The Hospital Data Project, Comparing hospital activity within Europe

• Main Author: Hugh F. Magee
• Other Authors: The Pennsylvania State University CiteSeerX Archives
• Format: Text
• Language: English
• Published: 2003
• Subjects: EU Health monitoring programme; health systems; hospital activity data; ICD coding From its commencement in 1997; the European Union’s (EU) Health Monitoring Programme (HMP)1 has; Iceland
• Online Access: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.614.7444; http://eurpub.oxfordjournals.org/content/13/suppl_1/73.full.pdf

Background: The ability to measure and compare hospital activity between EU member states is important for policy, planning, financing and assessment of population health. Earlier initiatives in this area have been largely directed at standardising high-level indicator definitions without proper account of differences in health systems and health information systems. The Hospital Data Project (HDP) develops a methodology for improved comparability of hospital inpatient and day case activity data across Europe and produces a pilot common data set. All EU members, Iceland and the World Health Organisation are participants. Methods: The approach comprises a detailed inventory of patient-level hospital data, identification of common areas, specification of data transformations and production of pilot data sets and metadata in a common format. An expert group developed a new diagnosis shortlist based on ICD-10. The project takes account of current work in the area of health care and morbidity indictors and applies the functional specification of health systems developed by the OECD. Results: Seventeen countries have submitted data and metadata in the common format for a single year. Data on inpatients and day cases are classified by age, gender, diagnosis and type of admission. Numbers of hospital discharges, mean and median lengths of stay and population rates are reported. Test data on selected hospital procedures has also been collected. The full data set contains approximately 500,000 records, and software has been designed to facilitate validation and use. Conclusion: Results to date are promising. It is a first step in a complex area, and further work is required to extend and refine this approach.