UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)
UvA-DARE (Digital Academic Repository)
Prevalence morbidity and mortality among heroin users and methadone patients
Buster, M.C.A.
Publication date
2003
Link to publication
Citation for published version (APA):
Buster, M. C. A. (2003). Prevalence morbidity and mortality among heroin users and
methadone patients.
General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).
Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You will be contacted as soon as possible.
© ©
M a r c e ll C . A . B u s t e r
DepartmentDepartment of Epidemiology, Documentation and Health Promotion EDC, MunicipalMunicipal Health Service, Amsterdam
G i e ll H . A . v a n B r u s s e l
DepartmentDepartment of Social and Mental Health, Municipal Health Service, Amsterdam.
W i mm v a n d e n B r i n k
P M A M A H U A M P P
NN
I
N
C
R
E
A
S
E I N
O V E R D O S EE M O R T A L I T Y
D U R I N GG T H E F I R S T T W O
W E E K SS A F T E R E N T E R I N G
O RR
R
E
-
E
N
T
E
R
I
N
G
M E T H A D O N EE T R E A T M E N T
I NN
A
M
S
T
E
R
D
A
M
AddictionAddiction 2002; 97: 993-1001 A b s t r a c t tAims:Aims: It has been s u g g e s t e d t h a t s t a r t i n g and t e m p o r a r i l y d i s c o n t i n u
-ingg m e t h a d o n e t r e a t m e n t is related to an increased risk in overdose m o r t a l i t y .. This s t u d y describes t h e incidence o f o v e r d o s e m o r t a l i t y in r e l a t i o nn t o t i m e after ( r e ) e n t e r i n g or leaving t r e a t m e n t .
Design:Design: A d y n a m i c c o h o r t of 5 2 0 0 A m s t e r d a m m e t h a d o n e clients
wass o b s e r v e d d u r i n g t r e a t m e n t and a maximum of 1 year- after
treat-m e n t . .
Findings:Findings: Between 1 9 8 6 a n d 1 9 9 8 , 29 729 person-years - py- and
688 overdose deaths were r e c o r d e d , leading t o an o v e r d o s e m o r t a l i t y ratee of 2 . 3 / 1 0 0 0 py 2.2 during and 2.4 after treatment- A m o d e s t increasee was o b s e r v e d d u r i n g the f i r s t 2 weeks a f t e r ( r e ) e n t e r i n g treat-m e n t ;; 6 . 0 / 1 0 0 0 py Rate Ratio: 2.9; 95% Confidence Interval: 1.4; 5.8- D i r e c t l y afterr l e a v i n g t r e a t m e n t no increase was o b s e r v e d .
Conclusions:Conclusions: I n h a l i n g h e r o i n , c o m m o n a m o n g A m s t e r d a m h e r o i n
users,, is t h o u g h t t o a c c o u n t f o r l o w OD m o r t a l i t y rates b o t h d u r i n g andd after t r e a t m e n t . A c c u m u l a t i o n of m e t h a d o n e , i n a d e q u a t e assess-mentt of t o l e r a n c e of k n o w n clients re-entering t r e a t m e n t and concur-rentt periods of stress or e x t r e m e h e r o i n use w h e n e n t e r i n g t r e a t m e n t aree m e n t i o n e d as p o s s i b l e e x p l a n a t i o n s of the increased risk w i t h i n thee f i r s t 2 w e e k s . An A u s t r a l i a n s t u d y r e p o r t e d a m u c h h i g h e r increase.. The m o d e s t increase in A m s t e r d a m is e x p l a i n e d by low back-g r o u n dd risk o f o v e r d o s e m o r t a l i t y , low s t a r t i n back-g dosaback-ge and the low-t h r e s h o l dd low-t o low-t r e a low-t m e n low-t .
P M A M A H U A M P P
1 44 T r a n s i t i o n p e r i o d s s o d o v a r d o s * m o r t a l i t y
I n t r o d u c t i o n n
Heroinn users have a high risk of death, especially death due to over-dosee OD. It has been shown repeatedly that methadone treatment reducess the risk of a fatal heroin overdose.1"2 | J H Daily methadone dosagess enhance the opiate users' tolerance to opiates and may decreasee the amount of heroin used.
However,, methadone may not only prevent overdose death but alsoo be the cause of it.15 This can be accidental due to misjudge-mentt of a client's tolerance by the physician or misjudgement by the opiatee user buying methadone at the black market or combining methadonee with heroin. Methadone can also be used as a means to committ suicide. Two recent studies indicate that the number of deaths preventedd exceeds the number of deaths caused by methadone treat-ment.1617 7
Dailyy regular methadone use is presumed to be safe. The risk off overdose may be expected to be higher during periods of transi-t i o n ,, i.e. periods aftransi-ter (re)entransi-tering or discontransi-tinuatransi-tion of metransi-thadone treatment.. Some studies indicate increased overdose mortality during thee first weeks of treatment,,6,* Until now, overdose mortality rates duringg these first weeks were calculated by only Caplehorn and Drummer.. They observed an extremely high -70.5/1000 person-years py- OD-mortalityy rate, that was 97.8 fold higher than the rates observed dur-ingg treatment and 6.5 fold higher compared to the after-treatment rate. .
Inn Amsterdam, different types of methadone treatment pro-grammess are available.19 Annually, approximately 2200 opiate users withh social, behavioural or psychiatric problems and tow motivation to quitt their habit of illicit drug use receive methadone in 'low-threshold programmes'' at the Municipal Health Service MHS * In the near future, abstinencee is not a realistic objective for these people and the reduc-tionn of harm caused by the continuous use of illicit opiates and other drugss is a major goal.1'0 In addition, approximately 1100 socially sta-bilisedbilised opiate users those with stable housing, regular income, health insurance-insurance- receive methadone treatment from their general practi-tionerr CP. They receive a methadone prescription once every week or
oncee every fortnight. Furthermore, approximately 300 opiate users whoo are motivated to achieve abstinence of illicit drug use participate inn methadone reduction programmes annually. The latter two pro-grammess will be regarded as 'high-threshold programmes' in this study.. Ideally, someone would enter and maintain in treatment until sustainedd abstinence of drugs is reached. In reality however, tempo-raryy withdrawal from treatment is common and people may have mul-tiplee experiences with entering and discontinuation of treatment. If thee OD mortality rates show a high increase during these transition periodss this would have a substantial effect on mortality among heroinn users.
P M A M A H U A M P P
T r a n s i t i o nn p e r i o d s n 4 o v e r d o * * m o r t a l i t y S S
Inn order to prevent methadone prescription at multiple loca-tionss to the same person at the same time, every methadone prescrip-tionn in Amsterdam is registered at the Central Methadone Register. Thiss register has been used to construct a study population and describee the incidence of overdose mortality.
Thee aim of this study is 1. to quantify the possible differences inn overdose mortality rates within different time-periods during or fol-lowingg discontinuation of methadone maintenance treatment and 2. to studyy the possible influence of patient characteristics and prior treat-mentt experience.
M e t h o d s s
SelectionSelection and description of the study population
Thee study population consists of methadone clients in methadone treatmentt in Amsterdam and former methadone clients who have left methadonee treatment less than one year ago. The study-period ran fromm January 1, 1986 until December 3 1 , 1998. In order to describe thee time at risk after methadone treatment properly, only those drug userss who were likely to stay in Amsterdam when they were not receivingg treatment were selected. Foreign heroin users are more likelyy to reside in Amsterdam temporarily, and therefore, the popula-tionn was restricted to methadone clients who had a known address in Amsterdamm and were born in The Netherlands, Surinam, the Dutch
Antilles,, Turkey or Morocco major ethnic minorities in the
Netherlands'Netherlands' A total of 5200 methadone clients met all selection cri-teria. .
ObservationObservation tint*
Observationn time was defined as a period during methadone treatment orr a period of maximal 1 year after methadone treatment. One year afterr discontinuation, individual observation time was censored. If a personn re-entered treatment, observation time started again. In case off death, observation time was terminated at date of death. Not all 52000 methadone clients participated in the study at the same time; thee average size of the population during the study-period was 2287. Thee study population was a dynamic cohort. As all methadone prescriptionss were registered at the CMR, the size of the treatment populationn and individual period of treatment participation could be measuredd precisely at any moment. Moreover, we were able to distin-guishh transition periods after (re)entering or discontinuation of treat-mentt from other periods during or after treatment. The first 10 weeks afterr entering or leaving treatment were classified from 1 to 10. After moree than ten weeks of observation time during or after treatment, personn time was classified in one category. In case of interruption, countingg started again with week 1.
P M A M A H U A M P P
B BB T r a n s i t i o n p e r i o d * a n d o v a r d o s a m o r t a l i t y
Observationn time was measured in weeks but commonly expressedd in person years py. Fifty-two first weeks of methadone-treatmentt -of multiple individuals' was equal to 1 py of first week
treatmenttreatment experience. Observation time after discontinuation of methadonee treatment was calculated individually by the summation of
alll full calendar weeks -Monday until Sunday- in which no methadone prescriptionn was registered. In total 41 448 first weeks of methadone treatmentt were observed. Only 2148 -5%- of these first weeks were absolutee first weeks, 95% of the first week observation time was con-tributedd by clients re-entering methadone treatment.
Cases Cases
Inn Amsterdam, coroners of the MHS examine all deaths that are not nat-urall or suspected to be not natural. Only in case of doubt as to whetherr death may be caused by a criminal offence, is autopsy and toxicologtcall analysis performed. Based on the coroners' records, the overdosee deaths are extracted annually. Overdose deaths are defined ass acute deaths occurring after the use of illicit drugs heroin, cocaine,cocaine, amphetamine, MDMA' or methadone, caused presumably by thesee drugs. Although drugs are often combined with alcohol or
ben-zodiazepines,1111 "i , n deaths that are due to alcohol and/or benzodi-azepiness only are not considered as overdose deaths. Information to determinee whether an overdose death is accidental or Intentional is oftenn insufficient. Therefore, both accidental and intentional overdose deathss were considered as cases in this study.
DateDate of death
Inn this study date of death was of major importance. The official date off death however, was the date of the declaration of death by the coronerr or physician. Date of death, as stated in the coroners' reports wass assumed to be the 'real date of death'. Overdose fatalities occur-ringg during treatment or within a period of 3 days after last methadonee prescription were considered as cases during treatment. Casualtiess occurring within the period from the 4th until the 17th day afterr leaving methadone treatment were considered to be overdose deathss during the first 2 weeks after treatment.
Analysis Analysis
Thee high-risk period was determined by comparing all person time of thee first, second third etc. week after starting and after leaving methadonee treatment with the number of overdose deaths occurring in thiss first, second third etc. week. We determined whether there was a higherr OD mortality rate after transition periods and, if so, how many weekss this elevation continued. Next, high risk periods were com-paredd to low risk periods using a Poisson regression analysis.
P M A M A H U A M P P
T r a n s i t i o nn p e r i o d s a n d o v o r d s s * m o r t a l i t y 8 7
Informationn available from the Central Methadone Register was usedd to distinguish other subgroups. Based on differences in gender, agee JO-years categories country of birth * The Netherlands versus otherother countries , calendar year three categories , former- treatment modalityy -low-threshold treatment or high-threshold treatment - and periodd since first treatment four categories- subgroups were distin-guished.. Multivariate Poisson-regression was used to describe differ-encess between these subgroups in respect to overdose mortality and too control for potential confounding. Differences in overdose mortal-ityy rates were expressed as rate ratios. Multivariate analyses were per-formedd using the Epidemiological Graphics. Estimation and Testing packagee EGRET."*
R e s u l t s s
DescriptionDescription of the study population
Tablee 1 describes the study population consisting of 5200 methadone clients.. It shows that the majority was male -77%, N = 4002- and born in Thee Netherlands -69%, N = 3586- As the study period proceeds the same individuall may appear in different categories of time dependent vari-abless e.g. age, duration of treatment, treatment status and modality. Manyy clients -7i%- participated in the study when they were in their 30s;; some of them also contributed person time to lower or higher age categories.. In total 50% of the observation time belongs to the 30-39 yearsyears age category.
Furthermore,, Table 1 shows that not all participants received methadonee treatment during the study period - 96 percent did so Four perr cent received only methadone in 1985, the year preceding the studyy period. Therefore, in 1986, they contributed observation time withinn the after treatment category only. Almost alt subjects *99%-contributedd to the after treatment category indicating that tempo-rary-- discontinuation of treatment is common. Most observation time wass obtained during methadone treatment -63%- During the study period,, time in treatment increased and time after treatment decreased.. Between 1986 and 1989, 57% of observation time was spentt during treatment; between 1995 and 1998, this was 68%. Many subjectss 49%* experienced both periods of low-threshold and periods off high-threshold treatment. In total, 86% of the participants received low-thresholdd and 63% received high-threshold methadone treatment. Moree than half -59%- of the participants already received treat-mentt before 1986; 41% received their first treatment during the study period.. Gradually, fewer new clients applied for methadone treatment. Inn 1986, 247 new clients entered the study population, in 1998 only 7 1 .. The average age of the 2226 people entering the population for thee first time during the study period was 30.4 years. The average age off the study-participants increased persistently from 30.2 years
P M A M A H U A M P P T r a n s i t i o nn p e r i o d s a n d o v e r d o s e rr I e I I t y NS:NS: nonsignificant Total Total Gender Gender Male e Female e CountryCountry of birth Thee Netherlands Otherr countries" YearYear of study 1 9 8 6 - 1 9 8 9 9 11 9 9 0 - 1 9 9 4 1 9 9 5 - 1 9 9 8 8
TimeTime since first MMT
Firstt year 11 -6 years 7-111 years >> 11 years Age Age << 2 0 years 2 0 - 2 99 years 3 0 - 3 99 years 4 0 - 4 99 years 5 0 - 5 99 years >> 6 0 years TreatmentTreatment modality High-thresholdd MMT Low-thresholdd MMT
DuringDuring / after treatment
Duringg treatment Afterr treatment
# ;; Number of individuals that attributed person time to this category
":": Surinam, DutchDutch Antilles.
Individuals* * NN % 5 2 0 00 100% 4 0 0 2 2 1198 8 3 5 8 6 6 1614 4 3 6 6 0 0 3 7 0 2 2 3 2 6 5 5 2 1 4 8 8 4 6 5 5 5 3 2 2 2 2 2 0 8 4 4 126 6 2 5 3 2 2 3 6 7 1 1 1 7 6 8 8 323 3 33 3 3 2 5 1 1 4 4 6 4 4 4 9 7 7 7 5 1 2 7 7 77% 77% 23% % 69% 69% 3 1 % % 70% % 7196 6 63% % 4 1 % % 90% % 62% % 4 0 % % 2% % 49% % 7 1 % % 34% % 6% % 1% % 63% % 86% % 96% % 99% % MoroccoMorocco or Turkey Person--N Person--N 299 7 2 9 222 9 1 0 6 8 1 9 9 200 3 4 8 9 3 8 2 2 9 3 8 2 2 111 5 0 2 8 8 4 6 6 1 1 2 2 2 9 9 6 9 9 100 7 0 9 7 9 2 9 9 138 8 7 2 8 4 4 144 7 4 2 6 4 7 5 5 9 6 8 8 122 2 11 2 9 2 8 166 801 188 7 4 7 100 9 8 3 yearss py % % 100% % 77% % 23% % 68% % 32% % 32% % 39% % 30% % 4% % 34% % 36% % 27% % 0% % 25% % 50% % 22% % 3% % 0% % 43% % 57% % 63% % 37% %
total may exceedexceed
5200-Fatall overdose numberr % 688 100% 61 1 7 7 61 1 7 7 20 0 36 6 12 2 7 7 24 4 29 9 8 8 0 0 19 9 40 0 8 8 1 1 0 0 2 2 4 4 42 2 26 6 90% % 10% % 90% % 10% % 29% % 53% % 18% % 10% % 35% % 43% % 12% % 0% % 28% % 59% % 12% % 1% % 0% % 32% % 68% % 62% % 38% % % p y y after r treatment t 37% % 38% % 35% % 34% % 42% % 44% % 35% % 32% % 5 1 % % 47% % 34% % 26% % 57% % 4 7 % % 36% % 28% % 26% % 20% % 29% % 43% % 0% % 100% % OD--mortatity y rate e 2.3 3 P-value e 2.7 7 1.0 0 < 0 . 0 5 5 3.0 0 0.77 < 0.001 2.1 1 3.1 1 1.4 4 6.2 2 2.4 4 2.7 7 1.0 0 0.0 0 2.6 6 2.7 7 1.2 2 0.0 0 0.0 0 1.7 7 2.7 7 2.2 2 2.4 4 N5 5 << 0.01 NS S NS S NS S
TableTable 1 Description of the study population: individuals,
T r a n s s P M A M A H U A M P P p e r i o d ss a n d o v e r d o s e m o r t a l i t y •• 9 AfterAfter treatment 2 5 0 0 0 2 0 0 0 0 >> 10, < 52 weeks 11 6-10 weeks 11 3 - 5 weeks mm 1 - 2 weeks
Duri Duri ngng treatment >> 10 weeks MM 6-10 weeks 11 3 - 5 weeks 11 1-2 weeks II 5 0 0 1 0 0 0 0 5 0 0 0
g f c — É B B ^ " * " "
0 0 0 0 1 1 CO O O i i CO O CO O O l l O l l CO O O I I o o OI I O I I rvj j Oi i CT1 CT1 O I I ' T T O I I O I I L/1 1 O I I O l l O l l CO O O l l FigureFigure 1 D e s c r i p t i o n of the s t u d y p o p u l a t i o n-- SD: 6.5, n = 2828- in 1 9 8 6 to 39.7 years • SD: 7.5, n = 2521 • in 1 9 9 8 . The me-diann p e r i o d since f i r s t m e t h a d o n e t r e a t m e n t increased f r o m 4 years •• Interquartile Range IQR: 2 - 5 - in 1 986 to t h i r t e e n years • IQR: 6 - 16 - in 1 9 9 8 .
Figuree 1 shows the size of the study p o p u l a t i o n d u r i n g the t o t a ll study p e r i o d . On average 2 2 8 7 clients were c o n t r i b u t i n g obser-v a t i o nn time to the study. In 13 years this resulted in 29 729 person-yearss of o b s e r v a t i o n t i m e . At each moment in t i m e , people were enter-ingg or leaving m e t h a d o n e t r e a t m e n t . On average 5.0% of the study p o p u l a t i o nn entered t r e a t m e n t one or t w o weeks p r e v i o u s l y and 4.5% leftt t r e a t m e n t 1 or 2 weeks previously. Most of the t i m e -69%- was o b t a i n e dd more than 10 weeks after leaving or e n t e r i n g m e t h a d o n e t r e a t m e n t .. From 1986 u n t i l 1989 64% of o b s e r v a t i o n t i m e was
P M A M A H U A M P P T r a n s i t i o nn p e r i o d s aa d o v o r d o s mm o r t o f i t y During During Weekss 1 Weekss 3 >> week After After Weekss t Weekss 3 >> week treatment treatment andd 2 -- 10 10 0 treatment treatment andd 2 -- 10 10 0 Numberr of weeks xx 10,000 7.8 8 18.9 9 70.7 7 Person--X Person--X years s 1000 1000 1.5 5 3.6 6 13.6 6 Numberr of ODD deaths g g 4 4 29 9 Incidence e perr 1000 py 6.0 0 1.1 1 2.1 1 95%% CI 3.// - 11.5 0.40.4 2.9 1.41.4 - 3.1 6.9 9 13.9 9 36.1 1 1.3 3 2.7 7 6.9 9 2 2 6 6 18 8 p-value e << 0.01 l.S S 2.2 2 2.6 6 0,40,4 - 6.0 1.0-4.8 1.0-4.8 1.61.6 4.1 N.S. .
TableTable 2 Number of weeks,
person-yearss and OD mortality rates att different time periods
obtainedd more than 10 weeks after entering or leaving treatment, dur-ingg the years 1995 - 1998 this increased to 72%. Person time spent duringg transition periods decreased during the study-period.
Sixty-eightt overdose casualties were observed. Relatively few overdosee casualties were observed among females, ethnic minorities andd high-threshold treatment participants. Moreover, the percentage off deaths during the last calendar period 1 9 9 5 - 1 9 9 8 -18%- is low comparedd to the person time that is collected during this period -30%-Tenn percent died during the first year of treatment although this cate-goryy comprised only 4% of the total person time.
High-riskHigh-risk period
Tablee 2 shows the number of overdose deaths and the number of weekss of observation time that were collected. During treatment 42 subjectss experienced a fatal OD. Nine of them suffered a fatal OD duringg the first 2 weeks after (re)entering treatment, 4 people died duringg the 3rd until the 10th week and 29 overdose casualties died moree than 10 weeks after entering methadone treatment. Overdose mortalityy rates were 6.0/1000 py, 1.1/1000 py and 2.1/1000 py respectively.. The overdose mortality rate among those who have par-ticipatedd in treatment continuously for a period longer than 2 weeks wass 1.9/1000 py.
Twenty-sixx people suffered a fatal overdose after treatment. Twoo of them died within the first 2 weeks after leaving methadone treatment.. Between the 3rd and 10th weeks after leaving methadone treatmentt 6 peopte died and the remaining 18 died more than 10 weekss after leaving methadone treatment. Overdose mortality rates
T r a n s i t i o nn p e r i o d s
P M A M A H U A M P P
nn d o v e r d o s e m o r t a l i t y •• 1
LRS:LRS: Likelihood Ratio Statistic;
DevianceDeviance on 9640 Df 440,9 LftS 7 OF 52.6 P < 0.001
FirstFirst 2 wemks after (re)enterlng treatment
VariablesVariables In the model
Bornn in the Netherlands Male e
11 until 6 years since first treatment 77 until 11 years since first treatment >> 11 years since first treatment Low-thresholdd modality
FirstFirst 2 wemks after (reentering treatment
Baseline: Baseline:
threshold! threshold!
female. female. ethnicethnic minority, year of first treatment, high-treatment,treatment, not treatmenttreatment aipha-LRS S 6.7 7 17.3 3 12.2 2 11.9 9 4.7 7 6.4 4 >> - 7.96, Df f 1 1 1 1 1 1 3 3 1 1 1 1
duringduring the first 2 weeks after Incidence Incidence (reentering (reentering :: 0.3/1000 py (9S%CI:0.1 -Beta a 1.05 5 1.62 2 1.19 9 .85 5 .66 6 -1.57 7 0.56 6 1.03 3 Rat t 1.3) 1.3) ee ratio 2.86 6 5.03 3 3.28 8 0.43 3 0.52 2 0.21 1 1.76 6 2.82 2 9 5 ** CI 1.421.42 - 5.78 p-value e < 0 . 0 1 1 2.32.3 - 11.1 1.51.5 - 7.2 0.180.18 - 1.0 0.230.23 - 1.2 0.080.08 - 0.6 1.1-2.9 1.1-2.9 .39.39 - 5.70 <0.001 1 <0.001 1 < 0 . 0 1 1 << o.os << 0.01
TableTable 3 Poisson-regression model, increased
riskk during the first 2 weeks with and without adjustmentt for confounding
weree 1.5/1000 py, 2.2/1000 py and 2.6/1000 py, respectively. In total,, overdose mortality rates outside treatment were 2.4/1000 py.
Thee high-risk transitional period seems to be limited to the firstt 2 weeks after (re)entering methadone treatment. These first 2 weekss were the very first 2 weeks for only 1 out of 9 fatalities. The otherss (re)entering treatment, two after a period of only 1 full week of absence,, 1 after a three weeks period of absence, 4 after a period betweenn 10 weeks and 6 months, the longest period of absence was 1 yearr and 1 month.
InfluenceInfluence of potential confonnéers and effect modification
Thee rate ratio of the first 2 weeks of methadone treatment compared too other periods during the study period was 2.9 -95%CI: 1.4; 5.8- This ratee ratio could be biased if a specific high-risk subgroup was over-representedd in this two weeks period. Although possibilities to control forr potential confounders were limited in this study, we could deter-minee subgroups with a potentially increased risk. The results of the multivariatee Poisson regression model conducted in order to adjust for potentiall confounding are shown in Table 3. Large rate ratios were observedd when comparing males and females • adjusted rate ratio of 3.3, 95%CII t.5-7.2- and Dutch and ethnic minorities • adjusted rate ratio of 5.0,95% CI: 2.3;; 11.1- Moreover, incidence rates were higher among clients of the
P M A M A H U A M P P
• 22 T r a n s i t i o n p e r i o d s a n d o v e r d o s e m o r t a l i t y
low-threshotdd treatment programmes -adjusted rate ratio: 1.8, 95%CI: 1.1 ; 2.9-Ratess were highest during the first year of treatment. The adjusted ratee among clients that were registered more than eleven years ago wass 4.8 - 95%CI: 1.7; 12.5- times lower. Comparing the crude and adjust-edd rate ratio, we may conclude that the potential confounders gender, ethnicity,, time since first treatment and treatment modality could not explainn the higher overdose mortality rate during the first 2 weeks afterr entering treatment.
D i s c u s s i o n n
Thee general overdose mortality rate in the study population •• 2.3/1000 pv- was low compared to figures presented in other stu-dies.1"151166 This may be related to the characteristics of the drug scene inn Amsterdam. The fact that the majority of opiate users in Amsterdam doo not inject their heroin - 60% of the opiate users treated at the MHS in 1990 and 85%% in 1999- is of major importance., l 7 n 8 Although overdose fatalities are nott restricted to injecting heroin users, intravenous use is a major risk factor."99 Among injecting opiate users in Amsterdam a higher overdose mortalityy rate * 6.3 per 1000 py. has been reported.120 injecting is especial-lyy rare among the major ethnic minorities in Amsterdam, who usually smokee their heroin • chasing the dragon • The adjusted overdose mor-talityy rate among them was 5.0 -88% 2.3-11.1- times lower than the rate amongg Dutch drug users. Furthermore, Amsterdam has an ageing and decreasingg population of chronic heroin users.118 The incidence of heroinn use has decreased since the 1980s.,z1 Consistent with findings off Warner-Smith et al. -2001- overdose mortality was not statistically significantlyy related with age,1" However, overdose mortality rates decreasedd with time since first treatment. This may be explained by a decreasingg frequency of heroin use in methadone treatment or by self selection;; those methadone clients who remain in treatment will differ fromm those who definitely leave treatment or die.
Thiss paper quantifies the differences in overdose mortality rates withinn different time periods during or following discontinuation of methadonee maintenance treatment. Conducting this study, three arbi-traryy decisions that may influence the results have been made and needd some comments. 1. The individual observation time was cen-suredd after one year without any registered prescription of methadone inn Amsterdam. Because of this, person time contribution by people whoo were no longer at risk those who became abstinent and did not relapserelapse and people who left Amsterdam- was limited. 2. Before 1989 methadonee registration was limited to the tast prescription of each week.. Therefore, the first full calendar week • Monday-Sunday- without methadonee treatment was considered as the first week after treat-ment,, in case smaller time periods or a period of 7 days was chosen, moree people should have left methadone treatment temporarily by
P M A M A H U A M P P
T r a n s i t i o nn p e r i o d s a n d o v « r d o s « m o r t a l i t y S 3
definition.. 3. All overdose deaths that occur within three days after methadonee treatment were counted as 'during treatment deaths'. First,, it was presumed that the protective effect of an elevated toler-ancee towards opiates would last some days after treatment. Secondly, ass mentioned above, the denominator within treatment was slightly overestimated.. This overestimation could lead to a underestimation of thee mortality rate during treatment if all deaths that did not occur on aa treatment day would be classified as 'after treatment overdose deaths'.deaths'. Thirdly, in case of misclassification of date of death, the wrongg date was likely to be after the last treatment contact.
OD-mortalityy rates after treatment were not significantly higherr than rates during treatment. In contrast to this, Langendam reportedd a significantly lower OD-mortality rates among injecting heroinn users of Amsterdam during methadone treatment compared to ratess outside treatment.'" The low risk of the non-injecting majority amongg Amsterdam opiate users could explain that the potential impactt of methadone treatment among all heroin users both injecting andand non-injecting- would be limited. Furthermore, rates after treat-mentt could have been influenced by periods of incarceration. An unknownn part of the time after discontinuation of treatment was spent inn prison. During their stay in prison heroin users would be at lower riskk of overdose. In case a heroin user is detained outside Amsterdam andd suffers a fatal overdose this may not be registered at the munici-pall health service in Amsterdam.
Ann important finding of this study was increased overdose mortalityy rates during the first two weeks after re-entering treatment; 6.0/10000 py compared to 1.9/1000 py during the succeeding weeks in treatment.. After discontinuation of treatment the OD mortality rate wass 2.4/1000 py. No significant increase was observed directly after discontinuation. .
Thee rate ratio of the first two weeks of methadone treatment comparedd to other periods during the study period was 2.9 •• 95%CI: 1.4; 5.8-, a difference that could not be explained by differences inn ethnicity, gender, period since registration, or treatment modality. Thee increased risk of ÜD death during the first 2 weeks is in line with otherr studies.18124 However, the OD mortality rate during the first 2 weekss observed in Caplehorn & Drummer's study was more than a 10-foldd higher; 70.4/1000 py.17 Apart from discussing the possible causess of the increase of OO mortality during the first 2 weeks after (re)enteringg methadone programmes, we will discuss the possible causess of the modest increase in the Amsterdam situation.
Thee period of increased risk during the first 2 weeks of methadonee treatment may be due to methadone itself. Accumulation off methadone is likely to occur because of the long elimination half-lifee of methadone • mean 55 hours after a single dose and 22-25 hours
P M A M A H U A M P P
1 44 T r a n s i t i o n p e r i o d s a n d o v e r d o s e m o r t a l i t y
duringduring chronic dosages-)2Sl26 Although, during the study period, the methadonee dosages prescribed at the municipal health service increased,, new clients still started at a low dose. The average dosage forr new clients during the first two weeks was 31 mg SD: 12- at 1986 andd 34 mg *SD:17- at 1998. The average dosage prescribed to clients duringg the first 2 weeks after re-entering treatment increased from 35 mgg -SDM6- at 1986 to 50 mg. SD:26- at 1998. Dosages during the remainingg periods increased from 35mg SD:15- to 61 mg
-SD:25-Becausee tolerance to opiates may vary over time, proper assessmentt and review of tolerance prior to commencement of methadonee treatment is important among both new clients demanding forr treatment and known clients re-entering treatment, especially thosee re-entries returning t o treatment after a period of incarceration. Thee period after release from prison has already been recognised as a highh risk period.1*7 IZ8 In case of drug users re-entering treatment imme-diatelyy after imprisonment, part of the overdose cases occurring dur-ingg the first 2 weeks after starting treatment might be attributed to thee increased risk after imprisonment.
Thee period during which a drug user demands treatment may bee a period of instability, stress and excessive drug use. Opiate users withh help-seeking behaviour are more likely to make a suicide attempt.1299 Therefore, accidental or intentional overdose fatalities may bee more prevalent during these periods. In Amsterdam, the step to (re)enterr treatment at one of the low-threshold methadone treatment programmess -or to demand for methadone treatment at one's own CP-mayy be lower and therefore opiate users entering treatment might be lesss desperate than in other countries.
Itt is tempting to calculate a proportion of overdose deaths that cann be attributed to the increased risk of overdose mortality within thee first 2 weeks of methadone treatment. It is true that, in case the observedd surplus of overdose mortality during the first 2 weeks could bee prevented, the number of overdose cases occurring during treat-mentt would decrease with 1596. However, in order to attribute the sur-pluss of OD deaths to methadone treatment, knowledge of a counter-factuall reality is needed; what overdose mortality rate would have beenn observed among these clients in the period that they (re)entered treatmentt in case methadone treatment did not exist at all. Taking into accountt the special circumstances around the period of entering treat-ment,, the number of prevented deaths could still exceed the number off OD deaths that can be attributed to treatment.
ODD mortality rates were lower among females and participants off high-threshold methadone treatment. The decreased risk among femalee opiate users is also observed in other studies. Females seem to havee more chance of surviving an overdose; the percentage of female overdosee cases presented in a hospital between 1992 and 1996 was 27966 which is thought to be more or less equal to the percentage of
P M A M A H U A M P P
T r a n s i t i o nn p e r i o d * a n d o v o r d o s * m o r t a l i t y B B
femalee heroin users in Amsterdam hospital-project, Municipal Health ServiceService • Females may use drugs more often in company and therefore medicall assistance can be alerted. The lower overdose mortality risk amongg married opiate users, reported by Davoli, could have a similar cause.u u
Thee lower risk among drug users at high-threshold treatment, isis due probably to thé differentiated treatment system in Amsterdam. GPss provide methadone - tablets- to opiate users in weekly take-home prescriptions.. This is unlikely to be a safer method than receiving methadonee five times a week • liquid taken on thé spot at working days andand take-home tablets during the weekend- within the low-threshöld programme.. Similarly, abstinence-oriented treatment could lead to a higherr risk for overdose mortality in case of relapse after losing toler-ancee to opiates.'" Within these high-threshold treatment populations wee observe a lower overdose mortality rate. Most probably because GP prescriptionn of methadone and abstinence oriented treatment is offeredd to specially indicated heroin users -i.e. considering abstinence orientedoriented treatment; motivated to quit drug use and considering treat-mentment from their general practitioner; social stability'
C o n c l u s i o n n
inn general, overdose mortality rates are low both during and after methadonee treatment, due presumably to the low prevalence of inject-ingg heroin use. A modest increase in overdose mortality rates during thee first two weeks after (re)entering treatment was observed. Accumulationn of methadone during the first period, inadequate assessmentt of tolerance of known clients e.g. after imprisonment-andd a concurrent period of stress, instability and extreme drug use at thee time of treatment demand are mentioned as possible explanations off the increased risk. An Australian study reported a much higher increase.. The modest increase in Amsterdam is explained by a low backgroundd risk of overdose mortality, low starting dosage and low-thresholdd to treatment.
P M A M A H U A M P P
s i t i o nn p e r i o
R E F E R E N C E S S
| ll C r ö n b l a d h L, Ö h l u n d LS, Cunne LM.
MortalityMortality in heroin addiction, the impactimpact of methadone treatment.
Actaa Psychiatrica Scandinavica 1 9 9 0 ;
8 2 :: 223 27.
!! Davoli M, Perucci CA, Forastiere F, Doylee P, Rapiti E, Zaccarelli M, Abeni
DD.. Risk factors for overdose
mortalitymortality a case control study withinwithin a cohort of intravenous drugdrug users. I n t e r n a t i o n a l Journal of
e p i d e m i o l o g yy 1 9 9 3 ; 22:273-76.
133 C a p l e h o r n JRM, Stella M, Dalton SYN, C l u f ff MC, Petrenas A M . Retention in
methadonemethadone maintenance and heroinheroin addicts1
risk of death.
A d d i c t i o nn 1 9 9 4 ; 89: 203-07.
|44 C a p l e h o r n JRM, D a l t o n SYN, Haldar F, Petrenass A M , Nisbet JC. Methadone
maintenancemaintenance and addicts' risk of fatalfatal heroin overdose. Substance
usee and misuse 1 9 9 6 ; 31: 711-21.
|55 Harding-Pink D. Methadone: one
person'sperson's maintenance dose is another'sanother's poison. Lancet 1993; 8846:8846: 665-66.
.. > Perret G, D é g l o n J, Kreek M, Ho A, La
Harpee R. Lethal methadone
intoxicationsintoxications in Geneva,
Switserland,Switserland, from 1994 to 1998.
A d d i c t i o nn 2 0 0 0 ; 9 5 : 1647-53.
\7\7 C a p l e h o r n JRM. D r u m m e r O H .
MortalityMortality associated with New SouthSouth Wales methadone programs inin 1994: lives lost and saved.
Medicall Journal of A u s t r a l i a 1 9 9 9 ;
170:170: 104-09.
|88 Zador D, Sunjic S. Deaths in
methadonemethadone maintenance treatmenttreatment in New South Wales, AustraliaAustralia 1990-1995. Addiction
2 0 0 0 ;; 95: 77-84.
|99 Plomp HN. van der Hek H, Ader HJ.
TheThe Amsterdam methadone dispensingdispensing circuit: genesis and effectivenesseffectiveness of a public health modelmodel for local drug policy.
A d d i c t i o nn 1 9 9 ) ; 91: 71 1-21.
111 0 Buning E, van Brussel C. The effects
ofof Harm reduction in Amsterdam.
Europeann A d d i c t i o n Research 1995;
/ :: 92 98.
| ll I Darke S, Zador D. Fatal heroin
'overdose':'overdose': a review. Addiction
1996;; 91: 1765-72.
|122 Ruttenberg AJ, Luke JL.
Heroin-RelatedRelated deaths: new
epidemiologicalepidemiological insights. Science
1984;; 226: 14-19.
111 3 Risser D, Schneider B. Drug related
deathsdeaths between 198S and 1992 examinedexamined at the institute of ForensicForensic Medicine in Vienna, Austria.Austria. Addiction 1994; 89: 851-57.
111 4 ECRET: Epidemiological CRaphics
EstimationEstimation and Testing package, versionversion 1.02.10. Statistics and
E p i d e m i o l o g yy Research C o r p o r a t i o n , Seattle,, W a s h i n g t o n . USA 1995.
| ll 5 Haastrup S, Jepsen PW. Eleven years
followfollow up of 300 young opioid addicts.addicts. Acta p s y c h i a t r i c a
Scandinavicaa 1 9 8 8 ; 77 : 22-26.
111 6 Engstrom A, Adamsson C, Allebeck P,
Rydbergg U. Mortality in patients
withwith substance abuse, a follow up inin Stockholm county, 1973-1984.
Thee I n t e r n a t i o n a l Journal of the a d d i c t i o n ss 1 9 9 1 ; 26: 91 106.
111 7 Hartgers C, van Santen CW, van Haastrechtt HJA. van den Hoek JAR, vann den Akker R, C o u t i n h o RA. De
HIV-prevalentieHIV-prevalentie onder druggebruikersdruggebruikers die methadon verstrektverstrekt krijgen bij de
drugsafdelingdrugsafdeling van de CC&CD. (HIV
prevalencee among d r u g users r e c e i v i n gg m e t h a d o n e at the d r u g d e p a r t m e n tt of the Municipal Health Service)) T i j d s c h r i f t v o o r sociale G e z o n d h e i d s z o r gg 1992; 70:275-79.
| ll 8 van Brussel GHA, Buster MCA. Care
forfor the future; opiate Addicts in Amsterdam.Amsterdam. Trends, Figures
19961996 1998. GG&GD A m s t e r d a m ,
A m s t e r d a mm 1999.
|199 Darke S, Ross J. Fatal heroin
overdosesoverdoses resulting from non-injectinginjecting routes of
administration.administration. Addiction 2 0 0 0 ;
9 5 :: 569-73.
|200 van Haastrecht HJA, van A m e i j d e n EJC,, van den Hoek JAR, Mientjes GHC, Baxx JS, C o u t h i n h o RA. Predictors of
mortalitymortality in the Amsterdam cohortcohort of positive and HIV-negativenegative drug users. American
Journall of e p i d e m i o l o g y I 9 9 6 ;
143:143: 380- 91.
II European M o n i t o r i n g Centre for Drug andd Drug A d d i c t i o n , EMCDDA. Study
onon incidence of problem drug use andand latency time to treatment in thethe European Union. CT. 99.EP.05.,
Lisbonn 2 0 0 0 .
1222 Warner-Smith M, Darke S, Lynskey M,
Halll W. Heroin overdose: causes
andand consequences. Addiction 2 0 0 1 ; 96:96: I 1 13-25.
1233 Langendam MW, van Brussel GHA. C o u t i n h oo RA, van A m e i j d e n EJC. The
impactimpact of harm-reduction-based methadonemethadone treatment on mortality.mortality. American Journal of Public
Healthh 2 0 0 1 ; 91: 774-80.
|244 D r u m m e r OH, O p e s k i n K, Syrjan SM,
Cordnerr M. Methadone toxicity
causingcausing death in ten subjects startingstarting on a methadone
maintenancemaintenance programme. American
Journall of Forensic Medical Pathology
1992;; 13: 346-50.
|255 I n t u r r i s i CE, Verebely K. The levels
ofof methadone in the plasma in
methadonemethadone maintenance. Clinical
Pharmacologicall Therapy 1 9 7 2 ;
13:13: 633-37.
1266 Verebely K, Volavka J, Mulé S. Resnick
R.. Methadone in man:
pharmacokineticpharmacokinetic and excretion studiesstudies in acute and chronic
treatment.treatment. Clinical Pharmacological
1277 Darke S, Ross J. Zador D, Sunjic S.
HeroinHeroin related deaths in New
SouthSouth Wales Australia, 1992-1996.
D r u g ,, A l c o h o l Dependence 2 0 0 0 :
60:60: 141-50.
1288 Seaman SR, Brettle RP, Core SM.
MortalityMortality from overdose among injectinginjecting drug users recently releasedreleased from prison: database linkagelinkage study. British Medical
Journall 1998; 316:426-28.
1299 Chodse AH, Sheehan M, Taylor C,
Edwardss C. Deaths of drug addicts
inin the United Kingdom 196 7-81.
Britishh Medical Journal 1985;
290:290: 427 28.
1300 Magura S, Rosenblum A. Leaving
methadonemethadone treatment: lessons learned,learned, lessons forgotten, lessonslessons ignored. M o u n t Sinai Journal
K n u tt R e i n a s
DepartmentDepartment of Social and Mental Health, Municipal Health Service, Amsterdam.
M a r c e ll C A . B u s t e r DepartmentDepartment of Epidemiology, Documentation and health Promotion EDC,
MunicipalMunicipal Health Service, Amsterdam
M e t t ee H i r b o
MedicalMedical Office of Health, City of Copenhagen and Frederiksberg, Denmark
P e t e rr N o l l e r
SoziologlschesSoziologlsches Forschungsinstitut (SOFI) Georg Augusi-Universitdt Cöttingen, Germany
O l i v e rr M u e l l e r EuropeanEuropean Cities on Drug Policy, Frankfurt am Main, Germany
S u s a n n aa S c h a r d t
EuropeanEuropean Cities on Drug Policy, Frankfurt am Main, Germany
H e l g ee W a a l
UniversityUniversity of Oslo, Oslo, Norway
W i mm v a n d e n B r i n k
P M A M A H U A M P P
F
A
T
A
LL
O
V
E
R
D
O
S
E
S
I NN O S L O , C O P E N H A G E N .
A
M
S
T
E
R
D
A
MM
A
N
D
F
R
A
N
K
F
U
R
T T
A b s t r a c t tInn the last decade Oslo has e x p e r i e n c e d a r i s i n g n u m b e r o f overdose deathss OD. In order t o u n d e r s t a n d this p r o b l e m , Oslo c o n t a c t e d three c i t i e s ,, A m s t e r d a m , C o p e n h a g e n and F r a n k f u r t , w h i c h are o f s i m i l a r sizee b u t have l o w e r a n d d e c r e a s i n g / s t a b l e n u m b e r s o f overdose deaths.. Available i n f o r m a t i o n on overdose d e a t h s , d r u g use and interv e n t i o n ss was g a t h e r e d and d r u g users, street leintervel p r o f e s s i o n a l s , p o l -icy-makerss and p o l i t i c i a n s were i n t e r v i e w e d . The d i f f e r e n c e s b e t w e e n ODD m o r t a l i t y f i g u r e s a n d p a r t i c u l a r l y the high and r i s i n g n u m b e r o f OD fatalitiess in Oslo c o u l d be e x p l a i n e d by d i f f e r e n c e s in m e t h o d s and d e f i n i t i o n ss used t o o b t a i n f i g u r e s , by t h e r i s i n g t r e n d s in h e r o i n use, byy a h i g h prevalence o f i n j e c t i n g use of m i x t u r e s o f h e r o i n and f l u n i -t r a z e p a m ,, by a g r o w i n g open d r u g scene and a l i m i -t e d a v a i l a b i l i -t y o f loww t h r e s h o l d t r e a t m e n t . To reduce overdose m o r t a l i t y in t h e f u t u r e , thee r e d u c t i o n and d i s p e r s a l of t h e open d r u g scene, a c c o m p a n i e d by a large-scalee i m p l e m e n t a t i o n of professional l o w - t h r e s h o l d f a c i l i t i e s is a d v i s e d . .
I n t r o d u c t i o n n
Osloo -Norway- e x p e r i e n c e d a steep increase in overdose deaths dur-ingg the n i n e t i e s . In o r d e r t o u n d e r s t a n d and deal w i t h t h i s p r o b l e m , Rusmiddeletatenn in Oslo c o n t a c t e d three N o r t h e r n European c i t i e s , A m s t e r d a mm • The Netherlands •, Copenhagen • Denmark •, and F r a n k f u r t amm Main • Germany- These are s i m i l a r in terms o f size and e c o n o m i c levell and have p o p u l a t i o n s r a n g i n g f r o m 500 0 0 0 t o 725 0 0 0 1.0 m i l -lionn to 1.8 m i l l i o n i n c l u d i n g the s u r r o u n d i n g d i s t r i c t s - and all o f t h e m facee a d r u g p r o b l e m . The e s t i m a t e d n u m b e r s o f o p i a t e users are r o u g h l yy similar, b e t w e e n 4 0 0 0 and 6 0 0 0 . The n u m b e r s of fatal ODs varyy w i d e l y , however. In 1999 the f i g u r e s f o r F r a n k f u r t N = 26- and A m s t e r d a mm N = 28- were r e l a t i v e l y low, as against a h i g h Oslo f i g u r e •• N = 104-, whereas m o d e r a t e f i g u r e s were o b s e r v e d in Copenhagen •• N = 44- A l s o , in c o n t r a s t t o the increasing t r e n d o b s e r v e d in Oslo, annuall fatal ODs have decreased in F r a n k f u r t and A m s t e r d a m and remainn stable in C o p e n h a g e n . This paper reports on a j o i n t EU-sup-portedd project t o analyse and evaluate the d i f f e r e n c e s b e t w e e n the c i t i e s . .
P M A M A H U A M P P
f B BB O v e r d o s e m o r t a l i t y i n 4 c i t i e s
Figuress for drug-related deaths are not directly comparable betweenn the countries of the European Community," so the first ques-tionn that needs to be asked is whether the variation in OD mortality figuress between cities is due to different methods and definitions.
Nott all drugs are equally dangerous. Heroin, which can cause fatall respiratory depression by influencing the respiratory centre of thee brain,IJ is considered to be the most important drug in relation to ODD mortality. The next question, then, is whether differences in OD mortalityy trends are explained by variations in the heroin epidemics. Norr do all heroin users run the same risk: this varies, depending on thee route of administration, and it is higher among injectors of heroin thann among those who inhale their drugs.131* Concurrent use of heroin andd barbiturates, benzodiazepines or alcohol increases the risk.1516 Injectionn of benzodiazepines is especially related to higher risk.17 Higherr rates of OD mortality are observed among homeless, HIV-infect-edd and single heroin users.1819 Use of heroin of an unexpectedly high purityy is thought to be the cause of fatal ODs in individual cases, but att regional level analysis of the relation between the purity of street sampless and the Incidence of OD deaths gives mixed results."81" The naturee of the drug scene -open, dispersed or bidden- might also be a factor,, but scientific evidence for effects on overdose prevalence is largelyy non-existent here. Toleration of an open drug scene is thought too be counterproductive, however112: the high availability of drugs, attractionn of new drug users, accumulation of health problems and loosee social contacts lead to chaos and an increased risk of transmis-sionn of infectious diseases. All these factors relate to characteristics off the drug user culture and patterns of use, so it is important to eval-uatee whether differences are explained by variations in these areas. Mostt overdoses do not lead to instantaneous death.113 If an ambulancee is alarmed immediately, the victim's life usually can be saved.. Fear of the police, however, can prevent drug users calling for help."** Also, if drug users in public areas try to find hiding places to injectt their drugs, safety from the police can become a danger in the eventt of an overdose.1'5 Safe injection rooms -or user rooms, or
super-visedvised injection facilities- permit an immediate response in the event off an overdose, thereby improving an individual's chance of avoiding overdosee morbidity or mortality compared to overdoses occurring elsewhere.1'66 If drug-free treatment results in long-term abstinence it minimisess the risk of an overdose. People who drop out and heroin userss who relapse after completing their treatment are at higher risk, however.1'77 Similarly, in the case of prisoners who undergo detoxifica-tionn during incarceration the period after release from prison is recog-nisedd to be a high-risk period.1'8 Methadone programmes have been shownn to reduce OD mortality by reducing heroin use, maintaining highh opioid tolerance and improving health through associated
ser-P M A M A H U A M ser-P ser-P
O v e r d o * ** m o r t a l i t y I n 4 c i t I • s 1 1 1
vices/199 Methadone can also cause OD deaths, however. Higher rates of ODD deaths are observed during the first two weeks of treatment., w u' Co-prescriptionn of heroin in addition to methadone- is also thought too reduce OD mortality: the initial results of the Dutch heroin trial indicatee an improvement in physical and mental health.1" These factorss are directly related to the interventions chosen by the city authorities.. The final question is whether the dtfferent overdose trendss in the four cities are caused by different choices and drug policyy options.
Thee project is designed to answer these questions is based on cityy reports and interviews with strategic informants in the cities, at bothh street level and the administrative and political levels. This paper focusess on the observed differences in the figures for, and trends in, ODD mortality in the four cities. First we discuss whether the differ-encess are real or due to different definitions and procedures. Having concludedd that there are real differences, we discuss the possible explanationss step by step: K differences in the onset and course of thee heroin epidemics, 2. differences and changes in the characteristics off the drug-use culture, and 3. differences and changes in interven-tionss and measures. Finally we make some recommendations on the choicess that could be made, considering the situation in Oslo in par-ticular. .
M e t h o d s s
Thee working group was set up in 1999 and consisted of a representa-tivee from each city plus a scientific consultant. The study was financedd by the Municipality of Oslo and the European Union. The reportt has recently been published.'" The city representatives are from institutionss with experience of drug research and policy at municipal level.. Meetings of the working group were organised in all the partici-patingg cities and they visited the 'drug scenes'.
Thee two-year-project was divided in three parts. The first part involvedd describing the local situation: the characteristics of the city, drugg use trends, drug treatment services and drug-related mortality. Mortalityy rates per 100 000 population were calculated, based on the numberr of residents among the OD deaths and the number of resi-dentss between IS and 65 years of age. The material is organised in cityy reports describing both the present situation and the history behindd it. The reports were used to analyse the findings and differ-encess in a time sequence perspective, in order to have any effect on thee overdose trend, changes in heroin use, drug user culture and mea-suress taken by the cities have to precede observed variations in over-dosee mortality. Obviously, effects might be masked or exaggerated by concurrentt other measures or influential factors. Also, it might be dif-ficultt to single out factors in drug policy or drug use patterns as they
P M A M A H U A M P P
1 B 22 O v « r « f o » e m o r t a l i t y i n 4 c i t i e s
oftenn relate to, and are influenced by, one another. The time sequence analysess should be interpreted with due consideration to interactions andd local variations.
Thee second and third parts of the project consisted of inter-views.. In each city, five heroin users and three street workers who deall with the OD problem professionally were interviewed using a semi-structuredd questionnaire. The selection of the respondents was basedd on diversity -e.g. drug users in- and outside treatment and a
paramedic,paramedic, police officer and social worker- and the ability to provide valuablee information • experience of a non-fatal 00 or involvement with
fatalfatal or nonfatal OD cases
-Basedd on the results of these interviews a second semi-struc-turedd questionnaire was developed to interview people who have an influencee on municipal drug policy: representatives of the police and healthh care services, an administrator responsible for drug policy and twoo politicians from the municipal council. This questionnaire dealt withh perceptions of the development of municipal drug policy and interventionss to reduce overdose mortality. This information was used too read the findings and observations in the city reports in the light of opinionss and experiences in the interviews. The final result was dis-cussedd in the research group to ensure comprehensive understanding off differences and similarities. Quotes from respondents are used to illustratee the results.
R e s u l t s s
ScrutinyScrutiny of th* OD figures
Beforee attempting to explain the differences in recorded OD mortality betweenn the cities it is important to find out whether they are real or spurious.. They could be caused by differences or changes in method-ologyy and definitions. Fig. 1 shows the number of overdose deaths accordingg to local registers in each city from the first known figures upp to 1999. The differences and similarities in the methods and defi-nitionss are shown in Table 1. In Copenhagen and Oslo ail cases sus-pectedd to be caused by an overdose are subjected to autopsy toxico* logicallogical analysis- after post mortem examination by the coroner. In contrast,, in Amsterdam and Frankfurt an autopsy is only conducted if thee coroner doubts whether death was the result of an overdose or a criminall offence.
Inn Amsterdam the figures are based on forensic coroners' reports andd collected by the Municipal Health Service. In the other cities deathss are recorded by the police. They are called 'drug-related deaths'deaths' rather than overdose deaths in Frankfurt and Oslo. Forensic medicinee in Oslo, however, confirms that almost all deaths are over-dosee deaths. In all the cities suicide by intoxication with methadone orr illicit drugs is considered as an overdose. OD deaths due to
intesti-O v e r d o s e e oo r l a
P M A M A H U A M P P
tt y i n 4 c 1 0 3 3
Amsterdam Amsterdam Copenhagen Copenhagen Franfurt Franfurt Oslo Oslo
66 0 140 0 22 0 00 0 8 0 0 66 0 44 0
i i
\ \ 20 0 to o r^ ^ r-» » r^ ^ CO O rv. . CT> CT> r^ ^ en n O O 00 0 en n _ _ CO O en n r g g eo o en n m m 00 0 en n CO O 01 1 LH H CO O en n 10 0 00 0 en n r--CO O en n CO O oo o en n CTl l CO O en n o o en n en n _ _ en n en n IN N en n en n m m en n en n en n en n L/1 1 en n en n <0 0 en n en n r-~ ~ en n en n CO O en n en n en n en n en n FigureFigure 1 OD deaths a c c o r d i n g to locall r e g i s t e r snall s m u g g l i n g o f cocaine have o n l y been observed in A m s t e r d a m •NN = 4- The figures for Copenhagen Fig. / • are l i m i t e d t o o f f i c i a l l y r e g i s t e r e dd residents of t h e city. In 1999 13 non-residents died f r o m an o v e r d o s e :: i n c l u d i n g those d e a t h s , 57 overdose deaths w o u l d have beenn r e p o r t e d . Non-residential OD deaths in the city accounted for a largerr p r o p o r t i o n o f OD deaths in Oslo -36%- and the m a j o r i t y of the casess in A m s t e r d a m • 54% • In F r a n k f u r t non-residents were i n c l u d e d b u tt c o u l d not be d i s t i n g u i s h e d . None o f the cities i n c l u d e d overdoses inn residents w h o died e l s e w h e r e . Table 2 shows the number o f 0D casess in 1999 and the rates per 100 0 0 0 residents in the 1 5 - 6 4 age g r o u p .. Expressed in rates per 100 0 0 0 r e s i d e n t s , the differences b e t w e e nn the numbers of OD deaths in Copenhagen and Oslo are s m a l l -err than i n d i c a t e d in Figure 1.
Thee average age o f t h e deceased was lowest in F r a n k f u r t -31
years-,years-, as a g a i n s t 34 in A m s t e r d a m , 35 in Oslo a n d 37 in
C o p e n h a g e n .. In A m s t e r d a m there was a big age difference b e t w e e n r e s i d e n t i a ll and n o n - r e s i d e n t i a l OD d e a t h s : the average was 28 and 41 r e s p e c t i v e l y . .
i a « « P M A M A H U A M P P O v a r d o s aa m o r t a l i t y II ii c i t i e s s City City Oslo o Copenhagen n Frankfurt t Amsterdam m
*DRD:*DRD: defined asas drug-related deaths, present proportion ofof non-00 deathsdeaths expected to be minimal in Oslo, possibly larger in Frankfurt. $$ Suicides using illicit drugs or methadone.
Derivedd from forensics s forensics s forensics s forensics s Toxicology y Conduct t Standardd non-natural Standardd non-riatural Doubtt OD/criminal offence Doubtt OD/crimfnal offence
Numberr may include e otherr DRD* + + --+ --+ --Intentionall OD includedd & Non-residents s Included d Fatall OD off inhabitants occurring g outsidee the city
TableTable Ï Definition, methods to
derivee the number of OD cases shown inn Figure 1
Obviously,, the official figures are not fully comparable and differencess should be considered indicative rather than exact. The dif-ferencess are large, however, and we can say that the number of OD fatalitiess in 1999 was relatively low in Amsterdam and Frankfurt, moderatee in Copenhagen and relatively high in Oslo. As records of locall OD mortality used methods and definitions that were constant overr time, the trends should be uninfluenced by the differences. Oslo hadd a low, stable number of overdoses during the 1980s, with a steady increasee after 1990 and a possible levelling-off in the last years recorded.. OD mortality increased slowly in Amsterdam over the 1 9 7 7 - 1 9 8 44 period, since when it has decreased in the same gradual manner.. In Frankfurt there was a relatively stable, moderately high fig-uree from 1979 to 1985, followed by a sharp increase from 1985 and a dramaticc decrease from 1992. OD fatalities in Copenhagen have remainedd relatively stable since recording started in 1982. Also, Copenhagenn and Amsterdam never experienced peaks as high as those observedd in Oslo and Frankfurt. During the project the perceptions of streett professionals 'ambulance personnel, police officers, social work-ers'ers' confirmed that the trends were decreasing, increasing or stable ass shown in Table 2.
DifferencesDifferences arising from the course of the heroin epidemics
Onee possible explanation for these differences is the timing of the heroinn epidemics. Although heroin was first seen in all four cities dur-ingg the seventies, the epidemics developed differently. Amsterdam
O w a r d o a * * P M A M A H U A M P P t yy I n 4 c i t i e s i a * * City City Oslo o Copenhagen n F r a n k f u r t t A m s t e r d a m m
## Cases limited to residents; this differentiation was Impossible
InIn Frankfurt Registered d ODD cases 1 9 9 9 104 4 57 7 26 6 28 8
wherewhere all eases are
N u m b e rr o f r e s i d e n t i a ll cases 67 7 4 4 4 <26 <26 13 3 consideredconsidered to be residents Ratess per 1000 0 0 0 p o p u l a t i o n 1 5 - 6 44 Included 2 0 . 0 0 14.2 2 << 5.9 2.5 5 Trends s
Increasee since 1989, stabilising since 1996 Stablee since first figures in 1982
Sharpp increase 1985-1991, sharp decrease afterwards Increasee until 1984, slow decrease afterwards
TableTable 2 Present OD figures • 1999-,
ratess per 100 000 population and agreedd 0 0 mortality situation as a basiss for this study
andd Copenhagen experienced a major increase in the prevalence of heroinn use during the mid-seventies and the first part of the eighties. Inn Frankfurt the number o f heroin users strongly increased during the latterr half of the 1980s and in Oslo during the 1990s. The overdose curvess correlate with the presumed prevalence of heroin use. Copenhagen,, however, never saw such a peaked curve as those found inn the other cities. Also, the sharp increase and decrease seen in Frankfurtt were unparalleled in the other cities. Therefore future trends fnn Oslo cannot be predicted on the basis of the material.
Whatt we see is that the timing of a heroin epidemic is only onee of several influential factors. One such factor is immigration of heroinn users. An increasing number of non-residential predominantly German-German- heroin users migrated to Amsterdam during the first half of thee 1980s, after which the number gradually decreased. The peak of thee Frankfurt curve is influenced by the immigration of heroin users fromm neighbouring cities. We are currently seeing an increasing num-berr of non-residential heroin users in Oslo: the steep increase may be influencedd by immigration. As Table 2 shows, however, even if ÜD casess are limited to residents, the prevalence in Oslo is still high.
P M A M A H U A M P P
I B BB O v e r d o s e m o r t a l i t y I n 4 c i t i e s
CulturalCultural explanations-characteristics of drugs, drugdrug usars and drug scene
Too what extent could differences in OD mortality be caused by cultur-all factors, such as differences and changes in the kind of drugs being used,, the route of administration, concurrent use of other drugs, the purityy and price of drugs and the nature of the drug scene?
Shiftingg drug trends
Osloo experienced a shift from an amphetamine to a heroin-oriented drugg scene in the years preceding the increase in overdoses, and this hass obviously contributed to the increase in mortality. During the 1990ss cocaine gradually became a more dominant drug in the heroin sceness in Frankfurt and Amsterdam,1" which might be influential in the decreasingg trend in OD mortality in these cities. The youth culture, and associatedd drug use, has gradually changed in all four cities. Among thee changes is an increase in the use of MDMA and cocaine -snorting-sincee the late 1980s in Amsterdam,1" Frankfurt and Copenhagen, wheree youngsters are reported to be avoiding heroin. In Oslo an increasee in the use of MDMA, and to a lesser extent cocaine, is a phe-nomenonn of the late nineties. Young people are also starting with heroin,, however.
Routee of administration
Differencess in the route of administration no doubt explain part of the differencee in OD mortality between Amsterdam and the other cities. In ann Amsterdam sample of methadone clients 37% ever injected heroin andd 13% did so last month. The main route of administration is inhal-ingg -chasing the dragon- which is obviously less dangerous in terms o ff OD m o r t a l i t y • in the same sample 88% ever inhaled and 50% did so last month- T h e numberr of needles exchanged almost halved during the 1990s, which indicatess a further decrease in injecting drug use and provides an-otherr explanation for the slow decrease in the numbers of OD deaths inn Amsterdam. In the other three cities injecting remains the most commonn route of administration, though Copenhagen has experienced aa gradual increase, to a level of 5096 of users inhaling by 1 999. In Oslo almostt all use is through injection -95% in 1999*
Mixturess of drugs
Thee risk-enhancing effect of the concurrent use of heroin and benzo-diazepiness and/or alcohol is well acknowledged. Differences and trendss in combining these substances are thought to be responsible forr part of the differences and trends in OD mortality. Intravenous use off Rohypnol, • flunitrazepam • is often mentioned as a risk factor in Oslo.. In the 1980s the use of Rohypnol*, was thought to be associ-atedd with increased OD mortality in Amsterdam.
