Study design

The register-linkage data of older patients with schizophrenia was obtained from the PERFormance, Effectiveness and Cost of Treatment episodes (PERFECT) project, which is a collaboration between five Finnish university hospital districts, the Social Insurance Institution of Finland (SII) and the National Institute for Health and Welfare (THL) (http://www.thl.fi/fi/tutkimus-ja-asiantuntijatyo/hankkeet-jaohjelmat/perfect). The basic data on adult patients with schizophrenia was collected from different national registers and collated by THL. Then those aged 65 or older were separated into their own group. The personal data on each patient was encrypted before delivery from THL to this study. The responsibility for this part was assigned to psychiatrist Tiina Talaslahti MD. The study protocol was approved by the Ethics Committee of the National Institute for Health and Welfare.

4.1.1 The Finnish Hospital Discharge Register

The history of the Finnish Hospital Discharge Register begins from 1967 when the National Board of Health administrated a new register covering all discharges from public hospitals. Since 1994 the FHDR, nowadays known as the Care Register for Health Care, also includes all social institutions, such as residential care homes for mentally disabled or the elderly. The Parliament of Finland passed the legislation on health registers in 1987. The purpose was to start to collect medical and health information on citizens and to protect individual’s privacy. The personal identification numbers used in the FHDR since 1968 make it possible to collate the information between the FHDR and different national registers (Gissler and Haukka 2004).

The FHDR contains data on all patients admitted to mental and general hospitals, municipal health centres, military hospitals, as well as prison and private hospitals

46

(Miettunen et al. 2010). It includes information on patients’ age, gender, place of residence, dates of admission and discharge as well as present status of future treatment or residence (long-term care, home, residential care home for the elderly etc.), information about hospital districts and main diagnosis and subdiagnoses of hospitalization at discharge. The statistics of the FHDR is structured on care notifications submitted by health care units all over Finland and carefully checked and corrected by THL if needed. Since 1998 it has also contained specialized outpatient care and day surgery, and since 2010 outpatient care in health centres.

The disease classification of the FHDR is based on the International Classification of Diseases (ICD) by the World Health Organization (WHO). The ICD was established in the 1960’s and it has gone through numerous revisions in recent decades. In the FHDR, diseases have been coded according to the ICD-10 since 1996, and the versions ICD-8 and ICD-9 (with Diagnostic and Statistical Manual, 3rd revised edition, DSM-III-R) were used in 1967-1986 and 1987-1995 respectively.

Those subjects of the present sample who have been collected from the FHDR had received schizophrenia or schizoaffective disorder as the main diagnosis at least once in the FHDR (ICD-8, ICD-9: 295, ICD-10: F20, F25). The diagnoses of physical or other diseases of each subject were also collected from the registers of the FHDR (main diagnosis between 1987 and 1998): high blood pressure (ICD-10:

I10-I15), atrial fibrillation (I48), congestive heart disease (I50), coronary artery disease (I20-I25), diabetes mellitus (E10-E14), chronic obstructive pulmonary disease (COPD) and asthma bronchiale (J44-J46), neoplasms (C00-C99), alcohol dependence (F10), dementia (F00-F03, G30), and high level of cholesterol (E78).

4.1.2 Registers of the Social Insurance Institution and the Finnish Center for Pensions

The Finnish pension scheme is based on two separate pensions: disability pensions provided by the SII since 1962 and also earnings-related pensions provided by the FCP since 1962. For the present Studies I and II, patients with a disability pension due to schizophrenia were extracted from these registers in order to identify as many

47 patients with schizophrenia as possible in the age group of 60 years or more. The registers of the SII also contain information on medication purchases and reimbursements for medical expenses used in these studies, as well as various details about the rates and incidence of benefits and their distribution by region and population group (SII 2015).

The diagnoses of physical or other diseases of each subject were collected from the SII in 1998 (reimbursement for medicine expenses and medicines purchased).

The data on medicines purchased consisted of antipsychotics and antidepressants used between 1998 and 2003. Antipsychotics were collected with the help of Anatomical Therapeutic Chemical (ATC) classification codes and further divided into four different groups according to the purpose of use and ATC-codes: 1. no use of antipsychotics, 2. use of First Generation Antipsychotics [FGAs, including N05AA01 (chlorpromazine), N05AA02 (levomepromazine), N05AA03 (promazine), N05AB02 (fluphenazine), N05AB03 (perphenazine), N05AC02 (tioridazine), N05AC01 (periciazine), N05AD03 (melperone), N05AD01 (haloperidol), N05AF01 (flupentixol), N05AF03 (chlorprothixene), N05AF05 (zuclopenthixol), N05AL01 (sulpiride)]), 3. use of Second Generation Antipsychotics [SGAs, including N05AX08 (risperidone), N05AH04 (quetiapine), N05AH03 (olanzapine), N05AE03 (sertindole), N05AH02 (clozapine)] and 4. use of both FGAs and SGAs. In addition, the use of antidepressants (code N06A) during 1998 was extracted from the SII.

4.1.3 The Finnish Causes of Death Register

The classification of causes of death was that used by Statistics Finland. The Finnish Causes of Death Register (FCDR) includes all deaths and causes of death among Finnish citizens and permanent residents since 1969. This register includes dates, places of death and diagnoses. The validity of this register has been scientifically demonstrated (Lahti and Penttilä 2001).

The death certificate is usually based on the diseases known before death and all the certificates issued from the physicians are checked by a forensic pathologist at the National Institute for Health and Welfare. If death is unexpected or if the deceased was not under medical supervision, a forensic autopsy is performed in order to ascertain the cause of death. In the period 1999-2008, an autopsy was done

48

after every fifth death in the age group 65 or older. The causes of death in the present Studies I and III have been categorized into two main groups: unnatural and natural causes of death. The former group includes the codes for suicides (ICD-10: X60-X84), accidents (ICD-10: V01-X59, Y10-Y86, Y972 and Y88-Y89) and homicides (ICD-10: 85-Y871), the latter includes all the other diagnoses such as various diseases.