- Email: [email protected]
Mortality and employment after in-patient opiate detoxification
European Psychiatry 27 (2012) 294–300
Original article
Mortality and employment after in-patient opiate detoxification A. Naderi-Heiden a,*, A. Gleiss b, C. Ba¨cker a, D. Bieber a, H. Nassan-Agha a, S. Kasper a, R. Frey a a b
Division of Biological Psychiatry, Department of Psychiatry and Psychotherapy, Medical University of Vienna, Waehringer Guertel 18-20, 1090 Vienna, Austria Core Unit for Medical Statistics and Informatics, Section of Clinical Biometrics, Medical University of Vienna, Spitalgasse 23, 1090 Vienna, Austria
A R T I C L E I N F O
A B S T R A C T
Article history: Received 29 January 2010 Received in revised form 6 May 2010 Accepted 8 May 2010 Available online 22 July 2010
Aim. – We considered that completed opiate detoxification resulted in increased life expectancy and earning capacity as compared to non-completed detoxification. Methods. – The cohort study sample included pure opioid or poly-substance addicts admitted for voluntary in-patient detoxification between 1997 and 2004. Of 404 patients, 58.7% completed the detoxification program and 41.3% did not. The Austrian Social Security Institution supplied data on survival and employment records for every single day in the individual observation period between discharge and December 2007. Statistical analyses included the calculation of standardized mortality rates for the follow-up period of up to 11 years. Results. – The standardized mortality ratios (SMRs) were between 13.5 and 17.9 during the first five years after discharge, thereafter they fell clearly with time. Mortality did not differ statistically significantly between completers and non-completers. The median employment rate was insignificantly higher in completers (12.0%) than in non-completers (5.5%). The odds for being employed were higher in pure opioid addicts than in poly-substance addicts (p = 0.003). Conclusions. – The assumption that completers of detoxification treatment have a better outcome than non-completers has not been confirmed. The decrease in mortality with time elapsed since detoxification is interesting. Pure opioid addicts had better employment prospects than poly-substance addicts. ß 2010 Elsevier Masson SAS. All rights reserved.
Keywords: Mortality Employment Opiate detoxification Opioid dependence Poly-substance dependence
1. Introduction National data sources have estimated the number of opiate abusers in Austria to be about 33,000 [15]. It was estimated that at least 27% of drug users had died within 20 years in a cohort of chronic drug users [31]. Detoxification is a necessary step prior to abstinence-oriented therapy. Generally, voluntary participation in detoxification programs with eligibility criteria is a positive prognostic feature in abstinence-oriented therapy [19]. The efficiency of opioid detoxification is critically discussed with respect to relapse and abstinence rates, respectively. Diversity in study design limited the comparability of studies. Results can be biased by patients’ inclusion criteria (e.g. co-morbidity), mode of detoxification and aftercare as well as abstinence definitions. Abstinence rates after step-by-step detoxification of methadone vary between 22 and 86% [19]. Both clinical experience and neurobiological evidence indicate that opioid dependence is a chronic relapsing disorder with potential social neglect and high mortality risk [32]. In the final
* Corresponding author. Tel.: +43 1 40 400 3568; fax: +43 1 40 400 3099. E-mail address: [email protected] (A. Naderi-Heiden). 0924-9338/$ – see front matter ß 2010 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.eurpsy.2010.05.002
analysis, the aim of detoxification and abstinence-oriented treatment is increased earning capacity and increased life expectancy. Our cohort study was conducted to evaluate the implication of detoxification treatment with respect to mortality and occupation in a follow-up period of up to 11 years. In-patients suffering from pure or poly-substance dependence were treated with slow release oral morphinsulphate/-hydrochloride, methadone or buprenorphine in a gradual detoxification program with strict guidelines that took about 2–4 weeks. We compared patients who completed in-patient opioid detoxification with those who failed to complete this program. Theoretically, detoxified patients with future prospects of opiate abstinence should have a better prognosis. The analyses considered a division into pure opioid addicts and poly-substance dependent patients. 2. Subjects and methods 2.1. Study population, in-patient therapeutic detoxification programme 1997–2004 The study sample included all opioid addicts (aged 15 or older) with clear commitment to abstinence, admitted at the Department of Biological Psychiatry, Medical University of Vienna, Austria, in the period between January 1997 and December 2004. Only
A. Naderi-Heiden et al. / European Psychiatry 27 (2012) 294–300
patients for voluntary in-patient detoxification were included. Patients were excluded from the present analyses if the following conditions occurred: [1] emergency treatments; [2] partial inpatient detoxification from illicit drugs, benzodiazepines or alcohol but continuously prescribed opioid maintenance therapy and; [3] in case of readmission to the withdrawal program. Note, that only the first entry (index-episode) was computed. In 1997, immediately before admission 73% of the subjects were treated by oral methadone or oral morphine maintenance therapy. Owing to revised treatment guidelines, the proportion of patients with former methadone, morphine or buprenorphine maintenance therapy increased year by year and was finally higher than 90% from year 2000 on. However, most of the patients had been noncompliant with maintenance programs and poly-substance dependent, respectively. We discriminated between patients with pure opioid dependence (ICD-10: F11.2) and poly-substance dependence (ICD-10: F19.2, including opioid, benzodiazepine, alcohol, cannabis, cocaine and amphetamine). The diagnosis of substance dependence was based on data derived from semi-structured diagnostic interviews and from existent medical records. A human immunodeficiency virus (HIV) test was performed by routine. Patients with HIV were detoxified in extraordinary cases only (n = 2). Detoxification treatment, usually 2–4 weeks in duration, was proposed. The patients’ opioid dose was replaced by an equivalent dosage of slow-release oral morphinsulphate/-hydrochloride, methadone or buprenorphine. After tapering off, the success of detoxification was confirmed by opiate, methadone, cocain, amphetamine, cannabinoide, barbiturate and benzodiazepine free urine specimen. The total sample of 404 patients (mean age 27.4 7.0 y) comprised 290 males (71.8%) and 114 females (28.2%). Twenty-five percent of the patients were recorded as employed (males 27%, females 20%). There were 125 patients with opioid dependence (30.9%) and 279 patients with poly-substance dependence (69.1%). Part of the patients completed the detoxification treatment (with drug-free urine specimen) and was discharged with abstinence-oriented recommendations. Another part discontinued the therapy and further illicid opioid intake or opiode maintenance therapy could be expected. To form the two comparison groups, patients were divided in completers and non-completers. From 404 patients, 237 (58.7%) completed the detoxification program (mean age: 27.7 6.3 y) and 167 (41.3%) did not (mean age: 27.2 7.5 y). The in-patient treatment of completers (median: 17 days) lasted longer than the in-patient treatment of noncompleters (median: 10 days). Completion rate was 63.2% in patients with pure opioid dependence and 56.6% in patients with polysubstance dependence (p = 0.215). Completion rate was 56.6% in male patients and 64.0% in females (p = 0.169). In the period between the index-episode and December 2004, a proportion of 59 out of 404 patients (14.6%) were readmitted at least one more time. The readmissions (n = 68) did not count for our analyses. 2.2. Study design and data sources This is a retrospective cohort study with arbitrarily chosen administrative censoring on Dec 31st, 2007. Data were collected in 2008. The Main Association of Austrian Social Security Institutions informed about daily documented employment status and data of death when indicated. Data on mortality was gained for the total sample of 404 subjects. The cause of death is not documented by the Main Association of Austrian Social Security Institutions. Data on employment was gained for 399 subjects only, since the identification numbers of five patients were not correctly delivered. The observation period lasted between a minimum of 3 and a maximum
295
of 11 years, i.e. from the day of patients’ discharge (between 01/1997 and 12/2004) to Dec 31st, 2007. At each single day in the observation period, patients were rated ‘‘employed’’ or ‘‘unemployed’’. In case of ‘‘no social security’’, patients were judged as ‘‘unemployed’’. The employment rate (%) was calculated for each subject. This is the duration of employment (specification in days) divided by the duration of the observation period (specification in days). 2.3. Confidentiality of data The protocol was approved by the local Ethics and Scientific Review Committee (EK.Nr: 489/2007). Personal identifiers were used following the rules of privacy regulation. For statistical evaluation only indirect individual-related data was used (§ 4 Z 1, Data Protection Act). The investigation set out to answer the following questions: have completers a lower mortality risk after discharge than noncompleters? have pure opioid dependents a lower mortality risk after discharge than poly-substance dependents? have completers more working days between discharge and the endpoint of the observation period than non-completers? have pure opioid dependents more working days between discharge and the endpoint of the observation period than polysubstance dependents? Gender differences and the influence of age at time of admission will also be explored with respect to mortality risk and working days. 2.4. Statistics Regarding to diagnostic features and gender, proportions of completers and non-completers were compared by Chi2 tests. In order to compare the observed number of deaths with what would be expected in the general population, the following approach was used [13]: the expected number of deaths within each year after discharge was calculated under the assumption that each patient’s risk would be equal to the risk in an age and sex matched segment of the general population. This risk was based on the life table provided by the Statistik Austria for the year 2002 [28]. The standardized mortality ratio (SMR) was the ratio of observed deaths to expected deaths. Ninety-five percent confidence intervals (CI) for SMRs as well as pvalues for comparing SMRs between two groups were calculated as proposed by Breslow and Day [8]. Observed and expected numbers of deaths were compared with each other using a Chi2 test. The calculation of percentages (and 95% CI) of patients reaching a certain time after discharge was based on the corresponding Kaplan-Meier survival distribution estimates. Kaplan-Meier curves are compared between groups using log-rank tests. The potential influence of various explanatory variables on survival after discharge was investigated using single- and multi-variable Cox proportional hazards models. Hazard ratios and 95% CI were given only for the multi-variable model since single-variable results were virtually the same. Both methods, Kaplan-Meier estimates and Cox models, are able to appropriately deal with the form of administrative censoring given in our data set. Employment rates were calculated from the daily raw data provided by the Main Association of Austrian Social Security Institutions as the number of days at which a patient was registered as being employed, divided by the total number of days within the observation period. This total number of days includes days where no information was available, thus producing worst case estimates. Employment rates are presented as median, quartiles, minimum and maximum due to their right skewed
A. Naderi-Heiden et al. / European Psychiatry 27 (2012) 294–300
296
Table 1 Number of subjects, person-years of follow-up, number of deaths and crude death rates per 1000 person-years of follow-up for the total sample and different subgroups.
Total sample Completers Non-completers Pure opiod dependence Poly-substance dependence Males Females a
Number of subjects
Person-years of follow-up
Number of deaths
a
404 237 167 125 279 290 114
2997.02 1714.91 1282.11 1303.36 1966.66 2090.57 906.45
35 21 14 8 27 30 5
11.68 12.25 10.92 6.14 13.73 14.35 5.52
Death rates per 1000 person-years of follow-up
Number of deaths/person-years of follow-up 1000.
distribution. The potential influence of various explanatory variables on the employment rate was investigated using a singleand multi-variable, proportional odds, logistic regression model where the dependent variable, employment rate, is categorized as follows: [category 1] employment rate of 0.0%, [category 2] employment rate of greater than 0.0 up to 17.8%, [category 3] employment rate greater than 17.8% (see results section: the median employment rate in patients with at least one working day turned out to be 17.8). Odds ratios with 95% CI are given only for the multi-variable model since single-variable results are virtually the same. No two-way interactions were significant. The reported p-values are the results of two-sided tests. pvalues less or equal to 0.05 were considered to be statistically significant. All computations have been performed using SAS software Version 9.2 (SAS Institute Inc., Cary, NC, USA, 2008). 3. Results
in males and 11.11 (CI 3.58–25.93) in females (p = 0.888), respectively, 10.77 (CI 5.50–16.90) in completers and 10.07 (CI 6.66–16.46) in non-completers (p = 0.982). The SMR was 7.08 (CI 3.05–13.95) in pure opioid addicts and 12.27 (CI 8.09–17.86) in poly-substance addicts (p = 0.228). Three subjects died within the first four months. The SMR for this short period was 15.8 (three deaths observed, 0.19 deaths expected), which did not differ significantly from the SMR 8.8 in the rest of the observation period (p = 0.560). The annual life table can be seen in Table 2. Relating to the whole observation period, the number of deaths in the patient group (Nd) was significantly higher than the number of expected deaths (Ne), i.e. the mortality risk was significantly higher in our patients than in the general population (p < 0.001, see Table 2). During the first years after discharge, SMRs were between 13.5 and 17.9, thereafter they fell with time. The Kaplan-Meier estimate of the survival distribution after discharge showed the following results for the total group:
3.1. Mortality Thirty-five subjects (30 males, five females) died in the observation period, 21 of which finished past detoxification treatments as completers, while 14 were rated as non-completers. Eight deceased subjects were diagnosed as pure opioid dependent, 27 as poly-substance dependent. Death rates per 1000 personyears of follow-up for the total sample and different subgroups can be seen in Table 1. With respect to deceased patients, a proportion of 27/35 (77%) had been treated by methadone, morphine or buprenorphine maintenance programs prior to the detoxification treatment, 8/35 (23%) had not. We have no information about the follow-up therapy and the cause of death. Between discharge and the year 2007, the SMR was 10.45 (CI 7.28–14.53) in the total group. The SMR was 10.56 (CI 7.13–15.08)
98.8% 95.5% 92.3% 90.4%
(CI (CI (CI (CI
97.1–99.5) 93.0–97.2) 89.2–94.6) 86.8–93.1)
survived survived survived survived
the the the the
1st year; 3rd year; 5th year; 10th year.
Fig. 1 shows the Kaplan Meier plot for the subgroups completers and non-completers, respectively. Inter-group differences were not significant (p = 0.806). Fig. 2 shows the Kaplan Meier plot for the subgroups pure opioid dependence and poly-substance dependence, respectively. Inter-group differences were not significant (p = 0.220). Four explanatory variables were investigated with respect to possible effects on the hazard of death in the observation period: discontinuation of detoxification treatment, age at the beginning of
Table 2 Life table of 404 opioid-addicted patients in an observation period up to 11 years after in-patient detoxification treatment. Time perioda (year)
Subjects starting annual observation period (N)b
Observed number of deaths (Nd)
Expected number of deaths (Ne)c
Standardized mortality ratio (Nd:Ne)
1st 2nd 3rd 4th 5th 6th 7th 8th 9th 10th 11th
404 399 393 386 349 317 286 250 198 132 78
5 6 7 6 6 2 1 1 1 0 0
0.37 0.38 0.39 0.39 0.37 0.35 0.33 0.30 0.24 0.16 0.07
13.54 15.77 17.90 15.55 16.28 5.69 3.01 3.36 4.23 0.00 0.00
a
After individual discharge from detoxification unit. The number of patients starting observation period decreased annually and depends on the number of patients lost by death or lost by administrative censoring on December 31st 2007. c The column shows the expected number of deceases in the annual patients’ sample under the assumption that each patient’s risk would be equal to the risk in the general population. The latter is based on the life table calculated by the Statistik Austria and comprised an age and sex matched general population of the year 2002 (21). Observed deaths versus expected deaths in the observation period (Nd vs. Ne): p < 0.0001. b
A. Naderi-Heiden et al. / European Psychiatry 27 (2012) 294–300
297
Table 3 Median employment rate (%) as well as 1st and 3rd quartile (%) in the observation period in different subgroups (total sample size = 399). Subgroups
Detoxification Completers Non-completers Dependence Pure opioid Poly-substance Gender Males Females
Fig. 1. Survival over time since discharge from psychiatry unit is shown for a cohort detoxified between 1997 and 2004. The graph shows Kaplan-Meier product-limit estimate, including deaths of completers and non-completers. Inter-group differences were not significant (p = 0.806). Note that the vertical axis scale starts at a minimal survival probability of 0.85.
the detoxification treatment, poly-substance dependence and gender. None of these showed a significant influence on survival: hazard ratios from the multi-variable Cox model were 0.85 (CI 0.43–1.66) for non-completers (p = 0.614), 1.60 (CI 0.72–3.53) for poly-substance addicts (p = 0.247), 1.15 (CI 0.73–1.82) for each 10 year increase in age (p = 0.549), and 0.41 (CI 0.16–1.07) for females (p = 0.070). 3.2. Employment The total sample (n = 399) showed a median employment rate of 9.1% (min: 0.0%, 1st quartile: 0.2%, 3rd quartile: 32.6%, max. 99.4%) in the observation period. The median employment rate of different subgroups can be seen in Table 3. The results were in favour of completers, pure opioid addicts and males.
1st quartile
Median employment rate
3rd quartile
233 166
0.3 0.1
12.0 5.5
39.1 27.7
122 277
1.8 0.0
16.3 6.1
49.4 25.1
285 114
0.2 0.0
11.3 5.7
39.1 22.1
Table 4 The total sample is subdivided into different subgroups (concerning detoxification, diagnosis and gender) and assigned to three employment rate categories. In case of at least one regular working day, the subjects showed a median employment rate of 17.8 %. Subgroups
Detoxification Completers Non-completers Dependence Pure opioid Poly-substance Gender Males Females
Fig. 2. Survival over time after discharge from psychiatry unit is shown for a cohort detoxified between 1997 and 2004. The graph shows Kaplan-Meier product-limit estimate, including deaths of subjects suffering from pure opioid dependence and poly-substance dependence. Inter-group differences were not significant (p = 0.220). Note that the vertical axis scale starts at a minimal survival probability of 0.85.
N
Employment rate (%) 0.0 % category 1
0.0–17.8 % category 2
> 17.8 % category 3
22.3 24.7
36.1 41.6
41.6 33.7
16.4 26.3
34.4 40.1
49.2 33.6
22.1 26.3
36.1 43.9
41.8 29.8
In the total group, 93 subjects (23.3%) did not work one single day (employment rate: 0.0%) in the observation period. The rest of the subjects (n = 306) showed a median employment rate of 17.8%, i.e. 153 subjects (38.3%) were employed less than 17.8% of the days in the observation period, 153 (38.3%) more than 17.8%. Table 4 introduces three categories: [category 1] employment rate of 0.0%, [category 2] employment rate of 0.0–17.8%, [category 3] employment rate greater than 17.8%. The completers, pure opioid addicts and males make for higher percentages of working subjects in [category 3]. The multi-variable logistic regression analysis investigated (dis) continuation of detoxification treatment, age at the beginning of detoxification treatment, pure opioid or poly-substance dependence and gender. The odds for being in a lower category (fewer days of employment) were 30.1% higher in non-completers than in completers (OR 1.301, CI 0.895–1.892, p = 0.168); 86.9% higher in poly-substance addicts than in pure opioid addicts (OR 1.869, CI 1.243–2.809, p = 0.003); 69.4% higher in females than in males (OR 1.694, CI 1.121–2.559, p = 0.012); 39.8% higher for each 10 additional years of age (OR 1.398, CI 1.071–1.827, p = 0.014). 4. Discussion One main question could be confirmed by a statistically significant finding: pure opioid dependent patients had a higher employment rate in the years after discharge than poly-substance dependent patients (logistic regression analysis, p = 0.003). The median employment rate was 16.3% in pure opioid dependence and 6.1% in poly-substance dependence. The diagnostic classification was documented at the time of admission at the detoxification unit. The aftercare of our patients has varied or was unknown during an observation period of up to 11 years. We are not aware of
298
A. Naderi-Heiden et al. / European Psychiatry 27 (2012) 294–300
consumption patterns after discharge. Maybe co-consumption patterns of our patients did not change considerably over years and pure opiate addicts had better working abilities. Co-consumption of benzodiazepines may have substantial impact on morbidity, mortality and clinical course [1]. A higher abstinence rate and thus a higher employment rate in pure opiate addicts could be another reason for our finding [9]. Additional analysis showed that male gender (p = 0.012) and younger age (p = 0.014) were significantly associated with high employment rates. For both conditions, this can be generally explained by the better access to the labour force. It can be assumed that females are more involved in childcare. There was no indication for more chronicity or damage in our female sample, but a review article states that females are at greater risk for relapse following abstinence [5]. Young subjects might have better working abilities with respect to shorter duration or less severity of illness. In a group of 11,400 problem drug users attending 38 different treatment centres in Rome, 61% of the study population were recorded as employed. Males were more likely to be employed than females (64% versus 48%) [3]. In our study, the proportion of employed subjects was only 25% while admitted to in-patient treatment. Indeed, tolerant eligibility criteria for admission to our detoxification unit included social exclusion, co-morbidity and multiple drug use. Opiate detoxification was offered not only to adherent patients but also to patients non-compliant with maintenance programs, which is a formal stance opposite to that in many countries. Other authors stated that abstinence-oriented interventions are indicated only in a few motivated patients with stable living conditions and adequate familial, social and therapeutic support [32]. In practice, non-compliant addicts or those with otherwise displaying intolerable behaviour are in fact subject to involuntary detoxification or prison. Fifty-nine percent of the patients completed the detoxification program. Other Austrian authors also found completion rates of about 50% with either methadone or slow-release oral morphine tapering [20]. We postulated that successfully detoxified patients should have better working abilities than non-completers, but completion of detoxification treatment did not significantly increase the probability of high employment rates during the observation period. The low employment rate might be an indication of high relapse rates in completers. The broad public health system in Austria allocates not only abstinence-oriented therapy but also low-threshold maintenance therapy with methadone, morphine tablets or buprenorphine. This could have been helpful anyhow. The low employment rate might also be influenced by not reported illegal work and the tolerant Austrian social security system for long-term non-workers, which can be a matter of survival or convenience as well. Few studies have assessed employment functioning after detoxification: [1] Sees et al. compared outcomes of patients with opioid dependence treated with methadone maintenance versus psychosocially enriched methadone assisted detoxification and showed no difference between groups in employment functioning after 12 months [27]. [2] Hubbard et al. showed a 10% increase in full-time employment among drug-free outpatients after 5 years [18]. We are not aware of other follow-up studies comparing working days in completers and non-completers of detoxification treatments. Significantly better abstinence-oriented behaviour was detected amongst patients who completed detoxification and went on to spend at least 6 weeks in a rehabilitation unit [16]. In contrast, the authors observed no significant different abstinence rates between non-completers and completers who had no aftercare on the majority of measures of drug use during follow-up. Unfortunately, earning capacity was not included as an outcome criterion. Mortality rates among opioid addicts are higher than for the general population. Bargagli et al. stated mortality rates in eight
study sites in Europe with comparable methodology [2]. All cohorts consisted of opiate users entering treatment during the period 1990–2000. The authors did not specify the kind of treatment. SMRs were calculated for the total observation period and vary between 9.2 and 25.6 in different countries. In the area of Vienna, Bargagli et al. reported an SMR of 9.8 without gender differences [2]. Numerically, 195 from 4150 patients died. The cause of death varied as follows: 50% drug-related, 19% HIVrelated, other causes 31%. In our Viennese study, the sample showed nearly the same SMR of 10.5 during 11 years. The mortality risk of the detoxified patients was significantly higher than in the general population (p < 0.0001). Other Austrian authors stated a SMR of 17.8 over a 3-year study period in a cohort of opiate-users enrolled in opioid maintenance treatment from 1995–1997 [25] and a SMR of 29.13 in cohort of patients enrolled in maintenance therapy from 1998–1999 [4]. In the latter sample, many HIV seropositve patients were treated and the cause of death was in 35% related to AIDS. Methadone is an effective maintenance treatment of opioid dependence. About 90% of our patients received opioid replacement therapy before the index episode. We assume that noncompleters of detoxification program or relapsed completers might have chosen this treatment again. Degenhardt et al. revealed in a follow up study of about 42,000 persons receiving opioid pharmacotherapy that time in treatment (methadone or buprenorphine as opioid replacement therapy) was associated with lower mortality than time out of treatment, with an overall ‘‘in treatment SMR’’ of 4.5, compared to an ‘‘out of treatment’’ SMR of 8.0 [12]. When compared methadone maintenance therapy and treatments that did not involve opioid replacement therapy (i.e., detoxification), a systematic review did not show a statistically significant superior effect on criminal activity and mortality [21]. Long-term dependent users and patients on long term opiate agonist treatments respectively might have many risk factors for poor cardiac, pulmonary, hepatic and renal health. Darke et al. confirmed that systemic disease was prominent among cases of fatal opioid toxicity cases with increasing age [10]. The prevalence of physical health disorders among older methadone patients is very high [26]. This severe illness profile in its own right forms an important reason to opt for detoxification treatment. This treatment justification needs to be carefully considered, if one is aware that patients choosing this route face an inordinate initial mortality risk. Our sample revealed consecutively high SMRs (from 13.5 to 17.9, Table 2) in a five-year period after detoxification, but relatively low SMRs thereafter. Note, however, that due to the small numbers of observed and expected deaths the variability of the reported SMRs is quite high (expressed by broad CI). The decrease in SMR may partly be due to increased mortality in the general population over time, leading to a relative reduction in the ratio. It is suggested that those who take the highest risks will probably die earliest, leading to a decrease in mortality for the cohort as a whole. However, the final drop in SMR after detoxification is greater than that, which occurs in agonist maintained patients where SMR curves are rather flat during the observation period [12]. Even though opioid dependency is a complex and chronically relapsing disease with high risks of morbidity and mortality a long process of both ‘‘maturing out’’ and abstinence-oriented behaviour might characterize this disease, as well. Opiate detoxification can increase mortality in the months following treatment. This has been published for patients [11] and for released prisoners who were formerly opiate addicts [6]. The acute increase in mortality in detoxified subjects has been attributed to a loss of tolerance and consequent unpredictability of resumed heroin use [11,6,29]. Generally, the considerable
A. Naderi-Heiden et al. / European Psychiatry 27 (2012) 294–300
mortality risk in the months following treatment indicates the need for greater health education of drug users and implementation of relapse and overdose death prevention programmes. We did not observe an accumulation of deaths during the first year as compared to the following years. Anyhow, three subjects died in the first four months after discharge, but this concentration was also insignificant. New devices such as naltrexone implants do much to reduce the high risk of mortality in the initial period after detoxification [30,24]. Causes of death were not obtained by us. Deaths by AIDS cannot be very relevant in our sample since only 0.5% of the patients were HIV sero-positive while detoxified. As a matter of clinical routine, we do not offer detoxification treatment to HIV patients since detoxification may increase the risk of death in HIV-infected drug users. Individuals experiencing withdrawal symptoms have a fivefold increased risk of death [22]. From 290 males a number of 30 patients died in the observation period, from 114 females a number of five patients died. Nevertheless, SMR showed no statistically significant gender differences (males 10.6, females 11.1). The lack of differences could be partially explained by the females’ lower mortality rates in the cohort of patients and in the general population as well. Bargagli et al. stated no gender SMR differences in six of eight cohort studies but extremely high female SMRs in two cities (Barcelona 53.7, Rome 37.7) [2]. Gossop et al. found a higher SMR in males [17]. In our study, the hazard ratio was 0.41 for females, i.e. the hazard of death was 60% lower in females than in males (p = 0.07). Altogether, our mortality data indicated a trend towards a better outcome in females. The SMRs, the Kaplan-Meier estimate as well as the hazard ratios showed no statistically significant differences either between completers and non-completers or between pure opioid addicts and poly-substance addicts. This was in contrast to our hypotheses. The lack of affirmative findings in sufficiently detoxified patients might be due to the severity of illness in our sample, the low adherence, relapse and the potentially missed or failed aftercare in part of the patients. Davoli et al. emphasised that retention in any treatment has been protective against overdose mortality and published SMRs of 3.9 ‘‘during treatment’’ and 21.4 ‘‘out of treatment’’ [11]. A 20-year study in patients who were hospitalized for either self-poisoning or voluntary detoxification or both showed no significant differences in mortality with respect to the reason of admission [7]. Poly-drug use, benzodiazepines, amphetamines and heavy drinking were identified as risk factors for mortality [17]. This was not significantly shown by our data, where the diagnosis ‘‘poly-substance dependence’’ was made at baseline at the inpatient ward before starting an eleven years observation period. Unfortunately, in retrospect we are not aware of consumption patterns or comorbidities after discharge. However, research showed that even drug abusers who achieve abstinence suffer from an increased mortality rate compared with the general population [23]. Psychiatric status at 5-year follow-up was predictive of 15-year mortality, whereas abstinence was not [14]. 5. Conclusion The present study provides a longitudinal analysis including daily information during the period under review. The elevated mortality rate in comparison to the general population is consistent with other literature. During the first 5 years after discharge, SMRs were between 13.5 and 17.9, thereafter they clearly decreased with time. We are not aware of any other paper demonstrating conclusively the drop in mortality with time elapsed since detoxification. The low median employment rate of 9.1% was unfortunate. Our assumption that completers of opiate
299
detoxification treatment have lower mortality and higher employment rates than non-completers has not been confirmed. Pure opioid addicts had better employment rates than poly-substance addicts. Tolerant eligibility criteria for admission to our detoxification unit and unknown–potentially insufficient–aftercare in a number of the patients have been discussed. We do not know how the results would have turned out if only compliant patients with stable living conditions as well as adequate familial, social and therapeutic support had reached the abstinence-oriented intervention at the baseline of our study. Conflict of interest statement None. Acknowledgments We thank the Main Association of the Austrian Social Security ¨ sterreichischen SozialversicherInstitutions (Hauptverband der O ungstra¨ger) for supplying the data about daily documented employment records and dates of death when indicated. References [1] Backmund M, Meyer K, Henkel C, Soyka M, Reimer J, Schutz CG. Co-consumption of benzodiazepines in heroin users, methadone-substituted and codeinesubstituted patients. J Addict Dis 2005;24:17–29. [2] Bargagli AM, Hickman M, Davoli M, Perucci CA, Schifano P, Buster M, et al. Drug-related mortality and its impact on adult mortality in eight European countries. Eur J Public Health 2006;16:198–202. [3] Bargagli AM, Sperati A, Davoli M, Forastiere F, Perucci CA. Mortality among problem drug users in Rome: an 18-year follow-up study, 1980–97. Addiction 2001;96:1455–63. [4] Bauer SM, Loipl R, Jagsch R, Gruber D, Risser D, Thau K, et al. Mortality in opioid–maintened patients after release from an addiction clinic. Eur Addict Res 2008;14:82–91. [5] Becker JB, Hu M. Sex differences in drug abuse. Front Neuroendocrinol 2008;29:36–47. [6] Bird SM, Hutchinson SJ. Male drugs-related deaths in the fortnight after release from prison: Scotland, 1996–99. Addiction 2003;98:185–90. [7] Bjornaas MA, Bekken AS, Ojlert A, Haldorsen T, Jacobsen D, Rostrup M, et al. A 20-year prospective study of mortality and causes of death among hospitalized opioid addicts in Oslo. BMC Psychiatry 2008;8:8. [8] Breslow NE, Day NE. Statistical methods in cancer research volume II–the design and analysis of cohort studies. IARC Sci Publ 1987;1–406. [9] Callaghan RC, Cunningham JA. Gender differences in detoxification: predictors of completion and re-admission. J Subst Abuse Treat 2002;23:399–407. [10] Darke S, Kaye S, Duflou J. Systemic disease among cases of fatal opioid toxicity. Addiction 2006;101:1299–305. [11] Davoli M, Bargagli AM, Perucci CA, Schifano P, Belleudi V, Hickman M, et al. Risk of fatal overdose during and after specialist drug treatment: the VEdeTTE study, a national multi-site prospective cohort study. Addiction 2007;102: 1954–9. [12] Degenhardt L, Randall D, Hall W, Law M, Butler T, Burns L. Mortality among clients of a state-wide opioid pharmacotherapy program over 20 years: risk factors and lives saved. Drug Alcohol Depend 2009;105:9–15. [13] Ederer F, Axtell LM, Cutler SJ. The relative survival rate: a statistical methodology. Natl Cancer Inst Monogr 1961;6:101–21. [14] Fridell M, Hesse M. Psychiatric severity and mortality in substance abusers: a 15-year follow-up of drug users. Addict Behav 2006;31:559–65. ¨ sterreich GmbH Gescha¨ftsbereich O ¨ BIG. Bericht zur Drogensti[15] Gesundheit O ¨ sterreich 2009. tuation in O [16] Ghodse AH, Reynolds M, Baldacchino AM, Dunmore E, Byrne S, Oyefeso A, et al. Treating an opiate-dependent inpatient population: a one-year follow-up study of treatment completers and noncompleters. Addict Behav 2002;27:765–78. [17] Gossop M, Stewart D, Treacy S, Marsden J. A prospective study of mortality among drug misusers during a 4-year period after seeking treatment. Addiction 2002;97:39–47. [18] Hubbard RL, Craddock SG, Anderson J. Overview of 5-year followup outcomes in the drug abuse treatment outcome studies (DATOS). J Subst Abuse Treat 2003;25:125–34. [19] Kornor H, Waal H. From opioid maintenance to abstinence: a literature review. Drug Alcohol Rev 2005;24:267–74. [20] Madlung-Kratzer E, Spitzer B, Brosch R, Dunkel D, Haring C. A double-blind, randomized, parallel group study to compare the efficacy, safety and tolerability of slow-release oral morphine versus methadone in opioid-dependent in-patients willing to undergo detoxification. Addiction 2009;104:1549–57. [21] Mattick RP, Breen C, Kimber J, Davoli M. Methadone maintenance therapy versus no opioid replacement therapy for opioid dependence. Cochrane Database Syst Rev 2009;(3):CD002209.
300
A. Naderi-Heiden et al. / European Psychiatry 27 (2012) 294–300
[22] Michel L, Giorgi R, Villes V, Poizot-Martin I, Dellamonica P, Spire B, et al. Withdrawal symptoms as a predictor of mortality in patients HIV-infected through drug use and receiving highly active antiretroviral therapy (HAART). Drug Alcohol Depend 2009;99:96–104. [23] Price RK, Risk NK, Murray KS, Virgo KS, Spitznagel EL. Twenty-five year mortality of US servicemen deployed in Vietnam: predictive utility of early drug use. Drug Alcohol Depend 2001;64:309–18. [24] Reece AC. Favorable mortality profile of naltrexone implants for opiate addiction. J Addict Dis 2010;29:30–50. [25] Risser, Ho¨nigschnabl S, Stichenwirth M, Pfudl S, Sebald D, Kaff A, et al. Mortality of opiate users in Vienna Austria. Drug and alcohol dependence 2001;64:251–6. [26] Rosen D, Smith ML, Reynolds CF. The prevalence of mental and physical health disorders among older methadone patients. Am J Geriatr Psychiatry 2008;16: 488–97.
[27] Sees KL, Delucchi KL, Masson C, Rosen A, Clark HW, Robillard H, et al. Methadone maintenance vs 180-day psychosocially enriched detoxification for treatment of opioid dependence: a randomized controlled trial. JAMA 2000;283:1303–10. [28] Statistik Austria. Demographisches Jahrbuch 2005 Vienna 2006. [29] Strang J, McCambridge J, Best D, Beswick T, Bearn J, Rees S, et al. Loss of tolerance and overdose mortality after inpatient opiate detoxification: follow up study. BMJ 2003;326:959–60. [30] Tait RJ, Ngo HTT, Hulse GK. Mortality in heroin users 3 years after naltrexone implant or methadone maintenance treatment. J Subst Abuse Treat 2008;35: 116–24. [31] Termorshuizen F, Krol A, Prins M, van Ameijden EJ. Long-term outcome of chronic drug use: the Amsterdam Cohort Study among drug users. Am J Epidemiol 2005;161:271–9. [32] Van den Brink W, Haasen C. Evidenced-based treatment of opioid-dependent patients. Can J Psychiatry 2006;51:635–46.
Original article
Mortality and employment after in-patient opiate detoxification A. Naderi-Heiden a,*, A. Gleiss b, C. Ba¨cker a, D. Bieber a, H. Nassan-Agha a, S. Kasper a, R. Frey a a b
Division of Biological Psychiatry, Department of Psychiatry and Psychotherapy, Medical University of Vienna, Waehringer Guertel 18-20, 1090 Vienna, Austria Core Unit for Medical Statistics and Informatics, Section of Clinical Biometrics, Medical University of Vienna, Spitalgasse 23, 1090 Vienna, Austria
A R T I C L E I N F O
A B S T R A C T
Article history: Received 29 January 2010 Received in revised form 6 May 2010 Accepted 8 May 2010 Available online 22 July 2010
Aim. – We considered that completed opiate detoxification resulted in increased life expectancy and earning capacity as compared to non-completed detoxification. Methods. – The cohort study sample included pure opioid or poly-substance addicts admitted for voluntary in-patient detoxification between 1997 and 2004. Of 404 patients, 58.7% completed the detoxification program and 41.3% did not. The Austrian Social Security Institution supplied data on survival and employment records for every single day in the individual observation period between discharge and December 2007. Statistical analyses included the calculation of standardized mortality rates for the follow-up period of up to 11 years. Results. – The standardized mortality ratios (SMRs) were between 13.5 and 17.9 during the first five years after discharge, thereafter they fell clearly with time. Mortality did not differ statistically significantly between completers and non-completers. The median employment rate was insignificantly higher in completers (12.0%) than in non-completers (5.5%). The odds for being employed were higher in pure opioid addicts than in poly-substance addicts (p = 0.003). Conclusions. – The assumption that completers of detoxification treatment have a better outcome than non-completers has not been confirmed. The decrease in mortality with time elapsed since detoxification is interesting. Pure opioid addicts had better employment prospects than poly-substance addicts. ß 2010 Elsevier Masson SAS. All rights reserved.
Keywords: Mortality Employment Opiate detoxification Opioid dependence Poly-substance dependence
1. Introduction National data sources have estimated the number of opiate abusers in Austria to be about 33,000 [15]. It was estimated that at least 27% of drug users had died within 20 years in a cohort of chronic drug users [31]. Detoxification is a necessary step prior to abstinence-oriented therapy. Generally, voluntary participation in detoxification programs with eligibility criteria is a positive prognostic feature in abstinence-oriented therapy [19]. The efficiency of opioid detoxification is critically discussed with respect to relapse and abstinence rates, respectively. Diversity in study design limited the comparability of studies. Results can be biased by patients’ inclusion criteria (e.g. co-morbidity), mode of detoxification and aftercare as well as abstinence definitions. Abstinence rates after step-by-step detoxification of methadone vary between 22 and 86% [19]. Both clinical experience and neurobiological evidence indicate that opioid dependence is a chronic relapsing disorder with potential social neglect and high mortality risk [32]. In the final
* Corresponding author. Tel.: +43 1 40 400 3568; fax: +43 1 40 400 3099. E-mail address: [email protected] (A. Naderi-Heiden). 0924-9338/$ – see front matter ß 2010 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.eurpsy.2010.05.002
analysis, the aim of detoxification and abstinence-oriented treatment is increased earning capacity and increased life expectancy. Our cohort study was conducted to evaluate the implication of detoxification treatment with respect to mortality and occupation in a follow-up period of up to 11 years. In-patients suffering from pure or poly-substance dependence were treated with slow release oral morphinsulphate/-hydrochloride, methadone or buprenorphine in a gradual detoxification program with strict guidelines that took about 2–4 weeks. We compared patients who completed in-patient opioid detoxification with those who failed to complete this program. Theoretically, detoxified patients with future prospects of opiate abstinence should have a better prognosis. The analyses considered a division into pure opioid addicts and poly-substance dependent patients. 2. Subjects and methods 2.1. Study population, in-patient therapeutic detoxification programme 1997–2004 The study sample included all opioid addicts (aged 15 or older) with clear commitment to abstinence, admitted at the Department of Biological Psychiatry, Medical University of Vienna, Austria, in the period between January 1997 and December 2004. Only
A. Naderi-Heiden et al. / European Psychiatry 27 (2012) 294–300
patients for voluntary in-patient detoxification were included. Patients were excluded from the present analyses if the following conditions occurred: [1] emergency treatments; [2] partial inpatient detoxification from illicit drugs, benzodiazepines or alcohol but continuously prescribed opioid maintenance therapy and; [3] in case of readmission to the withdrawal program. Note, that only the first entry (index-episode) was computed. In 1997, immediately before admission 73% of the subjects were treated by oral methadone or oral morphine maintenance therapy. Owing to revised treatment guidelines, the proportion of patients with former methadone, morphine or buprenorphine maintenance therapy increased year by year and was finally higher than 90% from year 2000 on. However, most of the patients had been noncompliant with maintenance programs and poly-substance dependent, respectively. We discriminated between patients with pure opioid dependence (ICD-10: F11.2) and poly-substance dependence (ICD-10: F19.2, including opioid, benzodiazepine, alcohol, cannabis, cocaine and amphetamine). The diagnosis of substance dependence was based on data derived from semi-structured diagnostic interviews and from existent medical records. A human immunodeficiency virus (HIV) test was performed by routine. Patients with HIV were detoxified in extraordinary cases only (n = 2). Detoxification treatment, usually 2–4 weeks in duration, was proposed. The patients’ opioid dose was replaced by an equivalent dosage of slow-release oral morphinsulphate/-hydrochloride, methadone or buprenorphine. After tapering off, the success of detoxification was confirmed by opiate, methadone, cocain, amphetamine, cannabinoide, barbiturate and benzodiazepine free urine specimen. The total sample of 404 patients (mean age 27.4 7.0 y) comprised 290 males (71.8%) and 114 females (28.2%). Twenty-five percent of the patients were recorded as employed (males 27%, females 20%). There were 125 patients with opioid dependence (30.9%) and 279 patients with poly-substance dependence (69.1%). Part of the patients completed the detoxification treatment (with drug-free urine specimen) and was discharged with abstinence-oriented recommendations. Another part discontinued the therapy and further illicid opioid intake or opiode maintenance therapy could be expected. To form the two comparison groups, patients were divided in completers and non-completers. From 404 patients, 237 (58.7%) completed the detoxification program (mean age: 27.7 6.3 y) and 167 (41.3%) did not (mean age: 27.2 7.5 y). The in-patient treatment of completers (median: 17 days) lasted longer than the in-patient treatment of noncompleters (median: 10 days). Completion rate was 63.2% in patients with pure opioid dependence and 56.6% in patients with polysubstance dependence (p = 0.215). Completion rate was 56.6% in male patients and 64.0% in females (p = 0.169). In the period between the index-episode and December 2004, a proportion of 59 out of 404 patients (14.6%) were readmitted at least one more time. The readmissions (n = 68) did not count for our analyses. 2.2. Study design and data sources This is a retrospective cohort study with arbitrarily chosen administrative censoring on Dec 31st, 2007. Data were collected in 2008. The Main Association of Austrian Social Security Institutions informed about daily documented employment status and data of death when indicated. Data on mortality was gained for the total sample of 404 subjects. The cause of death is not documented by the Main Association of Austrian Social Security Institutions. Data on employment was gained for 399 subjects only, since the identification numbers of five patients were not correctly delivered. The observation period lasted between a minimum of 3 and a maximum
295
of 11 years, i.e. from the day of patients’ discharge (between 01/1997 and 12/2004) to Dec 31st, 2007. At each single day in the observation period, patients were rated ‘‘employed’’ or ‘‘unemployed’’. In case of ‘‘no social security’’, patients were judged as ‘‘unemployed’’. The employment rate (%) was calculated for each subject. This is the duration of employment (specification in days) divided by the duration of the observation period (specification in days). 2.3. Confidentiality of data The protocol was approved by the local Ethics and Scientific Review Committee (EK.Nr: 489/2007). Personal identifiers were used following the rules of privacy regulation. For statistical evaluation only indirect individual-related data was used (§ 4 Z 1, Data Protection Act). The investigation set out to answer the following questions: have completers a lower mortality risk after discharge than noncompleters? have pure opioid dependents a lower mortality risk after discharge than poly-substance dependents? have completers more working days between discharge and the endpoint of the observation period than non-completers? have pure opioid dependents more working days between discharge and the endpoint of the observation period than polysubstance dependents? Gender differences and the influence of age at time of admission will also be explored with respect to mortality risk and working days. 2.4. Statistics Regarding to diagnostic features and gender, proportions of completers and non-completers were compared by Chi2 tests. In order to compare the observed number of deaths with what would be expected in the general population, the following approach was used [13]: the expected number of deaths within each year after discharge was calculated under the assumption that each patient’s risk would be equal to the risk in an age and sex matched segment of the general population. This risk was based on the life table provided by the Statistik Austria for the year 2002 [28]. The standardized mortality ratio (SMR) was the ratio of observed deaths to expected deaths. Ninety-five percent confidence intervals (CI) for SMRs as well as pvalues for comparing SMRs between two groups were calculated as proposed by Breslow and Day [8]. Observed and expected numbers of deaths were compared with each other using a Chi2 test. The calculation of percentages (and 95% CI) of patients reaching a certain time after discharge was based on the corresponding Kaplan-Meier survival distribution estimates. Kaplan-Meier curves are compared between groups using log-rank tests. The potential influence of various explanatory variables on survival after discharge was investigated using single- and multi-variable Cox proportional hazards models. Hazard ratios and 95% CI were given only for the multi-variable model since single-variable results were virtually the same. Both methods, Kaplan-Meier estimates and Cox models, are able to appropriately deal with the form of administrative censoring given in our data set. Employment rates were calculated from the daily raw data provided by the Main Association of Austrian Social Security Institutions as the number of days at which a patient was registered as being employed, divided by the total number of days within the observation period. This total number of days includes days where no information was available, thus producing worst case estimates. Employment rates are presented as median, quartiles, minimum and maximum due to their right skewed
A. Naderi-Heiden et al. / European Psychiatry 27 (2012) 294–300
296
Table 1 Number of subjects, person-years of follow-up, number of deaths and crude death rates per 1000 person-years of follow-up for the total sample and different subgroups.
Total sample Completers Non-completers Pure opiod dependence Poly-substance dependence Males Females a
Number of subjects
Person-years of follow-up
Number of deaths
a
404 237 167 125 279 290 114
2997.02 1714.91 1282.11 1303.36 1966.66 2090.57 906.45
35 21 14 8 27 30 5
11.68 12.25 10.92 6.14 13.73 14.35 5.52
Death rates per 1000 person-years of follow-up
Number of deaths/person-years of follow-up 1000.
distribution. The potential influence of various explanatory variables on the employment rate was investigated using a singleand multi-variable, proportional odds, logistic regression model where the dependent variable, employment rate, is categorized as follows: [category 1] employment rate of 0.0%, [category 2] employment rate of greater than 0.0 up to 17.8%, [category 3] employment rate greater than 17.8% (see results section: the median employment rate in patients with at least one working day turned out to be 17.8). Odds ratios with 95% CI are given only for the multi-variable model since single-variable results are virtually the same. No two-way interactions were significant. The reported p-values are the results of two-sided tests. pvalues less or equal to 0.05 were considered to be statistically significant. All computations have been performed using SAS software Version 9.2 (SAS Institute Inc., Cary, NC, USA, 2008). 3. Results
in males and 11.11 (CI 3.58–25.93) in females (p = 0.888), respectively, 10.77 (CI 5.50–16.90) in completers and 10.07 (CI 6.66–16.46) in non-completers (p = 0.982). The SMR was 7.08 (CI 3.05–13.95) in pure opioid addicts and 12.27 (CI 8.09–17.86) in poly-substance addicts (p = 0.228). Three subjects died within the first four months. The SMR for this short period was 15.8 (three deaths observed, 0.19 deaths expected), which did not differ significantly from the SMR 8.8 in the rest of the observation period (p = 0.560). The annual life table can be seen in Table 2. Relating to the whole observation period, the number of deaths in the patient group (Nd) was significantly higher than the number of expected deaths (Ne), i.e. the mortality risk was significantly higher in our patients than in the general population (p < 0.001, see Table 2). During the first years after discharge, SMRs were between 13.5 and 17.9, thereafter they fell with time. The Kaplan-Meier estimate of the survival distribution after discharge showed the following results for the total group:
3.1. Mortality Thirty-five subjects (30 males, five females) died in the observation period, 21 of which finished past detoxification treatments as completers, while 14 were rated as non-completers. Eight deceased subjects were diagnosed as pure opioid dependent, 27 as poly-substance dependent. Death rates per 1000 personyears of follow-up for the total sample and different subgroups can be seen in Table 1. With respect to deceased patients, a proportion of 27/35 (77%) had been treated by methadone, morphine or buprenorphine maintenance programs prior to the detoxification treatment, 8/35 (23%) had not. We have no information about the follow-up therapy and the cause of death. Between discharge and the year 2007, the SMR was 10.45 (CI 7.28–14.53) in the total group. The SMR was 10.56 (CI 7.13–15.08)
98.8% 95.5% 92.3% 90.4%
(CI (CI (CI (CI
97.1–99.5) 93.0–97.2) 89.2–94.6) 86.8–93.1)
survived survived survived survived
the the the the
1st year; 3rd year; 5th year; 10th year.
Fig. 1 shows the Kaplan Meier plot for the subgroups completers and non-completers, respectively. Inter-group differences were not significant (p = 0.806). Fig. 2 shows the Kaplan Meier plot for the subgroups pure opioid dependence and poly-substance dependence, respectively. Inter-group differences were not significant (p = 0.220). Four explanatory variables were investigated with respect to possible effects on the hazard of death in the observation period: discontinuation of detoxification treatment, age at the beginning of
Table 2 Life table of 404 opioid-addicted patients in an observation period up to 11 years after in-patient detoxification treatment. Time perioda (year)
Subjects starting annual observation period (N)b
Observed number of deaths (Nd)
Expected number of deaths (Ne)c
Standardized mortality ratio (Nd:Ne)
1st 2nd 3rd 4th 5th 6th 7th 8th 9th 10th 11th
404 399 393 386 349 317 286 250 198 132 78
5 6 7 6 6 2 1 1 1 0 0
0.37 0.38 0.39 0.39 0.37 0.35 0.33 0.30 0.24 0.16 0.07
13.54 15.77 17.90 15.55 16.28 5.69 3.01 3.36 4.23 0.00 0.00
a
After individual discharge from detoxification unit. The number of patients starting observation period decreased annually and depends on the number of patients lost by death or lost by administrative censoring on December 31st 2007. c The column shows the expected number of deceases in the annual patients’ sample under the assumption that each patient’s risk would be equal to the risk in the general population. The latter is based on the life table calculated by the Statistik Austria and comprised an age and sex matched general population of the year 2002 (21). Observed deaths versus expected deaths in the observation period (Nd vs. Ne): p < 0.0001. b
A. Naderi-Heiden et al. / European Psychiatry 27 (2012) 294–300
297
Table 3 Median employment rate (%) as well as 1st and 3rd quartile (%) in the observation period in different subgroups (total sample size = 399). Subgroups
Detoxification Completers Non-completers Dependence Pure opioid Poly-substance Gender Males Females
Fig. 1. Survival over time since discharge from psychiatry unit is shown for a cohort detoxified between 1997 and 2004. The graph shows Kaplan-Meier product-limit estimate, including deaths of completers and non-completers. Inter-group differences were not significant (p = 0.806). Note that the vertical axis scale starts at a minimal survival probability of 0.85.
the detoxification treatment, poly-substance dependence and gender. None of these showed a significant influence on survival: hazard ratios from the multi-variable Cox model were 0.85 (CI 0.43–1.66) for non-completers (p = 0.614), 1.60 (CI 0.72–3.53) for poly-substance addicts (p = 0.247), 1.15 (CI 0.73–1.82) for each 10 year increase in age (p = 0.549), and 0.41 (CI 0.16–1.07) for females (p = 0.070). 3.2. Employment The total sample (n = 399) showed a median employment rate of 9.1% (min: 0.0%, 1st quartile: 0.2%, 3rd quartile: 32.6%, max. 99.4%) in the observation period. The median employment rate of different subgroups can be seen in Table 3. The results were in favour of completers, pure opioid addicts and males.
1st quartile
Median employment rate
3rd quartile
233 166
0.3 0.1
12.0 5.5
39.1 27.7
122 277
1.8 0.0
16.3 6.1
49.4 25.1
285 114
0.2 0.0
11.3 5.7
39.1 22.1
Table 4 The total sample is subdivided into different subgroups (concerning detoxification, diagnosis and gender) and assigned to three employment rate categories. In case of at least one regular working day, the subjects showed a median employment rate of 17.8 %. Subgroups
Detoxification Completers Non-completers Dependence Pure opioid Poly-substance Gender Males Females
Fig. 2. Survival over time after discharge from psychiatry unit is shown for a cohort detoxified between 1997 and 2004. The graph shows Kaplan-Meier product-limit estimate, including deaths of subjects suffering from pure opioid dependence and poly-substance dependence. Inter-group differences were not significant (p = 0.220). Note that the vertical axis scale starts at a minimal survival probability of 0.85.
N
Employment rate (%) 0.0 % category 1
0.0–17.8 % category 2
> 17.8 % category 3
22.3 24.7
36.1 41.6
41.6 33.7
16.4 26.3
34.4 40.1
49.2 33.6
22.1 26.3
36.1 43.9
41.8 29.8
In the total group, 93 subjects (23.3%) did not work one single day (employment rate: 0.0%) in the observation period. The rest of the subjects (n = 306) showed a median employment rate of 17.8%, i.e. 153 subjects (38.3%) were employed less than 17.8% of the days in the observation period, 153 (38.3%) more than 17.8%. Table 4 introduces three categories: [category 1] employment rate of 0.0%, [category 2] employment rate of 0.0–17.8%, [category 3] employment rate greater than 17.8%. The completers, pure opioid addicts and males make for higher percentages of working subjects in [category 3]. The multi-variable logistic regression analysis investigated (dis) continuation of detoxification treatment, age at the beginning of detoxification treatment, pure opioid or poly-substance dependence and gender. The odds for being in a lower category (fewer days of employment) were 30.1% higher in non-completers than in completers (OR 1.301, CI 0.895–1.892, p = 0.168); 86.9% higher in poly-substance addicts than in pure opioid addicts (OR 1.869, CI 1.243–2.809, p = 0.003); 69.4% higher in females than in males (OR 1.694, CI 1.121–2.559, p = 0.012); 39.8% higher for each 10 additional years of age (OR 1.398, CI 1.071–1.827, p = 0.014). 4. Discussion One main question could be confirmed by a statistically significant finding: pure opioid dependent patients had a higher employment rate in the years after discharge than poly-substance dependent patients (logistic regression analysis, p = 0.003). The median employment rate was 16.3% in pure opioid dependence and 6.1% in poly-substance dependence. The diagnostic classification was documented at the time of admission at the detoxification unit. The aftercare of our patients has varied or was unknown during an observation period of up to 11 years. We are not aware of
298
A. Naderi-Heiden et al. / European Psychiatry 27 (2012) 294–300
consumption patterns after discharge. Maybe co-consumption patterns of our patients did not change considerably over years and pure opiate addicts had better working abilities. Co-consumption of benzodiazepines may have substantial impact on morbidity, mortality and clinical course [1]. A higher abstinence rate and thus a higher employment rate in pure opiate addicts could be another reason for our finding [9]. Additional analysis showed that male gender (p = 0.012) and younger age (p = 0.014) were significantly associated with high employment rates. For both conditions, this can be generally explained by the better access to the labour force. It can be assumed that females are more involved in childcare. There was no indication for more chronicity or damage in our female sample, but a review article states that females are at greater risk for relapse following abstinence [5]. Young subjects might have better working abilities with respect to shorter duration or less severity of illness. In a group of 11,400 problem drug users attending 38 different treatment centres in Rome, 61% of the study population were recorded as employed. Males were more likely to be employed than females (64% versus 48%) [3]. In our study, the proportion of employed subjects was only 25% while admitted to in-patient treatment. Indeed, tolerant eligibility criteria for admission to our detoxification unit included social exclusion, co-morbidity and multiple drug use. Opiate detoxification was offered not only to adherent patients but also to patients non-compliant with maintenance programs, which is a formal stance opposite to that in many countries. Other authors stated that abstinence-oriented interventions are indicated only in a few motivated patients with stable living conditions and adequate familial, social and therapeutic support [32]. In practice, non-compliant addicts or those with otherwise displaying intolerable behaviour are in fact subject to involuntary detoxification or prison. Fifty-nine percent of the patients completed the detoxification program. Other Austrian authors also found completion rates of about 50% with either methadone or slow-release oral morphine tapering [20]. We postulated that successfully detoxified patients should have better working abilities than non-completers, but completion of detoxification treatment did not significantly increase the probability of high employment rates during the observation period. The low employment rate might be an indication of high relapse rates in completers. The broad public health system in Austria allocates not only abstinence-oriented therapy but also low-threshold maintenance therapy with methadone, morphine tablets or buprenorphine. This could have been helpful anyhow. The low employment rate might also be influenced by not reported illegal work and the tolerant Austrian social security system for long-term non-workers, which can be a matter of survival or convenience as well. Few studies have assessed employment functioning after detoxification: [1] Sees et al. compared outcomes of patients with opioid dependence treated with methadone maintenance versus psychosocially enriched methadone assisted detoxification and showed no difference between groups in employment functioning after 12 months [27]. [2] Hubbard et al. showed a 10% increase in full-time employment among drug-free outpatients after 5 years [18]. We are not aware of other follow-up studies comparing working days in completers and non-completers of detoxification treatments. Significantly better abstinence-oriented behaviour was detected amongst patients who completed detoxification and went on to spend at least 6 weeks in a rehabilitation unit [16]. In contrast, the authors observed no significant different abstinence rates between non-completers and completers who had no aftercare on the majority of measures of drug use during follow-up. Unfortunately, earning capacity was not included as an outcome criterion. Mortality rates among opioid addicts are higher than for the general population. Bargagli et al. stated mortality rates in eight
study sites in Europe with comparable methodology [2]. All cohorts consisted of opiate users entering treatment during the period 1990–2000. The authors did not specify the kind of treatment. SMRs were calculated for the total observation period and vary between 9.2 and 25.6 in different countries. In the area of Vienna, Bargagli et al. reported an SMR of 9.8 without gender differences [2]. Numerically, 195 from 4150 patients died. The cause of death varied as follows: 50% drug-related, 19% HIVrelated, other causes 31%. In our Viennese study, the sample showed nearly the same SMR of 10.5 during 11 years. The mortality risk of the detoxified patients was significantly higher than in the general population (p < 0.0001). Other Austrian authors stated a SMR of 17.8 over a 3-year study period in a cohort of opiate-users enrolled in opioid maintenance treatment from 1995–1997 [25] and a SMR of 29.13 in cohort of patients enrolled in maintenance therapy from 1998–1999 [4]. In the latter sample, many HIV seropositve patients were treated and the cause of death was in 35% related to AIDS. Methadone is an effective maintenance treatment of opioid dependence. About 90% of our patients received opioid replacement therapy before the index episode. We assume that noncompleters of detoxification program or relapsed completers might have chosen this treatment again. Degenhardt et al. revealed in a follow up study of about 42,000 persons receiving opioid pharmacotherapy that time in treatment (methadone or buprenorphine as opioid replacement therapy) was associated with lower mortality than time out of treatment, with an overall ‘‘in treatment SMR’’ of 4.5, compared to an ‘‘out of treatment’’ SMR of 8.0 [12]. When compared methadone maintenance therapy and treatments that did not involve opioid replacement therapy (i.e., detoxification), a systematic review did not show a statistically significant superior effect on criminal activity and mortality [21]. Long-term dependent users and patients on long term opiate agonist treatments respectively might have many risk factors for poor cardiac, pulmonary, hepatic and renal health. Darke et al. confirmed that systemic disease was prominent among cases of fatal opioid toxicity cases with increasing age [10]. The prevalence of physical health disorders among older methadone patients is very high [26]. This severe illness profile in its own right forms an important reason to opt for detoxification treatment. This treatment justification needs to be carefully considered, if one is aware that patients choosing this route face an inordinate initial mortality risk. Our sample revealed consecutively high SMRs (from 13.5 to 17.9, Table 2) in a five-year period after detoxification, but relatively low SMRs thereafter. Note, however, that due to the small numbers of observed and expected deaths the variability of the reported SMRs is quite high (expressed by broad CI). The decrease in SMR may partly be due to increased mortality in the general population over time, leading to a relative reduction in the ratio. It is suggested that those who take the highest risks will probably die earliest, leading to a decrease in mortality for the cohort as a whole. However, the final drop in SMR after detoxification is greater than that, which occurs in agonist maintained patients where SMR curves are rather flat during the observation period [12]. Even though opioid dependency is a complex and chronically relapsing disease with high risks of morbidity and mortality a long process of both ‘‘maturing out’’ and abstinence-oriented behaviour might characterize this disease, as well. Opiate detoxification can increase mortality in the months following treatment. This has been published for patients [11] and for released prisoners who were formerly opiate addicts [6]. The acute increase in mortality in detoxified subjects has been attributed to a loss of tolerance and consequent unpredictability of resumed heroin use [11,6,29]. Generally, the considerable
A. Naderi-Heiden et al. / European Psychiatry 27 (2012) 294–300
mortality risk in the months following treatment indicates the need for greater health education of drug users and implementation of relapse and overdose death prevention programmes. We did not observe an accumulation of deaths during the first year as compared to the following years. Anyhow, three subjects died in the first four months after discharge, but this concentration was also insignificant. New devices such as naltrexone implants do much to reduce the high risk of mortality in the initial period after detoxification [30,24]. Causes of death were not obtained by us. Deaths by AIDS cannot be very relevant in our sample since only 0.5% of the patients were HIV sero-positive while detoxified. As a matter of clinical routine, we do not offer detoxification treatment to HIV patients since detoxification may increase the risk of death in HIV-infected drug users. Individuals experiencing withdrawal symptoms have a fivefold increased risk of death [22]. From 290 males a number of 30 patients died in the observation period, from 114 females a number of five patients died. Nevertheless, SMR showed no statistically significant gender differences (males 10.6, females 11.1). The lack of differences could be partially explained by the females’ lower mortality rates in the cohort of patients and in the general population as well. Bargagli et al. stated no gender SMR differences in six of eight cohort studies but extremely high female SMRs in two cities (Barcelona 53.7, Rome 37.7) [2]. Gossop et al. found a higher SMR in males [17]. In our study, the hazard ratio was 0.41 for females, i.e. the hazard of death was 60% lower in females than in males (p = 0.07). Altogether, our mortality data indicated a trend towards a better outcome in females. The SMRs, the Kaplan-Meier estimate as well as the hazard ratios showed no statistically significant differences either between completers and non-completers or between pure opioid addicts and poly-substance addicts. This was in contrast to our hypotheses. The lack of affirmative findings in sufficiently detoxified patients might be due to the severity of illness in our sample, the low adherence, relapse and the potentially missed or failed aftercare in part of the patients. Davoli et al. emphasised that retention in any treatment has been protective against overdose mortality and published SMRs of 3.9 ‘‘during treatment’’ and 21.4 ‘‘out of treatment’’ [11]. A 20-year study in patients who were hospitalized for either self-poisoning or voluntary detoxification or both showed no significant differences in mortality with respect to the reason of admission [7]. Poly-drug use, benzodiazepines, amphetamines and heavy drinking were identified as risk factors for mortality [17]. This was not significantly shown by our data, where the diagnosis ‘‘poly-substance dependence’’ was made at baseline at the inpatient ward before starting an eleven years observation period. Unfortunately, in retrospect we are not aware of consumption patterns or comorbidities after discharge. However, research showed that even drug abusers who achieve abstinence suffer from an increased mortality rate compared with the general population [23]. Psychiatric status at 5-year follow-up was predictive of 15-year mortality, whereas abstinence was not [14]. 5. Conclusion The present study provides a longitudinal analysis including daily information during the period under review. The elevated mortality rate in comparison to the general population is consistent with other literature. During the first 5 years after discharge, SMRs were between 13.5 and 17.9, thereafter they clearly decreased with time. We are not aware of any other paper demonstrating conclusively the drop in mortality with time elapsed since detoxification. The low median employment rate of 9.1% was unfortunate. Our assumption that completers of opiate
299
detoxification treatment have lower mortality and higher employment rates than non-completers has not been confirmed. Pure opioid addicts had better employment rates than poly-substance addicts. Tolerant eligibility criteria for admission to our detoxification unit and unknown–potentially insufficient–aftercare in a number of the patients have been discussed. We do not know how the results would have turned out if only compliant patients with stable living conditions as well as adequate familial, social and therapeutic support had reached the abstinence-oriented intervention at the baseline of our study. Conflict of interest statement None. Acknowledgments We thank the Main Association of the Austrian Social Security ¨ sterreichischen SozialversicherInstitutions (Hauptverband der O ungstra¨ger) for supplying the data about daily documented employment records and dates of death when indicated. References [1] Backmund M, Meyer K, Henkel C, Soyka M, Reimer J, Schutz CG. Co-consumption of benzodiazepines in heroin users, methadone-substituted and codeinesubstituted patients. J Addict Dis 2005;24:17–29. [2] Bargagli AM, Hickman M, Davoli M, Perucci CA, Schifano P, Buster M, et al. Drug-related mortality and its impact on adult mortality in eight European countries. Eur J Public Health 2006;16:198–202. [3] Bargagli AM, Sperati A, Davoli M, Forastiere F, Perucci CA. Mortality among problem drug users in Rome: an 18-year follow-up study, 1980–97. Addiction 2001;96:1455–63. [4] Bauer SM, Loipl R, Jagsch R, Gruber D, Risser D, Thau K, et al. Mortality in opioid–maintened patients after release from an addiction clinic. Eur Addict Res 2008;14:82–91. [5] Becker JB, Hu M. Sex differences in drug abuse. Front Neuroendocrinol 2008;29:36–47. [6] Bird SM, Hutchinson SJ. Male drugs-related deaths in the fortnight after release from prison: Scotland, 1996–99. Addiction 2003;98:185–90. [7] Bjornaas MA, Bekken AS, Ojlert A, Haldorsen T, Jacobsen D, Rostrup M, et al. A 20-year prospective study of mortality and causes of death among hospitalized opioid addicts in Oslo. BMC Psychiatry 2008;8:8. [8] Breslow NE, Day NE. Statistical methods in cancer research volume II–the design and analysis of cohort studies. IARC Sci Publ 1987;1–406. [9] Callaghan RC, Cunningham JA. Gender differences in detoxification: predictors of completion and re-admission. J Subst Abuse Treat 2002;23:399–407. [10] Darke S, Kaye S, Duflou J. Systemic disease among cases of fatal opioid toxicity. Addiction 2006;101:1299–305. [11] Davoli M, Bargagli AM, Perucci CA, Schifano P, Belleudi V, Hickman M, et al. Risk of fatal overdose during and after specialist drug treatment: the VEdeTTE study, a national multi-site prospective cohort study. Addiction 2007;102: 1954–9. [12] Degenhardt L, Randall D, Hall W, Law M, Butler T, Burns L. Mortality among clients of a state-wide opioid pharmacotherapy program over 20 years: risk factors and lives saved. Drug Alcohol Depend 2009;105:9–15. [13] Ederer F, Axtell LM, Cutler SJ. The relative survival rate: a statistical methodology. Natl Cancer Inst Monogr 1961;6:101–21. [14] Fridell M, Hesse M. Psychiatric severity and mortality in substance abusers: a 15-year follow-up of drug users. Addict Behav 2006;31:559–65. ¨ sterreich GmbH Gescha¨ftsbereich O ¨ BIG. Bericht zur Drogensti[15] Gesundheit O ¨ sterreich 2009. tuation in O [16] Ghodse AH, Reynolds M, Baldacchino AM, Dunmore E, Byrne S, Oyefeso A, et al. Treating an opiate-dependent inpatient population: a one-year follow-up study of treatment completers and noncompleters. Addict Behav 2002;27:765–78. [17] Gossop M, Stewart D, Treacy S, Marsden J. A prospective study of mortality among drug misusers during a 4-year period after seeking treatment. Addiction 2002;97:39–47. [18] Hubbard RL, Craddock SG, Anderson J. Overview of 5-year followup outcomes in the drug abuse treatment outcome studies (DATOS). J Subst Abuse Treat 2003;25:125–34. [19] Kornor H, Waal H. From opioid maintenance to abstinence: a literature review. Drug Alcohol Rev 2005;24:267–74. [20] Madlung-Kratzer E, Spitzer B, Brosch R, Dunkel D, Haring C. A double-blind, randomized, parallel group study to compare the efficacy, safety and tolerability of slow-release oral morphine versus methadone in opioid-dependent in-patients willing to undergo detoxification. Addiction 2009;104:1549–57. [21] Mattick RP, Breen C, Kimber J, Davoli M. Methadone maintenance therapy versus no opioid replacement therapy for opioid dependence. Cochrane Database Syst Rev 2009;(3):CD002209.
300
A. Naderi-Heiden et al. / European Psychiatry 27 (2012) 294–300
[22] Michel L, Giorgi R, Villes V, Poizot-Martin I, Dellamonica P, Spire B, et al. Withdrawal symptoms as a predictor of mortality in patients HIV-infected through drug use and receiving highly active antiretroviral therapy (HAART). Drug Alcohol Depend 2009;99:96–104. [23] Price RK, Risk NK, Murray KS, Virgo KS, Spitznagel EL. Twenty-five year mortality of US servicemen deployed in Vietnam: predictive utility of early drug use. Drug Alcohol Depend 2001;64:309–18. [24] Reece AC. Favorable mortality profile of naltrexone implants for opiate addiction. J Addict Dis 2010;29:30–50. [25] Risser, Ho¨nigschnabl S, Stichenwirth M, Pfudl S, Sebald D, Kaff A, et al. Mortality of opiate users in Vienna Austria. Drug and alcohol dependence 2001;64:251–6. [26] Rosen D, Smith ML, Reynolds CF. The prevalence of mental and physical health disorders among older methadone patients. Am J Geriatr Psychiatry 2008;16: 488–97.
[27] Sees KL, Delucchi KL, Masson C, Rosen A, Clark HW, Robillard H, et al. Methadone maintenance vs 180-day psychosocially enriched detoxification for treatment of opioid dependence: a randomized controlled trial. JAMA 2000;283:1303–10. [28] Statistik Austria. Demographisches Jahrbuch 2005 Vienna 2006. [29] Strang J, McCambridge J, Best D, Beswick T, Bearn J, Rees S, et al. Loss of tolerance and overdose mortality after inpatient opiate detoxification: follow up study. BMJ 2003;326:959–60. [30] Tait RJ, Ngo HTT, Hulse GK. Mortality in heroin users 3 years after naltrexone implant or methadone maintenance treatment. J Subst Abuse Treat 2008;35: 116–24. [31] Termorshuizen F, Krol A, Prins M, van Ameijden EJ. Long-term outcome of chronic drug use: the Amsterdam Cohort Study among drug users. Am J Epidemiol 2005;161:271–9. [32] Van den Brink W, Haasen C. Evidenced-based treatment of opioid-dependent patients. Can J Psychiatry 2006;51:635–46.