1
RATIONALE Schizophrenia is a chronic, severe and disabling mental disorder. The disorder is characterized by a variety of clinical manifestaons, including symptoms that are both posive and negave, reduced social funcon and community parcipaon, and cognive impairment. Medicaon nonadherence rates are high in schizophrenia, as they are in many chronic medical condions. Currently there is a dearth of long-term data comparing the use and effecveness of depot and oral anpsychocs aſter first- episode psychosis. AIM OF THE STUDY 1) To evaluate the disconnuaon of the inial anpsychoc treatment aſter the first hospital episode of schizophrenia, and 2) To compare the risk of rehospitalisaon due to schizophrenia between different anpsychocs aſter the first hospital episode of schizophrenia in a naonwide cohort of 2,588 paents hospitalized for the first me with schizophrenia during 2000-2007 in Finland [1]. METHODS The study was conducted as a register-based case linkage study ulizing data on hospitalizaon, mortality and anpsychoc prescripons. Hazard raos (HR) with 95% confidence intervals (CI) were calculated adjusng for the effects of sociodemographic and clinical variables, the temporal sequence of the anpsychocs used, and the choice of the inial anpsychoc. The primary outcome measures were: 1) all-cause disconnuaon of the inial anpsychoc medicaon, 2) rehospitalizaon due to schizophrenia, and 3) death from any cause. ACKNOWLEDGEMENTS • The study was supported by the Annual EVO Financing (special government subsidies from the Ministry of Health and Welfare, Finland) and by Janssen-Cilag. • The funders were not involved in the conduct of the study or in the collecon, management, analysis, or interpretaon of the data. • The research permission numbers to use the data were obtained from • the Social Insurance Instute (Kela 14/522/2009) • the Naonal Instute for Health and Welfare (Dnro 206/5.05.00/2009) • Stascs Finland (TK-53-739-09) Pasi Korhonen 1,2 , Jari Haukka 2,3,4 , Mark Taylor 5 , Peter M. Haddad 6,7 , Maxine X. Patel 8 , Jari Tiihonen 9,10 1 StatFinn Oy, Espoo, Finland; 2 EPID Research Oy, Espoo, Finland; 3 Department of Mental Health and Alcohol Research, National Institute for Health and Welfare, Helsinki; 4 Tampere School of Public Health, University of Tampere, Finland; 5 NHS Lothian, Edinburgh, U.K.; 6 Greater Manchester West Mental Health NHS Foundation Trust, Salford, U.K.; 7 University of Manchester, Manchester, U.K.; 8 Department of Psychosis Studies, Institute of Psychiatry, King’s College London; 9 Department of Forensic Psychiatry, University of Eastern Finland, Niuvanniemi Hospital, Kuopio; 10 Department of Clinical Physiology, Kuopio University Hospital, Kuopio Schizoaffecve disorders ICD-10 F25 N paents = 452 (6.1%) Longer illness cohort ICD-10 F20-F25 N paents = 25 884 (78%) Naïve cohort ICD-10 F20-F25 N paents = 7 434 (22%) Hospitalizaons in 2000-2007 Diagnosis ICD-10 codes F20-F25 N paents = 33 318 N hospitalizaons = 122 193 Schizophrenia ICD-10 F20 N paents = 2588 (34.8%) Definion of the naïve cohort • No prior hospitalizaon with F20 – F25 diagnoses • No anpsycocs use within 6 months prior to the first hospitalizaon Study populaon: • 16-65 years of age, mean age 37.5 years • 62% male, 38% female Induced delusional disorder ICD-10 F24 N paents = 11 (0.1%) Acute and transient psychoc disorders ICD-10 F23 N paents = 2689 (36.2%) Persistent delusional disorder ICD-10 F22 N paents = 1363 (18.3%) Schizotypal disorder ICD-20 F21 N paents = 331 (4.5%) 2000 2007 Prescripon register Drug substance (ATC) Date of purchase Amount (DDD) VNR-code (package code) Cause of death register Causes of death (ICD10) Date of death Hospital care register Diagnosis (ICD10) Start and discharge date Age, sex Hospital & hospital district d o i r e p p u - w o l l o F y r o t s i H Paent level informaon DATA SOURCES STUDY POPULATION Figure 2. Descripon of the study populaon Figure 1. Linkage of paent level data on hospitalizaons, anpsychoc prescripons and mortality RESULTS Pairwise comparison of depot vs. oral formulaons Figure 5. Risk of disconnuaon of inial medicaon Figure 6. Risk of rehospitalizaon In pooled analysis the depot anpsychocs were associated with a 59% lower risk of disconnuaon (HR=0.41; 95% CI 0.27-0.61; p<0.0001) and with a 64% lower risk of rehospitalizaon (HR=0.36; 95% CI 0.17-0.75; p=0.007) than their respecve oral formulaons. ABSTRACT Objecve: To compare the risk of rehospitalizaon and disconnuaon of inial treatment between different anpsychocs aſter first hospital episode of schizophrenia. Methods: A register-based linkage study ulizing data on hospitalizaons, mortality, and anpsychoc prescripons in a cohort of 2588 paents hospitalized for the first me due to schizophrenia in Finland between 2000 and 2007. Adjusted hazard raos with 95% confidence intervals (CI) were calculated. Results: About 58% of paents collected a prescripon for an anpsychoc within 30 days aſter discharge from hospital and 46% connued their inial treatment for at least 30 days. Among oral anpsychocs, clozapine and olanzapine were associated with more favourable outcomes. Depot injecons were associated with a 59% (95% CI 39%-73%) lower risk for treatment disconnuaon and a 64% (95% CI 25%-83%) lower risk for rehospitalisaon than their oral equivalents. Use of any anpsychoc vs. no anpsychoc was associated with a 55% (95% CI 33%-69%) lower mortality. Conclusions: Adherence to the inial anpsychoc treatment aſter hospitalisaon is low. Use of depot anpsychocs was associated with a significantly lower risk of treatment disconnuaon and rehospitalizaon than use of oral formulaons of the same compounds. Use of any anpsychoc was associated with lower mortality. Key words: register-based study, schizophrenia, anpsychocs, delayed acon preparaons, effecveness REHOSPITALIZATION RISK AND DISCONTINUATION OF INITIAL ANTIPSYCHOTIC TREATMENT AFTER FIRST HOSPITAL EPISODE OF SCHIZOPHRENIA CONCLUSIONS • The first study of the adherence and comparave effecveness of specific anpsychoc treatments in a large unselected populaon of paents in a real-world seng. • The findings are not aributable to selecon bias since the study populaon included all paents in Finland with a first hospitalizaon for schizophrenia. Also we adjusted for a number of clinical and sociodemographic factors in the analyses and made a further adjustment by using the choice of the inial anpsychoc, reflecng the paent’s clinical status at baseline. • Only a minority of paents were adherent to their inial anpsychoc during the first 60 days aſter discharge from their first hospitalizaon for schizophrenia. • The observed higher disconnuaon rates in our study in comparison with randomized controlled trials, the CATIE and EUFEST studies [2-4], may be partly explained by paents in real-life seng being less movated or co-operave because roune clinical care and follow-up are less intensive than in randomized controlled trials. • Oral clozapine and olanzapine were associated with more favorable outcomes regarding all-cause treatment disconnuaon and rehospitalizaon, similarly as in the CATIE and EUFEST studies [2-4]. • Depot injecons were associated with lower risk of treatment disconnuaon and rehospitalizaon than their equivalent oral formulaons. A systemac meta-analysis of randomized controlled trials [5] did not find such a difference which may be explained by selecon because nonadherent paents cannot be forced to parcipate in randomized trials, observaonal studies are the only way to invesgate this issue. • Because there is no compulsory outpaent treatment in Finland, our results on the comparave effecveness of specific anpsychocs can be generalized only to paents who are willing to use anpsychoc treatment in outpaent care. Figure 3. Risk of all-cause disconnuaon of inial an- psychoc treatment started within 30 days of discharge aſter the index hospital episode (N=1,507)* Of 2,588 paents with a first hospitalizaon • 1,507 (58.2%) started using anpsychoc medicaon during the first 30 days aſter discharge. • 1,182 (45.7% of total) connued using the inial anpsychoc medicaon for 30 days and longer. • During a mean follow-up period of 2 years 1,394 paents (53.9%) disconnued their inial anpsychoc. • Reasons for disconnuaons were 50.6% (n=705) because of a change in anpsychoc prescripon, 34.3% (n=478) because of disconnuaon, 14.8% (n=207) because of hospitalizaon, and 0.3% (n=4) because of death. *)The hazard raos are adjusted by using age at diagnosis, sex, length of first hospital episode, and current and previous use of anxiolycs, hypnocs and sedaves, andepressants, drugs used in addicve disorders, analgesics, an-parkinsonian drugs, blood glucose-lowering drugs, and lipid-modifying agents as covariates. Figure 4. Risk of rehospitalizaon by anpsychoc treatment (N=2,588)** • During a mean follow-up period of 2 years, 1,496 paents (57.8%) were rehospitalized because of relapse of schizophrenia symptoms. • The use of any anpsychoc was associated with a 62% (95% CI 57%-66%) lower risk when compared to no treatment at all. • Among oral anpsychocs, compared with risperidone, clozapine (HR=0.48; 95% CI 0.31-0.76) and olanzapine (HR=0.54; 95% CI 0.40-0.73) were each associated with a lower rehospitalizaon risk. **)The hazard raos are adjusted by using age at diagnosis, sex, length of first hospital episode, previous use of an-psychocs during the follow-up, the choice of inial anpsychoc, current and previous use of anxiolycs, hypnocs and sedaves, andepressants, drugs used in addicve disorders, analgesics, an-parkinsonian drugs, blood glucose-lowering drugs, and lipid-modifying agents as covariates. Risk of all-cause death • During the follow-up period 160 (6.2%) paents died. • The use of any anpsychoc compared with no an- psychoc was associated with a 55% lower mortality (HR=0.45; 95% CI; 0.31-0.67). 1. Tiihonen J et al., Am J Psychiatry (published online March 1, 2011) 2. Lieberman JA et al., N Engl J Med 2005; 353:1209-1223 3. Kahn RS et al., Lancet 2008; 371:1085-1097 4. McEvoy JP et al., Am J Psychiatry 2006; 163:600-610 5. Adams CE et al., Br J Psychiatry 2001; 179:290-299 REFERENCES Contact: Pasi Korhonen [email protected] tel. +358 50 365 2990
