% had only one treatment episode during the study period, including a group (19.7% of all patients) that had initiated treatment before the beginning of the study period (i.e., before January 1, 2000) without any interruption until June 30, 2008, the date of the end of the study period. Among those who started a treatment episode between January 1, 2000 and the end of the period, 69% remained in treatment after one year and 45% after two years. The risk of interrupting treatment appeared higher during the first 12 months of treatment.

The duration of treatment retention varies widely across different studies. A German study [18] observed that 50% of patients continued their treatment for more than 7 years, with a 65% retention rate at one year. Another study in Spain reported similar results, with a 60% retention rate at one year and 38% at 3 years [25]. In a recent US study, shorter treatment duration (mean 8 months) and lower retention rates were observed [13]. Low retention rates at one year (25–50% in certain studies) were reported in a 2001 review by Magura et al. [23], who identified the restrictive and coercive nature of specific treatment regimens as factors associated with low retention rates. A study of the treatment system in Norway [4] that demonstrated low retention rates resulted in the loosening of conditions of access to treatment and a modification of exclusion rules. A recent Italian longitudinal study conducted on more than 5000 patients [27] concluded that abstinence-oriented treatments were associated with lower retention rates than substitution-oriented treatments. A previous study had also demonstrated poorer outcomes with methadone reduction regimens than with methadone maintenance regimens [20].

The high retention rate measured in the canton of Vaud may be linked with the availability of a wide range of substitution treatment options, from treatment with abstinence objectives (high-threshold treatments) to treatment with risk reduction objectives (low-threshold treatments). This high retention rate is expected to improve patients outcomes in particular in terms of decreasing heroin consumption, decreasing i/v drug use and associated risks [28].

In our study, the probability of interrupting treatment was higher in cases of first treatment episode, in persons without a fixed abode, and in persons younger than 30 years of age. Bell et al. already identified lower retention rates in cases of first treatment episode [15], and lower age has been associated with lower retention rates in several studies [13, 16, 22]. Health-related variables and methadone dose were not associated with treatment duration, although methadone dose has been identified as a factor predictive of retention [16, 29]. It is likely that the methadone dosage reported in the entry document in our study does not reflect the final maintenance dosage. In another study using registry data, conducted in Ontario, Canada, where methadone treatment may also be administered by general practitioners, the probability to have longer treatments episodes (730 days or more) was negatively associated with the number of episodes [30].

The duration of the interruption between two treatment episodes has not been studied, although it is known that many patients experience several treatment episodes [25, 31‐33]. We observed that, after a treatment interruption, a one-year probability of re-entering treatment of 21% and a five-year probability of 43%; the probabilities were lower when the treatment interruption was a “true” end of treatment featuring methadone withdrawal. After five years, the probability of re-entering was very low regardless of the reason for interruption.

Gossop et al. [32] observed that 50% to 60% of persons who had stopped treatment had re-entered into one or another modality of treatment. Ball and Ross [33] estimated that approximately 70% of patients who ended a treatment with a withdrawal from methadone had either re-entered treatment or resumed consumption of opiate drugs of abuse.

Three factors at the interruption of treatment were associated with a higher probability of re-entering: an interruption not due to methadone withdrawal, bad physical health, and higher methadone dose. Bell et al. [34] demonstrated that the significant predictors of re-entry to treatment were younger age and shorter duration of the first treatment episode.

The long treatment durations associated with high rates of re-entry after interruption observed in our study contribute additional thoughts to the debate regarding the real possibilities of methadone withdrawal in OSTM. In light of the known protective effects of methadone maintenance on mortality [35, 36], some professionals question the relevance of methadone withdrawal, particularly whether it would be an inappropriate solution for many patients on MMT [5]. In this context, Clausen et al. [36] insist on the importance of retaining patients in treatment or rapidly reintegrating them in cases of interruption. Strike et al. [30], in their study of predictors of retention, suggest that repeat episodes may not be as beneficial as previous research reports.

In our study, the probability of re-entering was associated with bad conditions at the time of interruption and most cases who re-entered did so within one year, suggesting that drop-outs have the capacity to reintegrate rapidly and therefore to reduce their risks (in particular linked with injection). However, a non-negligible proportion of patients who interrupted treatment were still out of treatment after five years, and even if we do not know anything about their situation (particularly regarding mortality after interruption for another cause), it is possible that successful methadone withdrawal occurred in a part of this group. However, it was not possible to link this database with other databases, especially on mortality. More research is still needed to assess long-term outcomes after treatment interruption (regardless of cause) in long-term MMT.

Our study has limitations. The first limitation is that the analysis covers only one region of Switzerland. However, this region represents 10% of the population, and treatment recommendations issued by the Federal Office of Public Health are applied in the canton of Vaud.

The probability of staying in treatment was calculated on the basis of the first “entry” document registered during the considered period. Many of these treatments are not first treatments (the information is available) and this overestimates the general probability of staying in treatment since the analysis shows a lower retention for first treatments. We nevertheless chose this option to increase the number of patients in the model.

Because long delays have been observed in the administrative transmission of documents when the patient was transferred to a different doctor [37], we considered that when the duration of the interruption was less than 2 months, there was no treatment interruption. We may have overestimated treatment retention and underestimated the number of drop-outs by doing so. Furthermore, the register only includes patients living in the canton of Vaud. When a person begins treatment, a previous treatment in another canton may not be mentioned. This would lead to a possible underestimation of the duration of treatment.

The poor quality of data on drug use during the treatment was another limitation to the analysis of the database, as was the fact that the assessment of the psychological and physical state of the patient as well as its social inclusion was left to physician appreciation without standardized measurement. The quality of the information on socio-demographic variables and other health information (HIV/ hepatitis) was good. Overall, we feel that these registry-based data are robust and provide useful information to health providers.

OSTM are long-term (maintenance) treatments in Switzerland. Younger age, bad living conditions at entry, and first treatment are predictors of lower retention. Approximately one-half of patients who interrupt treatment will re-enter treatment within 5 years.