Finnish health care registers are frequently used for research, but little attention has been paid on the validity of diagnoses in these registers, especially in the field of rheumatology. This dissertation investigated the validity of rheumatoid arthritis (RA) biobank diagnoses that were supplemented with data from the Finnish Care Register for Health Care and the Drug Reimbursement Register. Additionally, this dissertation evaluated the validity of hospital registry diagnoses of systemic sclerosis (SSc) and polymyalgia rheumatica (PMR). The validity of the diagnoses was evaluated retrospectively by a thorough manual review of all data in the medical records from the follow-up period of patients from five university and central hospitals in Finland. The goal was to understand the strengths and limitations of Finnish registry data in the field of rheumatology and how to best use this data for rheumatological research. Additionally, we examined the most common reasons and risk factors for misdiagnosis, especially for PMR.

The validity of an RA diagnosis could be greatly improved by including data on number of visits and medication reimbursement rights. By utilising this additional data, validity was high, particularly for seropositive RA. On the contrary, the validity of a single RA diagnosis was only moderate, especially for seronegative RA.

The registry diagnosis of SSc was moderately accurate, although the validity clearly improved when the diagnosis was made in the rheumatology department. The validity of a more specific diagnosis of limited cutaneous SSc, which is the most common SSc type in Finland, was high even for diagnoses made in any departments.

Among patients diagnosed with PMR at a university hospital, a third of the diagnoses were found to be incorrect following a more comprehensive assessment and clinical follow-up. The risk of misdiagnosis was particularly high in patients with atypical disease patterns, inadequate response to medication, and those not meeting classification criteria. The most common conditions that mimicked PMR were other types of inflammatory arthritis, noninflammatory musculoskeletal conditions, infectious diseases and malignancy.

Combining data from different sources, especially information on special reimbursement for medication, markedly improves the validity of registry diagnoses. Careful consideration of differential diagnosis is essential when diagnosing PMR to avoid misdiagnosis.