Available online at www.sciencedirect.com
Drug and Alcohol Dependence 94 (2008) 151–157
Mortality prior to, during and after opioid maintenance treatment (OMT):
A national prospective cross-registry study
Thomas Clausen∗, Katinka Anchersen, Helge Waal
University of Oslo, Medical Faculty, Institute of Psychiatry, SERAF – National Centre for Addiction Research, Oslo, Norway
Received 13 August 2007; received in revised form 6 November 2007; accepted 6 November 2007
Available online 21 December 2007
reduction is still uncertain. This study investigates mortality reductions in an “intention-to-treat” perspective including all dropouts. The mortality
reducing effects of OMT are examined both within treatment and post-treatment. The study separates overdose and total mortality reductions.
Methods: The study is a prospective cross-registry study with up to 7 years follow-up. All opiate dependents in Norway who applied for OMT
(a total of 3789 subjects) were cross-linked with data from the death registry from Statistics Norway. Date and cause of death were crossed
with dates for initiation and termination of OMT, and subjects’ age and gender. A baseline was established from the waiting list mortality rate.
Intention-to-treat was investigated by analysing mortality among the entire population that started OMT.
Results: Mortality in treatment was reduced to RR 0.5 (relative risk) compared with pre-treatment. In the “intention-to-treat” perspective, the
mortality risk was reduced to RR 0.6 compared with pre-treatment. The patients who left the treatment programme showed a high-mortality rate,
Conclusions: OMT significantly reduces risk of mortality also when examined in an intention-to-treat perspective. Studies that evaluate effects of
studies as the current one are an important supplement to other observational designs in this field.
© 2007 Elsevier Ireland Ltd. All rights reserved.
Keywords: Maintenance treatment; Mortality; Opioid; Methadone; Buprenorphine; Intention-to-treat
Mortality among untreated opioid dependents is internation-
ally estimated with relatively wide variations; between 1 and
4 per 100 person years (Amato et al., 2005; Bargagli et al.,
2006; Concool et al., 1979; Darke et al., 2007; Hser et al., 1993;
Oppenheimer et al., 1994; Rossow and Lauritzen, 1999). As
the majority of deaths among opiate dependents are reported
to be overdose-related (Darke and Hall, 2003; Digiusto et al.,
2004; Gossop et al., 2002), the level of overdose mortality is
Further, the level of mortality in a given population will be
∗Corresponding author at: SERAF, LAR ∅st, Kirkeveien 166, N-0407 Oslo,
Norway. Tel.: +47 23368942; fax: +47 23368901.
E-mail address: email@example.com (T. Clausen).
eral health status and treatment availability (Darke et al., 2007).
et al., 2007). As risk reduction is influenced by baseline mortal-
ity, variation in mortality levels across different populations of
illicit drug users is thus the result.
Opioid maintenance treatment (OMT) is generally consid-
ered to be the most important harm-reducing measure (Darke et
al., 2007; van den Brink and Haasen, 2006; Zador, 2007), and
it is often stated that “OMT saves lives” (Brugal et al., 2005;
Gunne and Gronbladh, 1981). While the evidence favours this
view, there are several weaknesses in the research that underlie
these results, particularly in regard to level of risk reduction.
Still, estimating the level of risk reduction might be important
both in treatment policy, in balancing the need for control with
0376-8716/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved.
T. Clausen et al. / Drug and Alcohol Dependence 94 (2008) 151–157
However tragic, overdose deaths occur statistically infre-
quently, and research will need both adequate population size
and follow-up period to attain sufficient statistical strength.
Mortality has been included in the outcome measures of some
randomized clinical trials (RCTs) with short follow-up periods,
with little or no significant mortality reducing effects found
(Dole et al., 1969; Strain et al., 1993). Even within opioid-
dependent populations, research designs with large samples and
long-term follow-up are needed, but difficult to apply in RCTs,
in order to firmly examine mortality reducing effects. (Faggiano
et al., 2003). A systematic review concludes that it is not proven
that OMT significantly reduces mortality (Amato et al., 2005).
Nonetheless, the results from OMT research have been diver-
gent on the issue. In a Swedish study, a clinical trial comparing
buprenorphine maintenance to medically supervised buprenor-
phine withdrawal, observed a statistically significant difference
tality after detoxification with that in treatment, and the findings
ago, Gunne compared a population-denied entrance to OMT
found marked differences in mortality (Gunne and Gronbladh,
1981). However, that study was performed in a situation of pol-
icy conflict in Sweden and the group-denied treatment were in
a rather vulnerable situation.
The mortality-reducing effects of OMT are primarily estab-
lished through observational studies. However, these often
exclude dropouts, sometimes lack clear selection criteria for
treatment or might be seen as local area studies difficult to gen-
eralise from (Hser et al., 1993; Kleinman et al., 1977; Nich and
Carroll, 2002). Findings are, as a result, mostly applicable to
those selected by the treatment unit and maintained in treat-
ment. Even well designed cohort studies face challenges with
persons lost to follow-up (Soyka et al., 2006; Termorshuizen et
sion and exclusion criteria may vary. Comparison of effect of
different OMT programmes in different countries or regions is
therefore not always appropriate. For example, a recent study
from Stockholm found no opiate overdose deaths in their OMT
polydrug users and excluded patients with concurrent use of
drugs (Fuglestad et al., 2007). The in-treatment effects of such a
ing all systematic maintenance treatment offered (Waal, 2007).
Applicants wait for treatment for some time, enabling a wait-
ing list design, and establishing a pre-treatment mortality level
within the study population.
Patient registers include information on periods both in and
out of treatment.
According to current estimates there are between 8200 and
12500 injecting opiate addicts in Norway (Bretteville-Jensen
and Amundsen, 2006). Approximately 4600 persons are in
OMT, of whom 90–95% are intravenous drug users. (Statistics
from National OMT competence centre 2006). Buprenorphine
was registered as a therapeutic OMT drug in 2001 and by 2003;
23% used buprenorphine, the rest used methadone. The average
centre 2006) (Waal, 2007).
Statistics Norway (SSB) receives all death certificates con-
cerning deceased Norwegians (SSB, 2006). The relatively high
quality and accessibility of electronic public registries in Nor-
way permit a cross-registry study between an OMT patient
registry and the death registry.
OMT population, with calculated baseline mortality prior to
treatment, including all programme dropouts, is studied. The
study is prospective and with a population size sufficient for
appropriate statistical analysis. As nation wide registers are
available, it is possible to study the effects of treatment in an
“intention-to-treat” design. Our results complement previous
(1) To assess differences in mortality rates prior to, during and
(2) To evaluate mortality reductions in an “intention-to-treat”
(3) To examine the distribution of drug overdose versus non-
overdose as cause of death.
2. Materials and methods
2.1. Sample and data collection
The Norwegian OMT programme is designed to reach the
other types of treatment (Waal, 2007). The inclusion criteria for
priority. Treatment is based on cooperation between social ser-
All opioid-dependent people who applied for and were
accepted for OMT in Norway between 1 January 1997 and 31
December 2003 – a total of 3789 persons – were included in the
data for this study, with a total observation time of up to 7 years.
The design hence is a dynamic cohort, where persons were
included as they applied for OMT, resulting in individual and
varying observation times; from inclusion until 31 December
2003, which was the time set for examination of mortality.
The sample has been divided into pre-treatment, (applicants
of OMT). A national OMT registry including national ID num-
bers was established based on the electronic record system in
T. Clausen et al. / Drug and Alcohol Dependence 94 (2008) 151–157
each OMT centre. Each centre provided lists of all persons who
had applied for, entered and left OMT during the observation
period. These lists were sent to Statistics Norway and infor-
mation on the date and cause(s) of any deaths were attached
to the data files. The merging of data registers was performed
towards the end of 2005. Thus, all deaths in the observation
period (through 2003) are included in the register.
2.3. Measures and definitions
Death certificates registered with Statistics Norway are in
most cases completed by a physician after examination of the
deceased. In about one-third of cases, additional information as
a result of autopsy is included (SSB, 2006). Death certificates
include one principal cause of death, and up to four underlying
causes (ICD 10 codes) (SSB, 2006; WHO, 2006). For the pur-
poses of this paper, only the principal cause of death is used.
Acute intoxications/deaths from all substances were combined
F19.0, X42.0 and X44.0 diagnoses.
The non-overdose groups included both somatic and sud-
den/violent deaths (such as suicide, traffic accidents and
Some subjects included in the study (167 individuals) under-
went several treatment periods. “In-treatment” refers in this
study to the actual number of days in treatment (sum of days in
treatment, excluding days “post-treatment” between treatment
periods). “Post-treatment” is the number of days out of treat-
ment both between and following treatment periods within the
study period. If subjects had several application dates, the first
date was chosen.
The registry initially contained some individuals that for
varying reasons did not start treatment. Some were not fulfill-
ing the criteria of opioid dependence. Others chose long-term
drug-free residential treatment and some chose not to start for
ment initiation within 1 year, and were all re-examined and
verification of the application status confirmed. Subjects who
were ineligible or withdrew their application for OMT consti-
tute a mixed group that is termed “ineligible for treatment” (403
persons in total). Possible cases of misclassifications between
the utmost care has been taken to reduce the problem by manu-
ally cross-checking the data with each centre. Mortality rate is
not calculated for this group as a whole, as no definite obser-
vation time was available. Some persons with application status
(pre-treatment) between 1 and 365 days, (i.e. included during
the final year of observation (2003)), may have withdrawn the
application prior to commencing treatment or been found non-
eligible, without this being captured during data collection, as
beyond 31 December 2003.
“Intention-to-treat” in this paper includes every person who
ever started on OMT.
Most analyses and descriptive statistics were performed by
SPSS version 14.0.2. (Inc., Chicago, IL, USA). Mortality rates
were calculated per 100 person years, being equivalent in this
A Cox regression with a time-dependent covariate was per-
formed (by SAS 9.2) (SAS Institute Inc.), to assess statistical
differences between the treatment categories, as each individual
could have changed status from pre-treatment to treatment and
subsequently to post-treatment during the observation period.
ject’s placement within the groups: pre-treatment, in-treatment
and post-treatment. Calculated hazard ratios should be inter-
preted as relative risk (RR) between groups.
The project was approved by the National Committees for
Research Ethics and by the Data Inspectorate of Norway.
3.1. Sample characteristics
at the end of 2003.
Female subjects constituted 31.9%. Ages ranged from 23 to
66 years of age. Twenty-three percent of the subjects, who had
commenced OMT, terminated the treatment without restarting
within the observation period. The mean age of male subjects
was slightly higher than for the female subjects (live males 2.1
Characteristics of the study population, as of 31 December 2003
Pre-treatment In treatment Post- treatmentSub-total aliveDeceasedTotal
483 23827113576 2133789
Mean age (S.D.)
Percent and mean age. Mean age of subjects and (S.D.) (as on 31 December 2003) included in the study.
T. Clausen et al. / Drug and Alcohol Dependence 94 (2008) 151–157
Cause of mortality and observation time
Cause of mortality
Total mortality (%)
Median no. of days in group (S.D.)
Total observation time (p.y.)
Real numbers (percent). Chi-square P=0.001. p.y.: person years.
years (mean) older than females). This was true for all groups
except for deceased subjects, which had similar mean ages for
3.2. Overdose versus non-overdose
Overdose deaths comprised the majority of all deaths in all
groups combined and 79% of deaths in the pre-treatment group
(Table 2). For those leaving treatment, overdose mortality was
not statistically different from that of the pre-treatment group.
The frequency of overdose deaths in treatment was clearly
24 cases. The in-treatment group is dominated by non-overdose
ber of days in the different treatment groups showed variation,
but median pre-treatment time (on waiting list) was between 5
and 6 months, and in-treatment time close to 2 years. The total
observation time (from the first application date to 31 Decem-
per person. Of the total observation time, the in-treatment time
made up the main bulk.
3.3. Gender and mortality
Table 3 shows gender-specific mortality rates differentiated
tion in overdose mortality for the group in OMT (1.4) compared
with the pre-treatment group (2.4) was found. Overdose mor-
tality was particularly prevalent among males who had ceased
treatment (4.1). From an “intention-to-treat” perspective the
mortality reduction was smaller than the in-treatment results,
with a significant reduction from 2.4 pre-treatment, to 1.8 in
Mortality rates per 100 person years, by gender (CI)
Intention-to-treat group, including the “in-treatment” and “post-treatment”
The “post-treatment” subgroup had a higher mortality rate
than any other group; this particularly applied to male subjects.
Female subjects using heroin or undergoing OMT had a similar
mortality rate to their male counterparts, but after OMT treat-
ment the mortality of female subjects was lower than that of
3.4. Mortality risk reduction with OMT
A significant overall reduction in mortality risk between
the pre-treatment and the “intention-to-treat” populations
with a hazard ratio of 0.6, P=0.004 was found (Table 4).
Similarly, there was a significant overall reduction in mor-
Hazard ratios for death according to treatment group, by gender
Calculations based on Cox regression with a time-dependent covariate.
Hazard ratios to be interpreted as relative risks (RR).
T. Clausen et al. / Drug and Alcohol Dependence 94 (2008) 151–157
tality risk between pre-treatment and in-treatment groups
(hazard ratio 0.5, P=0.001). The post-treatment group as
a whole was not significantly different from the pre-
treatment group in regard to mortality; however, a tendency
toward gender differences was observed. Following treat-
ment, males had significantly higher mortality than those in
the pre-treatment levels. The post-treatment females had a
non-significant tendency toward reduced mortality. Risk for
overdose death was significantly reduced with OMT, both for
the in-treatment group separately and in “intention-to-treat”
4.1. Discussion of main results
The core trait of this study is the approach that allows an
examination of mortality reduction from a defined baseline
level both within OMT and post-OMT with overall mortality
separated from overdose mortality. The study confirms ear-
lier findings of reduced mortality in treatment and points to
the change in pattern with overdose mortality dominant before
and after and non-overdose mortality in treatment. Further,
the approach enables calculation of concrete risk reductions
in relation to the pre-treatment level, demonstrating significant
reductions both in overdose deaths (RR 0.2) and all-cause mor-
tality (RR 0.5).
This presented level of mortality reduction might be influ-
enced by particular characteristics of OMT in Norway. For
example, a long-term opiate addiction history is a precondition
to be eligible for OMT in Norway; the findings should never-
theless indicate a level of reductions that might be generally
In addition to these findings, the study also demonstrates
that a significant mortality risk reduction is found using the
“intention-to-treat” perspective; meaning mortality in dropouts
included. In this perspective the overall mortality reduction,
as a result of our national OMT programme, is RR 0.6 com-
pared with the pre-treatment level. The benefits of OMT are
supported from the “intention-to-treat” perspective; however,
the level is reduced compared with the results from the “in-
will therefore tend to exaggerate the positive outcomes of OMT
due to selection mechanisms.
Overdose mortality was still found in treatment although
reduced from the pre-treatment level. One explanation may be
that overdoses were caused by polydrug intoxications, of which
some of the death certificates were supplemented with informa-
tion supported on autopsy and toxicological analyses, hence the
risk of misclassification. Deaths among patients with a known
The mortality pattern demonstrated in the study is on level
holm (Fuglestad et al., 2007; Langendam et al., 2001; van
ment (0.4 per 100 person years) is similar to what has been
reported by Caplehorn et al. (1996) with a mortality of 0.5 per
100 person years during treatment.
The patterns of mortality and mortality reductions are gen-
erally similar for both genders, with a couple of exceptions.
Male subjects who cease OMT treatment have twice the risk
of death as female subjects in the same situation. The increase
in overdose mortality post-treatment was more or less confined
to males. However, males in the general population also have
about twice the age-adjusted mortality risk as females (SSB,
2006). The increase in mortality for the “post-treatment” group
is by and large caused by an increase in the frequency of over-
2005; Desmond and Maddux, 2000; Digiusto et al., 2004). Our
observation is that this differs for the two genders. We have no
The nature of the study population, comprising the entire
OMT population in Norway is important to notice. It is likely
that the group “post-treatment” includes individuals with the
largest burden of psychiatric co-morbidity, and that they have a
polydrug use. It is generally accepted that it is the most heavily
burdened and non-compliant patients who leave OMT (Dole et
al., 1969; Fischer et al., 2005; Fuglestad et al., 2007; Johanson,
1981; Skodol et al., 2002).
Another noteworthy finding is that non-overdose deaths in
the in-treatment group are dominating. One explanation for this
is that to be eligible for OMT, somatic health problems promote
swift acceptance into the programme.
No exact figures for HIV prevalence in the study population
are available in the dataset, but it is known to be relatively low
in Norway (SIRUS, 2006). Estimated HIV prevalence, based on
4.2. Methodological considerations
The “intention-to-treat ”perspective in this area of research
may be regarded controversial, as OMT is usually considered
to be a life-long treatment. This does not, however, justify the
necessity to exclude the mortality in treatment dropouts in the
reduction in treatment is therefore only one of several goals in
an OMT programme, and has to be assessed against competing
interests. This underlines the importance of knowing the poten-
tial in reducing mortality risk for an entire population of opiate
addicts eligible for OMT, not only those currently in treatment.
The dataset includes no information about individual type
of maintenance drug or dosage. During the observation period
for this study, methadone was the most widely used medication,
whereas buprenorphine introduced in 2001 was used by 23%
by the end of 2003. Additionally, no information on the sub-
T. Clausen et al. / Drug and Alcohol Dependence 94 (2008) 151–157
jects’ drug use history was available, nor whether termination
of treatment was voluntary or not.
age limit of 25 years. Additionally, age is measured at the end
of the observation time, not at inclusion.
the registries employed. In this study, the national OMT patient
registry was based on electronic patient records information
from the regional OMT centres. For patients in-treatment and
for the post-treatment group, we believe that the information is
up to date and of good quality, as budgets and monitoring of
expenses are based on continuously updating the data.
In spite of thorough control of each case that did not enter
treatment (ineligible), some subjects might wrongly have been
classified in the “pre-treatment” group. As the “ineligible for
treatment” group had a low crude mortality rate, misclassifica-
the issue has been thoroughly examined and manually checked,
we find it unlikely that results are significantly affected.
The current study population represents those who were
found eligible to OMT in Norway during the study period. At
present this comprises between 1/3 and 1/2 of all injecting opi-
ate dependents in Norway, being characterised by a long term
and severe opiate-dependence syndrome. The Norwegian OMT
programme has distinct characteristics related to high-threshold
level and rehabilitation goals (Waal, 2007).
However, as OMT programmes vary internationally due to
different inclusion procedures, the absolute levels of mortality
relative level of risk reduction of mortality as a result of OMT
is more comparable across OMT populations, than the absolute
mortality levels are.
The quality of the data from the death registry in Norway is
viewed as good in terms of overall mortality. Still, the mortality
cause is dependent on the individual practice of each doctor
of overdose as a cause of death.
Monitoring of mortality and other aspects of treatment out-
come via cross-registry studies, as done here, is a rational
and efficient way of evaluating OMT. However, there are to
date only a few countries, mostly Scandinavian, that have the
infrastructure in terms of national registries feasible for our
study approach. Registry data are typically limited in terms
of covariate variables available. Therefore, other observational
designs providing more extensive and varied information are
needed in addition. This can yield important data on dif-
ferences and characteristics of subgroups within the OMT
population and the programme dropouts, e.g. on co-morbidity
In conclusion, the present study adds systematic evi-
dence to the notion that OMT reduces mortality. The results
support earlier studies and add the finding of reduced mor-
tality also from an established baseline level and in an
“intention-to-treat” perspective, with a mortality risk reduc-
tion in the order of RR 0.6. The cross-registry strategy is
favourable in terms of allowing analyses without concerns for
The findings argue strongly for OMT programme efforts to
reduce time on waiting lists, as the mortality reducing effect
in treatment is so evident. Future studies should be designed
to further investigate the underlying explanatory factors for the
gender differences recognized.
Conflict of interest
Thanks to all Norwegian OMT centres, providing patient
Thanks to Hege Kornør (Norwegian Knowledge Centre for
the Health Services) for participation in design and discussions
leading to the paper.
Funding: The project is carried by employees at the Insti-
tute of Psychiatry, University of Oslo, with no separate external
Contributors: Thomas Clausen has taken part in the final
rounds of data collection, performed the statistical analyses and
drafted the first version of the paper.
Katinka Anchersen took part during the data collection and
discussions leading to the paper and has taken part and edited
during analyses and writing up.
leading to the paper and writing up.
All authors have read and approved this final version of the
An overview of systematic reviews of the effectiveness of opiate mainte-
nance therapies: available evidence to inform clinical practice and research.
J. Subst. Abuse Treat. 28, 321–329.
Bargagli, A.M., Hickman, M., Davoli, M., Perucci, C.A., Schifano, P., Buster,
impact on adult mortality in eight European countries. Eur. J. Public Health
Bretteville-Jensen, A.L., Amundsen, E.J., 2006. Omfang av sprøytemisbruk i
Norge. Report 5. SIRUS, Oslo, p. 83.
Brugal, M.T., Domingo-Salvany, A., Puig, R., Barrio, G., Garcia de Olalla, P.,
de la Fuente, L., 2005. Evaluating the impact of methadone maintenance
programmes on mortality due to overdose and aids in a cohort of heroin
users in Spain. Addiction 100, 981–989.
Caplehorn, J.R., Dalton, M.S., Haldar, F., Petrenas, A.M., Nisbet, J.G., 1996.
Methadone maintenance and addicts’ risk of fatal heroin overdose. Subst.
Use Misuse 31, 177–196.
Concool, B., Smith, H., Stimmel, B., 1979. Mortality rates of persons entering
methadone maintenance: a seven-year study. Am. J. Drug Alcohol Abuse 6,
T. Clausen et al. / Drug and Alcohol Dependence 94 (2008) 151–157 Download full-text
Darke, S., Degenhardt, L., Mattick, R., 2007. Mortality Amongst Illicit Drug
Users. Camebridge University Press, Camebridge.
Darke, S., Hall, W., 2003. Heroin overdose: research and evidence-based inter-
vention. J. Urban Health 80, 189–200.
Desmond, D.P., Maddux, J.F., 2000. Deaths among heroin users in and out of
methadone treatment. J. Maint. Addict. 1, 45–61.
cotherapies for opioid dependence (NEPOD). Addiction 99, 450–460 (see
comment; erratum appears in Addiction. 2005 Jan;100(1):139).
Dole, V.P., Robinson, J.W., Orraca, J., Towns, E., Searcy, P., Caine, E., 1969.
Methadone treatment of randomly selected criminal addicts. N E J. Med.
Faggiano, F., Vigna-Taglianti, F., Versino, E., Lemma, P., 2003. Methadone
Syst. Rev., CD002208.
Fischer, B., Rehm, J., Kim, G., Kirst, M., 2005. Eyes wide shut?—A conceptual
and empirical critique of methadone maintenance treatment. Eur. Addict.
Res. 11, 1–9 (discussion 10–14).
Fuglestad, A., Stenbacka, M., Leifman, A., Nylander, M., Thiblin, I., 2007.
Methadone maintenance treatment: the balance between life-saving treat-
ment and fatal poisonings. Addiction 102, 406–412.
Gossop, M., Stewart, D., Treacy, S., Marsden, J., 2002. A prospective study
of mortality among drug misusers during a 4-year period after seeking
treatment. Addiction 97, 39–47.
Gunne, L.M., Gronbladh, L., 1981. The Swedish methadone maintenance pro-
gram: a controlled study. Drug Alcohol Depend. 7, 249–256.
Hser, Y.I., Anglin, D., Powers, K., 1993. A 24-year follow-up of California
narcotics addicts. Arch. Gen. Psychiatry 50, 577–584.
Kakko, J., Svanborg, K.D., Kreek, M.J., Heilig, M., 2003. 1-Year retention and
heroin dependence in Sweden: a randomised, placebo-controlled trial (see
comment). Lancet 361, 662–668.
Kleinman, P.H., Lukoff, I.F., Kail, B.L., 1977. Magic fix – critical analysis of
methadone – maintenance treatment. Soc. Probl. 25, 208–214.
Langendam, M.W., van Brussel, G.H., Coutinho, R.A., van Ameijden, E.J.,
2001. The impact of harm-reduction-based methadone treatment on mor-
tality among heroin users. Am. J. Public Health 91, 774–780.
a cohort of heroin addicts from London clinics: a 22-year follow-up study.
Addiction 89, 1299–1308.
and life-threatening overdoses among drug addicts. Addiction 94,209–219.
SIRUS, 2006. The drug situation in Norway 2006. In: SIRUS (Ed.), Annual
report to the EMCDDA, Oslo, p. 87.
Skodol, A.E., Gunderson, J.G., Pfohl, B., Widiger, T.A., Livesley, W.J., Siever,
L.J., 2002. The borderline diagnosis. I. Psychopathology, comorbidity, and
personality structure [see comment]. Biol. Psychiatry 51,936–950.
Soyka, M., Apelt, S.M., Lieb, M., Wittchen, H.U., 2006. One-year mortal-
ity rates of patients receiving methadone and buprenorphine maintenance
therapy: a nationally representative cohort study in 2694 patients. J. Clin.
Psychopharmacol. 26, 657–660.
SSB, 2006. Statistics Norway http://www.ssb.no/dodsarsak/. Access date:
Strain, E.C., Stitzer, M.L., Liebson, I.A., Bigelow, G.E., 1993. Dose–response
effects of methadone in the treatment of opioid dependence. Ann. Int. Med.
Termorshuizen, F., Krol, A., Prins, M., van Ameijden, E.J., 2005. Long-term
outcome of chronic drug use: the Amsterdam Cohort Study among drug
users. Am. J. Epidemiol. 161, 271–279.
van Ameijden, E.J., Langendam, M.W., Coutinho, R.A., 1999. Dose-effect rela-
tionship between overdose mortality and prescribed methadone dosage in
low-threshold maintenance programs. Addict. Behav. 24, 559–563.
van den Brink, W., Haasen, C., 2006. Evidenced-based treatment of opioid-
dependent patients. Can. J. Psychiatry 51, 635–646.
WHO, 2006. International Statistical Classification of Diseases and Related
Health Problems. 10th Revision.
Waal, H., 2007. Merits and problems in high-threshold methadone maintenance
treatment. Eur. Addict. Res. 13, 66–73.
Zador, D., 2007. Methadone maintenance: making it better. Addiction 102,