Denmark

DENMARK

COMPUTERIZATION OF DANISH PHARMACIES

Similar to the Netherlands, GPs in Denmark act as gatekeeper to second line health care and provide most of the medical care. The majority (97%) of citizens are assigned a GP, who generate 90% of the prescriptions. Although initiatives have been taken to create GP databases, the most important source for pharmacoepi-demiological studies in Denmark to date constitute data from the Danish pharmacies that became increas-ingly computerized in the 1990s and have allowed for the establishment of regional and national prescrip-tion registries (Gaist et al., 1997). As part of its tax funded health care for all inhabitants, the Danish NHS provides medical attendance free of charge and reim-burses 50% of all expenditure on a wide range of prescribed medicines independently of the presenters income and employment status. Measured in defined daily doses, 73% of all medication sold in Denmark in 1996 was on prescription (Gaist et al., 1997).

The Danish NHS is divided into 16 sections. Each community pharmacy collects data on all prescription drugs and forwards data on reim-bursable medicines to their local NHS section on a monthly basis. These data form the basis for the two prescription registries, the Odense University Pharmaco-epidemiological Database (OPED) and the Pharmaco-epidemiological Prescription Database of North Jutland (PDNJ) (Sorensen and Larsen, 1994; Gaist et al., 1997). Publications from these databases are numerous and can be obtained from the depart-mental websites.

A third prescription register was established in 1995, the Register of Drug Statistics (RDS) at the Medicines Division of the National Board of Health, that collects information on all pharmacy transac-tions of prescribed drugs independent of reimburse-ment status and covers the entire population of 5.2 million inhabitants. Since 2003, study-specific access to these data is available through National Board of Health.

Denmark has a long tradition in registration; church files were established in 1645, and in 1968 the Danish Civil Registration System was established. This civil registration system, which is updated daily, is used in all registries and comprises the civil registration number (CRN), civil status, CRN of father/mother/children, death data, immigration and emigration. These data allow for exact assessment of follow-up time of all citizens. Other nationwide registries in Denmark are the hospital discharge registry, the cancer registry, birth registry, mortality files and the social registries.

PRESCRIPTION DATABASE OF NORTH JUTLAND, ODENSE UNIVERSITY PHARMACO-EPIDEMIOLOGICAL DATABASE AND THE NATIONAL DATABASE

The OPED database covers the county of Funen (population approximately 470 000) persons, and PDNJ covers the county of Jutland and other (approx-imately 1.6 million persons) and together they cover a representative sample of 40% of the Danish popula-tion (Nielsen et al., 1996; Gaist et al., 1997). Dispens-ing claims data that are collected in the systems comprise a unique patient identifier, the CRN, that allows longitudinal tracking of the patient through different layers of the health care system, the date of dispensing, the product code (unique for brand, quan-tity and formulation) and ATC code. The computer-ized drug-dispensing histories contain data concerning the dispensed drug, type of prescriber, dispensing date and dispensed amount. The prescribed dosing regi-men (and therefore legend duration) is not recorded in the systems. Over-the-counter medication and non-reimbursed drugs (such as sedatives, hypnotics, oral contraceptives and laxatives) or in-hospital drugs use are not registered. Population data are obtained from the Central Registration System every 6 months to track migration or date of death. Completeness of reimbursed dispensed drugs has proven to be good (Gaist et al., 1997).

Prescription data have been linked to local, regional and national hospitalization discharge data, cancer registries, psychiatric registry, death and birth registries for specific projects through the CRN in the PDNJ database (Nielsen et al., 1999; Olesen et al., 1999; Thrane and Sorensen, 1999; Dalton et al., 2000; Fonager et al., 2000; Larsen et al., 2000; Sorensen et al., 2000). For a more recent list of publications, the website ‘www.clin-epi.dk’ should be visited.

Almost all the OPED studies are based on prescrip-tion data only (Gaist, 1999; Bjerrum and Bergman, 2000). Odense University Pharmaco-Epidemiological Database and Prescription Database of North Jutland are public institution research registries. Data can be accessed upon approval of a protocol by the Steering Committees.

Data from the national prescription database can be linked to the hospital discharge registry, the cancer registry, birth registry, mortality files and social registries on the basis of the CRN. The Danish hospi-tal discharge registry comprises data on 99.4% of all discharges from Danish hospitals and includes the CRN code, dates of admission and discharge, the surgical procedures performed and up to 20 diag-noses classified according to the ICD-10 classification of diseases (Andersen et al., 1999). Access can be obtained for specific projects, but all analyses must be done in Statistics Denmark or by modem. Statistics Denmark links the registries and deletes the CRN. As there is no access to the CRN, paper records validation is not possible. Currently, the procedure for getting access to the data might take up to 6 months and is project related.