Nordic Journal of Psychiatry
ISSN: 0803-9488 (Print) 1502-4725 (Online) Journal homepage: http://www.tandfonline.com/loi/ipsc20
Total population-based study of the impact of substance use disorders on the overall survival of psychiatric inpatients Steinn Steingrímsson, Martin I. Sigurdsson, Thor Aspelund, Sigmundur Sigfússon & Andrés Magnússon To cite this article: Steinn Steingrímsson, Martin I. Sigurdsson, Thor Aspelund, Sigmundur Sigfússon & Andrés Magnússon (2016) Total population-based study of the impact of substance use disorders on the overall survival of psychiatric inpatients, Nordic Journal of Psychiatry, 70:3, 161-166, DOI: 10.3109/08039488.2015.1062143 To link to this article: http://dx.doi.org/10.3109/08039488.2015.1062143
Published online: 28 Aug 2015.
Submit your article to this journal
Article views: 35
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ipsc20 Download by: [Orta Dogu Teknik Universitesi]
Date: 05 March 2016, At: 08:00
Total population-based study of the impact of substance use disorders on the overall survival of psychiatric inpatients
Downloaded by [Orta Dogu Teknik Universitesi] at 08:00 05 March 2016
STEINN STEINGRÍMSSON, MARTIN I. SIGURDSSON, THOR ASPELUND, SIGMUNDUR SIGFÚSSON, ANDRÉS MAGNÚSSON
Steingrímsson S, Sigurdsson M I, Aspelund T, Sigfússon S, Magnússon A. Total populationbased study of the impact of substance use disorders on the overall survival of psychiatric inpatients. Nord J Psychiatry 2015; Early Online:1–6. Background: Patients with severe mental illness have a shortened lifespan, and substance use disorder (SUD) is an especially important diagnosis in this respect. There have been no studies comparing directly SUD to other diagnoses in a nationwide cohort. Aims: To directly compare differences in mortality rates of psychiatric inpatients with a discharge diagnosis of SUD versus other psychiatric diagnoses. Methods: A register-based study was made of all patients admitted to psychiatric hospitals in Iceland between 1983 and 2007. Patients were grouped according to discharge diagnoses. Survival with respect to SUD was compared using Cox-proportional hazard ratio, excluding those with an organic mental disorder. Furthermore, the survival of patients with SUD and co-morbid diagnoses was evaluated. Results: A total of 14,281 patients (over the age of 18 years) were admitted to a psychiatric hospital in Iceland during the study period, with a total of 156,356 years of follow-up. For both men and women, a diagnosis of SUD conferred similar mortality as a diagnosis of schizophrenia without SUD, while individuals with a diagnosis of a mood disorder or “other disorders” had significantly lower mortality than SUD. For men with SUD, a co-occurring mental disorder was associated with an increased risk of dying, however, this was not found for women. Conclusions: SUD was the psychiatric diagnosis that had the highest mortality rate among psychiatric inpatients, in both men and women. An additional psychiatric diagnosis on a pre-existing SUD diagnosis did increase the risk for men but not women. • Dual diagnosis, Mental disorders, Substance use disorders, Survival analysis Steinn Steingrímsson, Department of Psychiatry, Sahlgrenska University Hospital, Gothenburg, Sweden, E-mail: [email protected]; Accepted 9 June 2015.
N
umerous studies have shown increased all-cause mortality among psychiatric patients; from natural causes as well as accidents, suicide and homicide (1–6). Harris and Barraclough (1) reviewed 152 reports and concluded that psychiatric inpatients had a 2.8 times higher mortality rate than the general population (1). Furthermore, Wahlbeck et al. reported that psychiatric inpatients in Denmark, Finland and Sweden lived an average of 15 to 20 years less than the general population (7). Substance use disorder (SUD) is a common mental illness, both in the community and among psychiatric inpatients. Community estimates of 1-year prevalence of SUD differ widely in different parts of the world but are often estimated close to 2–3% (8). International prevalence studies have revealed that individuals with severe mental illness have significantly higher rates of substance use, particularly of alcohol, cannabis, and amphetamines,
© 2016 Informa Healthcare
compared to the general population (9, 10). Dixon et al. surveyed consecutive patients admitted to US inner-city psychiatric hospitals and found that two thirds of the patients with severe mental illness were either past substance abusers in remission or current substance abusers (11). In Iceland a large proportion of the psychiatric inpatients receive at least one diagnosis of an SUD with a narrowing difference between the sexes (12). There are very few studies that directly compare mortality rates between SUD patients and other patients admitted to psychiatric hospitals. Most previous studies have only examined survival or mortality rates of a single or few diagnostic categories and used the general population as reference. Thus, comparisons of survival rates for SUD and other diagnostic categories have most often been indirect between-studies comparisons. DOI: 10.3109/08039488.2015.1062143
S STEINGRÍMSSON ET AL.
This study takes advantage of several special features of the Icelandic population, which is rather homogenous and individuals are rarely lost to follow up. Furthermore, Iceland has several detailed population registers. The aim of this nationwide cohort study was to compare the survival of inpatients with SUDs to that of other psychiatric inpatients, and also to examine psychiatric co-morbidity of SUD patients and its effect on mortality.
Methods
Downloaded by [Orta Dogu Teknik Universitesi] at 08:00 05 March 2016
Study design There are two psychiatric centres in Iceland: one at the University Hospital in Reykjavík, in the capital, and another in a regional hospital in Akureyri. These two centres provide specialist care to psychiatric patients for all of Iceland. The database used contains information on all patients (n ⫽ 16,191) admitted to either of these two psychiatric hospitals in the period from 1 January 1983 in Reykjavík and 1 January 1985 in Akureyri; until 31 March 2008. The patients were born between 1892 and 1989 and 47.6% of the cohort were women. The variables in the database were collected from the admission office in each hospital and collated into a single data file. The variables were gender, age, time of both admission and discharge and discharge diagnoses, all of which were made by a specialist in psychiatry. The Bioethical Committee of Iceland (registration number: 08-051) and the Data Protection Authority of Iceland (2008/189) approved the study.
Study population Iceland is a Nordic country with a high level of education and high standards of living. There is one urbanized area, Reykjavik, in the southwest corner where the majority of the population lives, while the other parts of Iceland are rural areas. The larger psychiatric centre is located in Reykjavik, while ⬍ 10% of the patient cohort was admitted in Akureyri. There is also a 12-step-orientated clinic in Reykjavik where most uncomplicated SUD cases are treated.
The first time a patient was admitted was defined as the index admission. Admission criteria vary on an individual basis but are most often acute severe mental illness, e.g. psychotic symptoms, severe depression, manic state, increased suicidality, life-threatening withdrawal, or planned withdrawal treatment in patients with dual-diagnosis (defined as concomitant psychiatric illness with SUD). Patients were assigned to groups according to the discharge diagnoses made at the first contact. If a diagnosis of substance use disorder was made, the patient was assigned to group A. Patients with a diagnosis of schizophrenia or related disorder without a diagnosis of substance use disorder were assigned to group B. Mood disorders without the criteria for group A or B were assigned to group C and patients with none of the aforementioned diagnoses were assigned to group D. Patients diagnosed with an organic mental disorder at index admission were excluded. Conversion of diagnosis made with ICD-9 into ICD-10 was performed using official conversion tables from the World Health Organization (13). The details of the ICD-9 and ICD-10 diagnosis conversion are presented in Table 1. Patients under 18 years of age were excluded for the present study.
Statistical analysis All statistical analyses were by STATA (version 10). Diagnostic groups were compared using Kaplan-Meier analyses and Cox-proportional hazard ratio calculations. Analyses were performed separately for each gender and adjusted for age and year of index admission. In survival analysis all cases were censured at 20 years follow-up in order to minimize bias of smaller groups since few individuals had such a long follow-up. A two-tailed p value of ⬍ 0.05 was considered significant in all calculations.
Results Every patient admitted to a psychiatric hospital in Iceland from 1985 to March 2008, and most patients from 1983 and 1984 were included in this study. This comprises a total of 14,281 individuals with 156,356 years of follow-up. There were a total of 7439 men (52.1%) with
Table 1. Categories of mental illness with conversion principle between ICD-9 and ICD-10. Group of mental disorder SUD Schizophrenia and related disorders Mood disorders Other mental disorders
162
ICD-9 codes
ICD-10 codes
291, 292, 303–305 (excluding 305.1 and 305.8) 295, 297, 298.2–298.9 296, 298.0, 298.1, 300.4, 301.1, 311 300 – 302, 305.8, 306, 307.1, 307.4, 307.5, 307.8, 308, 309, 316 (excluding 300.4, 301.1 and 302.0)
F10–F19 (excluding F17) F20–F29 F30–F39 F40–F48, F50–F59, F60–F69 and other diagnoses excluding organic mental disorder
NOrD J PSYChIaTrY·VOL 70 NO 3·2016
Downloaded by [Orta Dogu Teknik Universitesi] at 08:00 05 March 2016
PSYCHIATRIC INPATIENT SURVIVAL AND SUBSTANCE USE
a total of 82,556 years of follow up and 6842 women (47.9%) with 73,800 years. The number of patients in each of the four diagnostic groups is shown in Table 2. The most common discharge diagnosis category at index admission was SUD for men (mean age men 39.6 years, standard deviation (⫾) 14.3, women 39.1 ⫾ 14.6) and mood disorders for women (mean age men 44.5 ⫾ 18.6, women 45.4 ⫾ 18.4). The third most common diagnostic group was “Other disorders” (mean age men 35.1 ⫾ 14.6, women 36.9 ⫾ 15.8) and least common was “Schizophrenia and related disorders” (mean age men 37.1 ⫾ 15.3, women 44.3 ⫾ 16.7). A total of 2533 deaths occurred in the study cohort within 20 years following index admission, 1409 men and 1124 women. Adjusted survival for each gender by the four diagnostic groups after index admission is presented in Fig. 1 and 2 with the estimated hazard ratios presented in Table 3. For both men and women, a diagnosis of SUD conferred similar mortality as a diagnosis of schizophrenia without SUD while individuals with a diagnosis of a mood disorder or “other disorders” had significantly lower mortality than SUD. The patients that had received a diagnosis of SUD at index admission were divided into two groups; those who did receive a diagnosis of schizophrenia and related disorders or a mood disorder (dual diagnosis) and those who did not. The hazard ratio of these two groups is compared in Table 4. For men, dual diagnosis was associated with an increased risk of dying; however, this was not statistically significant for women. In other words, for women, a diagnosis of SUD alone already carries a similar risk of dying as a dual diagnosis.
Discussion This nationwide cohort study with over a 25-year period found that psychiatric inpatients diagnosed with a SUD had a higher mortality than other psychiatric diagnoses examined, surpassing mood disorders and other psychiatric disorders such as anxiety-related, personality and behavioural disorders but equaling schizophrenia. A cooccurring diagnosis of schizophrenia and related disorders or mood disorders in addition to a SUD diagnosis
Fig. 1. Survival analysis of time to death (all-cause mortality) for men admitted to a psychiatric hospital in Iceland 1983–2008 comparing directly between discharge diagnoses groups adjusted for age at admission.
conferred statistically significant increased risk for men but not women. Many studies have compared mortality rates of psychiatric patients, including SUD patients, to the general population (1). However, there are fewer studies that directly compare the survival of different inpatients categories. The present study found that SUD carried the highest mortality of all diagnostic groups. This is in accordance with the extensive review of Harris and Barraclough (1), which found that SUDs and eating disorders had the highest mortality rates of all diagnostic groups compared to a healthy population. In their analyses, schizophrenia had considerably lower mortality rate than alcohol and other SUDs, both in men and women, while in this study; men with schizophrenia without comorbid SUD had similar mortality rates as men with SUD. Our finding of lower hazard ratios in the mood disorder and “other mental illness” groups compared to
Table 2. Demographics of psychiatric diagnosis at index admission in Iceland between 1981–2008. Men n ⫽ 7439
Women n ⫽ 6842
Mean age at index admission (years ⫾ SD) 39.9 ⫾ 15.7 41.7 ⫾ 17.0 Diagnostic group Substance use disorder (n (%)) 4312 (58.0) 2185 (31.9) Schizophrenia and related disorders (n (%)) 739 (9.9) 588 (8.6) Mood disorder (n (%)) 1601 (21.5) 2776 (40.6) Other mental disorder (n (%)) 787 (10.6) 1293 (18.9)
NOrD J PSYChIaTrY·VOL 70 NO 3·2016
Fig. 2. Survival analysis of time to death (all-cause mortality) for women admitted to a psychiatric hospital in Iceland 1983–2008 comparing directly between discharge diagnoses groups adjusted for age at admission.
163
S STEINGRÍMSSON ET AL.
Table 3. Hazard ratio of psychiatric diagnosis according to diagnosis at index admission compared to substance use disorder including effect of age and time of admission.
Downloaded by [Orta Dogu Teknik Universitesi] at 08:00 05 March 2016
Men n ⫽ 7451
Substance use disorder Schizophrenia ⫹ related Mood disorder Other mental disorder Age at index admission 18–30 years 30–40 years 40–50 years 50–60 years 60–70 years Over 70 years Year of admission 1980–1990 1990–2000 After 2000
Women n ⫽ 6842
Hazard ratio
95% CI
p
Hazard ratio
95% CI
p
reference group 1.07 0.87 0.81
0.90–1.28 0.76–0.99 0.65–0.99
0.44 0.04 0.04
reference group 0.89 0.66 0.65
0.72–1.08 0.57–0.77 0.52–0.80
0.24 ⬍ 0.001 ⬍ 0.001
reference group 1.50 2.69 5.22 11.05 25.49
1.19–1.89 2.17–3.33 4.25–6.41 8.99–13.57 20.52–31.66
0.001 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001
reference group 1.95 4.08 9.21 18.78 47.14
1.35–2.81 2.93–5.68 6.71–12.64 13.73–25.68 34.54–64.33
⬍ 0.001 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001
reference group 0.74 0.47
0.65–0.84 0.35–0.63
⬍ 0.001 ⬍ 0.001
reference group 0.66 0.52
0.57–0.76 0.38–0.72
⬍ 0.001 ⬍ 0.001
SUD is in line with previous studies (14). An important point, as shown in Table 2, is that mortality seems to have improved in this population, which may be due to a number of factors. An important perspective is that psychiatric diagnoses vary in the need for admission to psychiatric services and outpatient treatment services have changed in the recent decades, decreasing the need for acute treatment (15). The pattern of drug use in Iceland may explain why SUD does not carry a higher mortality rate than schizophrenia in men. In contrast to most of the USA and Europe, heroin use is very rare in Iceland, while the intravenous use of psycho-stimulants is high (16). Thus, deaths due to acute overdoses are relatively rare. Furthermore, human immunodeficiency virus did not spread to the intravenous drug users’ circles until recently (17).
In addition, there are extensive treatment alternatives for SUD patients in Iceland, with over 50 beds per 100,000 inhabitants free of charge and extensive outpatient services including 12-step and cognitive behavioral therapy programmes, compared to Sweden, for example, with approximately 40 beds which is generally high compared to most other countries (WHO atlas) (18). The average alcohol consumption in Iceland has historically been relatively low, until quite recently (19). Thorarinsson showed that the mortality of patients with alcohol use disorder in Iceland during the period 1951 to 1974 was twice that of the total population (5). During that period, annual per capita consumption of absolute alcohol among individuals over 15 years of age increased from 1.50 to 3.04 L (5). The consumption continued to rise and had reached 7.4 L in 2007 (19). Since the pattern
Table 4. Hazard ratio of co-morbid psychiatric diagnosis at index admission compared to substance use disorder alone. Men n ⫽ 7451 Hazard ratio Substance use disorder Co-morbid diagnosis Age at index admission 18–30 years 30–40 years 40–50 years 50–60 years 60–70 years Over 70 years Year of admission 1980–1990 1990–2000 After 2000
164
p
Hazard ratio
95% CI
p
1.05–1.66
0.02
ref 1.27
0.93–1.74
0.13
ref 1.33 2.42 4.55 9.16 19.30
0.998–1.79 1.86–3.16 3.50–5.89 7.01–11.98 14.15–26.32
0.05 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001
ref 2.31 4.66 8.65 14.39 27.89
1.33–3.99 2.83–7.66 5.31–14.09 8.58–24.13 15.92–48.86
0.003 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001
ref 0.70 0.50
0.65–0.84 0.34–0.74
⬍ 0.001 ⬍ 0.001
ref 0.60 0.61
0.46–0.78 0.37–0.99
⬍ 0.001 0.04
ref 1.32
95% CI
Women n ⫽ 6842
NOrD J PSYChIaTrY·VOL 70 NO 3·2016
Downloaded by [Orta Dogu Teknik Universitesi] at 08:00 05 March 2016
PSYCHIATRIC INPATIENT SURVIVAL AND SUBSTANCE USE
of consumption in Iceland is predominantly binge drinking, screening tools have given a high estimate of alcohol use disorders (20). In men, the mortality for SUD was no higher than for schizophrenia, and psychiatric co-morbidity resulted in an increased mortality risk. For women psychiatric comorbidity did not increase statistically the mortality risk. An opposite difference by gender was found in a study of homeless people in Denmark; in men with SUD, an additional diagnosis of schizophrenia spectrum disorders did not confer increased mortality, while this did increase the mortality for women with an SUD (21). In a Swedish register study, psychosis was associated with a reduced mortality in patients with SUD (22). Diagnosis at index admission should, however, be regarded as a preliminary diagnosis and in many cases another diagnosis will be provided at a later admission or outpatient contact, especially when one diagnosis can mask another, e.g. SUD and schizophrenia. This could mean that a patient that is diagnosed with SUD alone will at another time point be diagnosed with another disorder. This study analyses mostly firm end points such as gender, age and mortality. However, there are some limitations to the study. Firstly, some uncertainty remains when register data is used to obtain discharge diagnoses. Secondly, using index admission as the categorical definition may pose problems as another underlying diagnosis may be identified at a later stage. Thirdly, because of the abundance of inpatient treatment for SUD outside psychiatric wards in Iceland, only the more severe cases of SUD are likely be admitted to the psychiatric wards, making generalizability to other countries more difficult. As well as SUD, patients may be admitted due to severe withdrawal, which can be an indicator of the life-threatening effect of SUD, and therefore the results are not applicable directly to outpatients. As in all register-based research an important thing to remember is that both local guidelines and law affect the diagnosis pattern as well as management. For example, in countries where there is a financial gain attached to certain diagnoses the pattern is likely to be different, but since Iceland has free of charge health care the use of diagnostic codes are more likely to be based on clinical presentation. In summary, this study found that SUD was the psychiatric diagnosis that carried the highest mortality rate among psychiatric inpatients, both in men and women. An additional psychiatric diagnosis on a pre-existing SUD diagnosis did increase the risk for men but not for women. Declaration of interest: The study was funded by the Public Health Institute of Iceland and The scientific fund of Landspitali University Hospital. The authors declare no conflict of interest. The authors alone are responsible for the content and writing of the article. NOrD J PSYChIaTrY·VOL 70 NO 3·2016
References 1. Harris EC, Barraclough B. Excess mortality of mental disorder. Br J Psychiatry 1998;173:11–53. 2. Mortensen PB, Agerbo E, Erikson T, Qin P, Westergaard-Nielsen N. Psychiatric illness and risk factors for suicide in Denmark. Lancet 2000;355:9–12. 3. Hiroeh U, Appleby L, Mortensen PB, Dunn G. Death by homicide, suicide, and other unnatural causes in people with mental illness: a population-based study. Lancet 2001;358: 2110–12. 4. Hiroeh U, Kapur N, Webb R, Dunn G, Mortensen PB, Appleby L. Deaths from natural causes in people with mental illness: a cohort study. J Psychosom Res 2008;64:275–83. 5. Thorarinsson AA. Mortality among men alcoholics in Iceland, 1951–74. J Stud Alcohol 1979;40:704–18. 6. Hser YI, Hoffman V, Grella CE, Anglin MD. A 33-year follow-up of narcotics addicts. Arch Gen Psychiatry 2001;58:503–8. 7. Wahlbeck K, Westman J, Nordentoft M, Gissler M, Laursen TM. Outcomes of Nordic mental health systems: life expectancy of patients with mental disorders. Br J Psychiatry 2011;199:453–8. 8. Demyttenaere K, Bruffaerts R, Posada-Villa J. Gasquet I, Kovess V, Lepine JP et al. Prevalence, severity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys. JAMA 2004;291:2581–90. 9. Menezes PR, Johnson S, Thornicroft G, Marshall J, Prosser D, Bebbington P et al. Drug and alcohol problems among individuals with severe mental illness in south London. Br J Psychiatry 1996;168:612–19. 10. Scott J. Homelessness and mental illness. Br J Psychiatry 1993;162:314–24. 11. Dixon L, McNary S, Lehman AF. Remission of substance use disorder among psychiatric inpatients with mental illness. Am J Psychiatry 1998;155:239–43. 12. Steingrimsson S, Carlsen HK, Sigfusson S, Magnusson A. The changing gender gap in substance use disorder: a total population-based study of psychiatric in-patients. Addiction 2012;107:1957–62. 13. World Health Organization (WHO). The ICD-10 classification of mental and behavioural disorders. Conversion tables between ICD-8, ICD-9 and ICD-10. Revision 1. WHO/MNH/92.16.REV.1. Geneva: World Health Organization, 1994. 14. Hannerz H, Borga P, Borritz M. Life expectancies for individuals with psychiatric diagnoses. Public Health 2001;115:328–37. 15. Moltke K, Høegh EB, Sæbye D, Larsen PL, Reff KT, Knop J. Psychiatric emergency services in Copenhagen 2012: a 27-year psychiatric and demographic follow-up study. Nord J Psychiatry 2015;69(6):459–65. 16. Bjarnadottir GD, Haraldsson HM, Rafnar BO, Sigurdsson E, Steingrimsson S, Johannsson M, et al. Prevalent intravenous abuse of methylphenidate among treatment-seeking patients with substance abuse disorders: a descriptive population-based study. J Addict Med 2015;9(3):188–94. 17. Indridason H, Gudmundsson S, Karlsdottir B, Löve A, Briem H, Gottfredsson M. Long term nationwide analysis of HIV and AIDS in Iceland, 1983–2012. J AIDS Clin Res 2014;5:387. 18. World Health Organization (WHO). Mental Health Atlas 2011. 2011. Geneva: World Health Organization 19. Consumption of alcoholic beverages 1980–2007. Statistics Iceland 2011. Retrieved from http://www.statice. is/?PageID ⫽ 1253 & src ⫽ https://rannsokn.hagstofa.is/pxen/Dialog/ varval.asp?ma ⫽ VIS05120%26ti ⫽ Consumption⫹ of ⫹ alcoholic⫹ beverages ⫹ 1980-2007⫹⫹⫹%26path ⫽ ./Database/ visitolur/neyslaeldra/%26lang ⫽ 1%26units ⫽ Litres 20. Stefansson JG, Lindal E. Algengi geðraskana á StórReykjavíkursvæðinu [The prevalence of mental disorders in the Greater-Reykjavik area]. Laeknabladid 2009;95: 559–64. 21. Nielsen SF, Hjorthoj CR, Erlangsen A, Nordentoft, M. Psychiatric disorders and mortality among people in homeless shelters in Denmark: a nationwide register-based cohort study. Lancet 2011;377:2205–14.
165
S STEINGRÍMSSON ET AL. 22. Nyhlen A, Fridell M, Backstrom M, Hesse M, Krantz, P. Substance abuse and psychiatric co-morbidity as predictors of premature mortality in Swedish drug abusers: a prospective longitudinal study 1970–2006. BMC Psychiatry 2011; 11:122.
Downloaded by [Orta Dogu Teknik Universitesi] at 08:00 05 March 2016
Steinn Steingrímsson, Mental health services, Landspitali the National University Hospital of Iceland, Reykjavik, Iceland and Centre of
Ethics, Law and Mental Health (CELAM), Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden. Martin I. Sigurdsson, Faculty of Medicine, University of Iceland, Reykjavik, Iceland. Thor Aspelund, Faculty of Medicine, University of Iceland, Reykjavik, Iceland. Sigmundur Sigfússon, Akureyri Hospital, Akureyri, Iceland. Andrés Magnússon, Mental health services, Landspitali the National University Hospital of Iceland, Reykjavik, Iceland.
166
NOrD J PSYChIaTrY·VOL 70 NO 3·2016
ISSN: 0803-9488 (Print) 1502-4725 (Online) Journal homepage: http://www.tandfonline.com/loi/ipsc20
Total population-based study of the impact of substance use disorders on the overall survival of psychiatric inpatients Steinn Steingrímsson, Martin I. Sigurdsson, Thor Aspelund, Sigmundur Sigfússon & Andrés Magnússon To cite this article: Steinn Steingrímsson, Martin I. Sigurdsson, Thor Aspelund, Sigmundur Sigfússon & Andrés Magnússon (2016) Total population-based study of the impact of substance use disorders on the overall survival of psychiatric inpatients, Nordic Journal of Psychiatry, 70:3, 161-166, DOI: 10.3109/08039488.2015.1062143 To link to this article: http://dx.doi.org/10.3109/08039488.2015.1062143
Published online: 28 Aug 2015.
Submit your article to this journal
Article views: 35
View related articles
View Crossmark data
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ipsc20 Download by: [Orta Dogu Teknik Universitesi]
Date: 05 March 2016, At: 08:00
Total population-based study of the impact of substance use disorders on the overall survival of psychiatric inpatients
Downloaded by [Orta Dogu Teknik Universitesi] at 08:00 05 March 2016
STEINN STEINGRÍMSSON, MARTIN I. SIGURDSSON, THOR ASPELUND, SIGMUNDUR SIGFÚSSON, ANDRÉS MAGNÚSSON
Steingrímsson S, Sigurdsson M I, Aspelund T, Sigfússon S, Magnússon A. Total populationbased study of the impact of substance use disorders on the overall survival of psychiatric inpatients. Nord J Psychiatry 2015; Early Online:1–6. Background: Patients with severe mental illness have a shortened lifespan, and substance use disorder (SUD) is an especially important diagnosis in this respect. There have been no studies comparing directly SUD to other diagnoses in a nationwide cohort. Aims: To directly compare differences in mortality rates of psychiatric inpatients with a discharge diagnosis of SUD versus other psychiatric diagnoses. Methods: A register-based study was made of all patients admitted to psychiatric hospitals in Iceland between 1983 and 2007. Patients were grouped according to discharge diagnoses. Survival with respect to SUD was compared using Cox-proportional hazard ratio, excluding those with an organic mental disorder. Furthermore, the survival of patients with SUD and co-morbid diagnoses was evaluated. Results: A total of 14,281 patients (over the age of 18 years) were admitted to a psychiatric hospital in Iceland during the study period, with a total of 156,356 years of follow-up. For both men and women, a diagnosis of SUD conferred similar mortality as a diagnosis of schizophrenia without SUD, while individuals with a diagnosis of a mood disorder or “other disorders” had significantly lower mortality than SUD. For men with SUD, a co-occurring mental disorder was associated with an increased risk of dying, however, this was not found for women. Conclusions: SUD was the psychiatric diagnosis that had the highest mortality rate among psychiatric inpatients, in both men and women. An additional psychiatric diagnosis on a pre-existing SUD diagnosis did increase the risk for men but not women. • Dual diagnosis, Mental disorders, Substance use disorders, Survival analysis Steinn Steingrímsson, Department of Psychiatry, Sahlgrenska University Hospital, Gothenburg, Sweden, E-mail: [email protected]; Accepted 9 June 2015.
N
umerous studies have shown increased all-cause mortality among psychiatric patients; from natural causes as well as accidents, suicide and homicide (1–6). Harris and Barraclough (1) reviewed 152 reports and concluded that psychiatric inpatients had a 2.8 times higher mortality rate than the general population (1). Furthermore, Wahlbeck et al. reported that psychiatric inpatients in Denmark, Finland and Sweden lived an average of 15 to 20 years less than the general population (7). Substance use disorder (SUD) is a common mental illness, both in the community and among psychiatric inpatients. Community estimates of 1-year prevalence of SUD differ widely in different parts of the world but are often estimated close to 2–3% (8). International prevalence studies have revealed that individuals with severe mental illness have significantly higher rates of substance use, particularly of alcohol, cannabis, and amphetamines,
© 2016 Informa Healthcare
compared to the general population (9, 10). Dixon et al. surveyed consecutive patients admitted to US inner-city psychiatric hospitals and found that two thirds of the patients with severe mental illness were either past substance abusers in remission or current substance abusers (11). In Iceland a large proportion of the psychiatric inpatients receive at least one diagnosis of an SUD with a narrowing difference between the sexes (12). There are very few studies that directly compare mortality rates between SUD patients and other patients admitted to psychiatric hospitals. Most previous studies have only examined survival or mortality rates of a single or few diagnostic categories and used the general population as reference. Thus, comparisons of survival rates for SUD and other diagnostic categories have most often been indirect between-studies comparisons. DOI: 10.3109/08039488.2015.1062143
S STEINGRÍMSSON ET AL.
This study takes advantage of several special features of the Icelandic population, which is rather homogenous and individuals are rarely lost to follow up. Furthermore, Iceland has several detailed population registers. The aim of this nationwide cohort study was to compare the survival of inpatients with SUDs to that of other psychiatric inpatients, and also to examine psychiatric co-morbidity of SUD patients and its effect on mortality.
Methods
Downloaded by [Orta Dogu Teknik Universitesi] at 08:00 05 March 2016
Study design There are two psychiatric centres in Iceland: one at the University Hospital in Reykjavík, in the capital, and another in a regional hospital in Akureyri. These two centres provide specialist care to psychiatric patients for all of Iceland. The database used contains information on all patients (n ⫽ 16,191) admitted to either of these two psychiatric hospitals in the period from 1 January 1983 in Reykjavík and 1 January 1985 in Akureyri; until 31 March 2008. The patients were born between 1892 and 1989 and 47.6% of the cohort were women. The variables in the database were collected from the admission office in each hospital and collated into a single data file. The variables were gender, age, time of both admission and discharge and discharge diagnoses, all of which were made by a specialist in psychiatry. The Bioethical Committee of Iceland (registration number: 08-051) and the Data Protection Authority of Iceland (2008/189) approved the study.
Study population Iceland is a Nordic country with a high level of education and high standards of living. There is one urbanized area, Reykjavik, in the southwest corner where the majority of the population lives, while the other parts of Iceland are rural areas. The larger psychiatric centre is located in Reykjavik, while ⬍ 10% of the patient cohort was admitted in Akureyri. There is also a 12-step-orientated clinic in Reykjavik where most uncomplicated SUD cases are treated.
The first time a patient was admitted was defined as the index admission. Admission criteria vary on an individual basis but are most often acute severe mental illness, e.g. psychotic symptoms, severe depression, manic state, increased suicidality, life-threatening withdrawal, or planned withdrawal treatment in patients with dual-diagnosis (defined as concomitant psychiatric illness with SUD). Patients were assigned to groups according to the discharge diagnoses made at the first contact. If a diagnosis of substance use disorder was made, the patient was assigned to group A. Patients with a diagnosis of schizophrenia or related disorder without a diagnosis of substance use disorder were assigned to group B. Mood disorders without the criteria for group A or B were assigned to group C and patients with none of the aforementioned diagnoses were assigned to group D. Patients diagnosed with an organic mental disorder at index admission were excluded. Conversion of diagnosis made with ICD-9 into ICD-10 was performed using official conversion tables from the World Health Organization (13). The details of the ICD-9 and ICD-10 diagnosis conversion are presented in Table 1. Patients under 18 years of age were excluded for the present study.
Statistical analysis All statistical analyses were by STATA (version 10). Diagnostic groups were compared using Kaplan-Meier analyses and Cox-proportional hazard ratio calculations. Analyses were performed separately for each gender and adjusted for age and year of index admission. In survival analysis all cases were censured at 20 years follow-up in order to minimize bias of smaller groups since few individuals had such a long follow-up. A two-tailed p value of ⬍ 0.05 was considered significant in all calculations.
Results Every patient admitted to a psychiatric hospital in Iceland from 1985 to March 2008, and most patients from 1983 and 1984 were included in this study. This comprises a total of 14,281 individuals with 156,356 years of follow-up. There were a total of 7439 men (52.1%) with
Table 1. Categories of mental illness with conversion principle between ICD-9 and ICD-10. Group of mental disorder SUD Schizophrenia and related disorders Mood disorders Other mental disorders
162
ICD-9 codes
ICD-10 codes
291, 292, 303–305 (excluding 305.1 and 305.8) 295, 297, 298.2–298.9 296, 298.0, 298.1, 300.4, 301.1, 311 300 – 302, 305.8, 306, 307.1, 307.4, 307.5, 307.8, 308, 309, 316 (excluding 300.4, 301.1 and 302.0)
F10–F19 (excluding F17) F20–F29 F30–F39 F40–F48, F50–F59, F60–F69 and other diagnoses excluding organic mental disorder
NOrD J PSYChIaTrY·VOL 70 NO 3·2016
Downloaded by [Orta Dogu Teknik Universitesi] at 08:00 05 March 2016
PSYCHIATRIC INPATIENT SURVIVAL AND SUBSTANCE USE
a total of 82,556 years of follow up and 6842 women (47.9%) with 73,800 years. The number of patients in each of the four diagnostic groups is shown in Table 2. The most common discharge diagnosis category at index admission was SUD for men (mean age men 39.6 years, standard deviation (⫾) 14.3, women 39.1 ⫾ 14.6) and mood disorders for women (mean age men 44.5 ⫾ 18.6, women 45.4 ⫾ 18.4). The third most common diagnostic group was “Other disorders” (mean age men 35.1 ⫾ 14.6, women 36.9 ⫾ 15.8) and least common was “Schizophrenia and related disorders” (mean age men 37.1 ⫾ 15.3, women 44.3 ⫾ 16.7). A total of 2533 deaths occurred in the study cohort within 20 years following index admission, 1409 men and 1124 women. Adjusted survival for each gender by the four diagnostic groups after index admission is presented in Fig. 1 and 2 with the estimated hazard ratios presented in Table 3. For both men and women, a diagnosis of SUD conferred similar mortality as a diagnosis of schizophrenia without SUD while individuals with a diagnosis of a mood disorder or “other disorders” had significantly lower mortality than SUD. The patients that had received a diagnosis of SUD at index admission were divided into two groups; those who did receive a diagnosis of schizophrenia and related disorders or a mood disorder (dual diagnosis) and those who did not. The hazard ratio of these two groups is compared in Table 4. For men, dual diagnosis was associated with an increased risk of dying; however, this was not statistically significant for women. In other words, for women, a diagnosis of SUD alone already carries a similar risk of dying as a dual diagnosis.
Discussion This nationwide cohort study with over a 25-year period found that psychiatric inpatients diagnosed with a SUD had a higher mortality than other psychiatric diagnoses examined, surpassing mood disorders and other psychiatric disorders such as anxiety-related, personality and behavioural disorders but equaling schizophrenia. A cooccurring diagnosis of schizophrenia and related disorders or mood disorders in addition to a SUD diagnosis
Fig. 1. Survival analysis of time to death (all-cause mortality) for men admitted to a psychiatric hospital in Iceland 1983–2008 comparing directly between discharge diagnoses groups adjusted for age at admission.
conferred statistically significant increased risk for men but not women. Many studies have compared mortality rates of psychiatric patients, including SUD patients, to the general population (1). However, there are fewer studies that directly compare the survival of different inpatients categories. The present study found that SUD carried the highest mortality of all diagnostic groups. This is in accordance with the extensive review of Harris and Barraclough (1), which found that SUDs and eating disorders had the highest mortality rates of all diagnostic groups compared to a healthy population. In their analyses, schizophrenia had considerably lower mortality rate than alcohol and other SUDs, both in men and women, while in this study; men with schizophrenia without comorbid SUD had similar mortality rates as men with SUD. Our finding of lower hazard ratios in the mood disorder and “other mental illness” groups compared to
Table 2. Demographics of psychiatric diagnosis at index admission in Iceland between 1981–2008. Men n ⫽ 7439
Women n ⫽ 6842
Mean age at index admission (years ⫾ SD) 39.9 ⫾ 15.7 41.7 ⫾ 17.0 Diagnostic group Substance use disorder (n (%)) 4312 (58.0) 2185 (31.9) Schizophrenia and related disorders (n (%)) 739 (9.9) 588 (8.6) Mood disorder (n (%)) 1601 (21.5) 2776 (40.6) Other mental disorder (n (%)) 787 (10.6) 1293 (18.9)
NOrD J PSYChIaTrY·VOL 70 NO 3·2016
Fig. 2. Survival analysis of time to death (all-cause mortality) for women admitted to a psychiatric hospital in Iceland 1983–2008 comparing directly between discharge diagnoses groups adjusted for age at admission.
163
S STEINGRÍMSSON ET AL.
Table 3. Hazard ratio of psychiatric diagnosis according to diagnosis at index admission compared to substance use disorder including effect of age and time of admission.
Downloaded by [Orta Dogu Teknik Universitesi] at 08:00 05 March 2016
Men n ⫽ 7451
Substance use disorder Schizophrenia ⫹ related Mood disorder Other mental disorder Age at index admission 18–30 years 30–40 years 40–50 years 50–60 years 60–70 years Over 70 years Year of admission 1980–1990 1990–2000 After 2000
Women n ⫽ 6842
Hazard ratio
95% CI
p
Hazard ratio
95% CI
p
reference group 1.07 0.87 0.81
0.90–1.28 0.76–0.99 0.65–0.99
0.44 0.04 0.04
reference group 0.89 0.66 0.65
0.72–1.08 0.57–0.77 0.52–0.80
0.24 ⬍ 0.001 ⬍ 0.001
reference group 1.50 2.69 5.22 11.05 25.49
1.19–1.89 2.17–3.33 4.25–6.41 8.99–13.57 20.52–31.66
0.001 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001
reference group 1.95 4.08 9.21 18.78 47.14
1.35–2.81 2.93–5.68 6.71–12.64 13.73–25.68 34.54–64.33
⬍ 0.001 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001
reference group 0.74 0.47
0.65–0.84 0.35–0.63
⬍ 0.001 ⬍ 0.001
reference group 0.66 0.52
0.57–0.76 0.38–0.72
⬍ 0.001 ⬍ 0.001
SUD is in line with previous studies (14). An important point, as shown in Table 2, is that mortality seems to have improved in this population, which may be due to a number of factors. An important perspective is that psychiatric diagnoses vary in the need for admission to psychiatric services and outpatient treatment services have changed in the recent decades, decreasing the need for acute treatment (15). The pattern of drug use in Iceland may explain why SUD does not carry a higher mortality rate than schizophrenia in men. In contrast to most of the USA and Europe, heroin use is very rare in Iceland, while the intravenous use of psycho-stimulants is high (16). Thus, deaths due to acute overdoses are relatively rare. Furthermore, human immunodeficiency virus did not spread to the intravenous drug users’ circles until recently (17).
In addition, there are extensive treatment alternatives for SUD patients in Iceland, with over 50 beds per 100,000 inhabitants free of charge and extensive outpatient services including 12-step and cognitive behavioral therapy programmes, compared to Sweden, for example, with approximately 40 beds which is generally high compared to most other countries (WHO atlas) (18). The average alcohol consumption in Iceland has historically been relatively low, until quite recently (19). Thorarinsson showed that the mortality of patients with alcohol use disorder in Iceland during the period 1951 to 1974 was twice that of the total population (5). During that period, annual per capita consumption of absolute alcohol among individuals over 15 years of age increased from 1.50 to 3.04 L (5). The consumption continued to rise and had reached 7.4 L in 2007 (19). Since the pattern
Table 4. Hazard ratio of co-morbid psychiatric diagnosis at index admission compared to substance use disorder alone. Men n ⫽ 7451 Hazard ratio Substance use disorder Co-morbid diagnosis Age at index admission 18–30 years 30–40 years 40–50 years 50–60 years 60–70 years Over 70 years Year of admission 1980–1990 1990–2000 After 2000
164
p
Hazard ratio
95% CI
p
1.05–1.66
0.02
ref 1.27
0.93–1.74
0.13
ref 1.33 2.42 4.55 9.16 19.30
0.998–1.79 1.86–3.16 3.50–5.89 7.01–11.98 14.15–26.32
0.05 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001
ref 2.31 4.66 8.65 14.39 27.89
1.33–3.99 2.83–7.66 5.31–14.09 8.58–24.13 15.92–48.86
0.003 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001 ⬍ 0.001
ref 0.70 0.50
0.65–0.84 0.34–0.74
⬍ 0.001 ⬍ 0.001
ref 0.60 0.61
0.46–0.78 0.37–0.99
⬍ 0.001 0.04
ref 1.32
95% CI
Women n ⫽ 6842
NOrD J PSYChIaTrY·VOL 70 NO 3·2016
Downloaded by [Orta Dogu Teknik Universitesi] at 08:00 05 March 2016
PSYCHIATRIC INPATIENT SURVIVAL AND SUBSTANCE USE
of consumption in Iceland is predominantly binge drinking, screening tools have given a high estimate of alcohol use disorders (20). In men, the mortality for SUD was no higher than for schizophrenia, and psychiatric co-morbidity resulted in an increased mortality risk. For women psychiatric comorbidity did not increase statistically the mortality risk. An opposite difference by gender was found in a study of homeless people in Denmark; in men with SUD, an additional diagnosis of schizophrenia spectrum disorders did not confer increased mortality, while this did increase the mortality for women with an SUD (21). In a Swedish register study, psychosis was associated with a reduced mortality in patients with SUD (22). Diagnosis at index admission should, however, be regarded as a preliminary diagnosis and in many cases another diagnosis will be provided at a later admission or outpatient contact, especially when one diagnosis can mask another, e.g. SUD and schizophrenia. This could mean that a patient that is diagnosed with SUD alone will at another time point be diagnosed with another disorder. This study analyses mostly firm end points such as gender, age and mortality. However, there are some limitations to the study. Firstly, some uncertainty remains when register data is used to obtain discharge diagnoses. Secondly, using index admission as the categorical definition may pose problems as another underlying diagnosis may be identified at a later stage. Thirdly, because of the abundance of inpatient treatment for SUD outside psychiatric wards in Iceland, only the more severe cases of SUD are likely be admitted to the psychiatric wards, making generalizability to other countries more difficult. As well as SUD, patients may be admitted due to severe withdrawal, which can be an indicator of the life-threatening effect of SUD, and therefore the results are not applicable directly to outpatients. As in all register-based research an important thing to remember is that both local guidelines and law affect the diagnosis pattern as well as management. For example, in countries where there is a financial gain attached to certain diagnoses the pattern is likely to be different, but since Iceland has free of charge health care the use of diagnostic codes are more likely to be based on clinical presentation. In summary, this study found that SUD was the psychiatric diagnosis that carried the highest mortality rate among psychiatric inpatients, both in men and women. An additional psychiatric diagnosis on a pre-existing SUD diagnosis did increase the risk for men but not for women. Declaration of interest: The study was funded by the Public Health Institute of Iceland and The scientific fund of Landspitali University Hospital. The authors declare no conflict of interest. The authors alone are responsible for the content and writing of the article. NOrD J PSYChIaTrY·VOL 70 NO 3·2016
References 1. Harris EC, Barraclough B. Excess mortality of mental disorder. Br J Psychiatry 1998;173:11–53. 2. Mortensen PB, Agerbo E, Erikson T, Qin P, Westergaard-Nielsen N. Psychiatric illness and risk factors for suicide in Denmark. Lancet 2000;355:9–12. 3. Hiroeh U, Appleby L, Mortensen PB, Dunn G. Death by homicide, suicide, and other unnatural causes in people with mental illness: a population-based study. Lancet 2001;358: 2110–12. 4. Hiroeh U, Kapur N, Webb R, Dunn G, Mortensen PB, Appleby L. Deaths from natural causes in people with mental illness: a cohort study. J Psychosom Res 2008;64:275–83. 5. Thorarinsson AA. Mortality among men alcoholics in Iceland, 1951–74. J Stud Alcohol 1979;40:704–18. 6. Hser YI, Hoffman V, Grella CE, Anglin MD. A 33-year follow-up of narcotics addicts. Arch Gen Psychiatry 2001;58:503–8. 7. Wahlbeck K, Westman J, Nordentoft M, Gissler M, Laursen TM. Outcomes of Nordic mental health systems: life expectancy of patients with mental disorders. Br J Psychiatry 2011;199:453–8. 8. Demyttenaere K, Bruffaerts R, Posada-Villa J. Gasquet I, Kovess V, Lepine JP et al. Prevalence, severity, and unmet need for treatment of mental disorders in the World Health Organization World Mental Health Surveys. JAMA 2004;291:2581–90. 9. Menezes PR, Johnson S, Thornicroft G, Marshall J, Prosser D, Bebbington P et al. Drug and alcohol problems among individuals with severe mental illness in south London. Br J Psychiatry 1996;168:612–19. 10. Scott J. Homelessness and mental illness. Br J Psychiatry 1993;162:314–24. 11. Dixon L, McNary S, Lehman AF. Remission of substance use disorder among psychiatric inpatients with mental illness. Am J Psychiatry 1998;155:239–43. 12. Steingrimsson S, Carlsen HK, Sigfusson S, Magnusson A. The changing gender gap in substance use disorder: a total population-based study of psychiatric in-patients. Addiction 2012;107:1957–62. 13. World Health Organization (WHO). The ICD-10 classification of mental and behavioural disorders. Conversion tables between ICD-8, ICD-9 and ICD-10. Revision 1. WHO/MNH/92.16.REV.1. Geneva: World Health Organization, 1994. 14. Hannerz H, Borga P, Borritz M. Life expectancies for individuals with psychiatric diagnoses. Public Health 2001;115:328–37. 15. Moltke K, Høegh EB, Sæbye D, Larsen PL, Reff KT, Knop J. Psychiatric emergency services in Copenhagen 2012: a 27-year psychiatric and demographic follow-up study. Nord J Psychiatry 2015;69(6):459–65. 16. Bjarnadottir GD, Haraldsson HM, Rafnar BO, Sigurdsson E, Steingrimsson S, Johannsson M, et al. Prevalent intravenous abuse of methylphenidate among treatment-seeking patients with substance abuse disorders: a descriptive population-based study. J Addict Med 2015;9(3):188–94. 17. Indridason H, Gudmundsson S, Karlsdottir B, Löve A, Briem H, Gottfredsson M. Long term nationwide analysis of HIV and AIDS in Iceland, 1983–2012. J AIDS Clin Res 2014;5:387. 18. World Health Organization (WHO). Mental Health Atlas 2011. 2011. Geneva: World Health Organization 19. Consumption of alcoholic beverages 1980–2007. Statistics Iceland 2011. Retrieved from http://www.statice. is/?PageID ⫽ 1253 & src ⫽ https://rannsokn.hagstofa.is/pxen/Dialog/ varval.asp?ma ⫽ VIS05120%26ti ⫽ Consumption⫹ of ⫹ alcoholic⫹ beverages ⫹ 1980-2007⫹⫹⫹%26path ⫽ ./Database/ visitolur/neyslaeldra/%26lang ⫽ 1%26units ⫽ Litres 20. Stefansson JG, Lindal E. Algengi geðraskana á StórReykjavíkursvæðinu [The prevalence of mental disorders in the Greater-Reykjavik area]. Laeknabladid 2009;95: 559–64. 21. Nielsen SF, Hjorthoj CR, Erlangsen A, Nordentoft, M. Psychiatric disorders and mortality among people in homeless shelters in Denmark: a nationwide register-based cohort study. Lancet 2011;377:2205–14.
165
S STEINGRÍMSSON ET AL. 22. Nyhlen A, Fridell M, Backstrom M, Hesse M, Krantz, P. Substance abuse and psychiatric co-morbidity as predictors of premature mortality in Swedish drug abusers: a prospective longitudinal study 1970–2006. BMC Psychiatry 2011; 11:122.
Downloaded by [Orta Dogu Teknik Universitesi] at 08:00 05 March 2016
Steinn Steingrímsson, Mental health services, Landspitali the National University Hospital of Iceland, Reykjavik, Iceland and Centre of
Ethics, Law and Mental Health (CELAM), Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden. Martin I. Sigurdsson, Faculty of Medicine, University of Iceland, Reykjavik, Iceland. Thor Aspelund, Faculty of Medicine, University of Iceland, Reykjavik, Iceland. Sigmundur Sigfússon, Akureyri Hospital, Akureyri, Iceland. Andrés Magnússon, Mental health services, Landspitali the National University Hospital of Iceland, Reykjavik, Iceland.
166
NOrD J PSYChIaTrY·VOL 70 NO 3·2016
