The routine collation of health outcomes data from hospital treated subjects in the Health Outcomes Data Repository (HODaR): descriptive analysis from the first 20,000 subjects.

OBJECTIVES Health technology assessment requires data covering many different facets of treatment. A new resource, the Health Outcomes Data Repository (HODaR), is described and evaluated for its use in the pharmaceutical research and development process. METHODS Data were collated for subjects treated at Cardiff and Vale National Health Service (NHS) Hospitals Trust, United Kingdom. Inpatients are surveyed 6 weeks postdischarge by postal survey, whilst outpatients are handed a survey pack when they attend. Survey data cover sociodemographics, resource use, production losses, and quality of life. Electronic hospital data are available for all responders, and linked with survey returns. Sample characteristics, coverage of disease areas, and a more detailed description of data values for diabetes are described. RESULTS Survey responses relating to 16,188 admissions and 4476 outpatient attendances were available relating to around 2000 different diagnoses. Over 5000 pharmacy items and 400,000 biochemistry test results were available. Analysis of utility data showed a broad coverage of diseases. For patients with diabetes the pattern of EQ-5D scores across subgroups is not clear. Health service resource use showed a linear relationship with respect to number of comorbidities. CONCLUSIONS HODaR represents a new approach to accessing patient data, and gathers both routine and survey-based data. Although linking survey data to routine hospital systems is a complex task, which produces some limitations, it can produce health outcomes data at relatively low cost. Its performance within the pharmaceutical research and development process needs to be further evaluated in order to assess its most appropriate role.