Estimating the incidence of systemic lupus erythematosus in a defined population using multiple sources of retrieval.

We studied the age-specific incidence of SLE in a defined population in southern Sweden during 1981-86. Female incidence was 5.4/100,000/year and male incidence 1/100,000/year. In comparison with previous studies, incident cases were characterized by a high age at diagnosis with the highest incidence in age groups 55-64 and 65-74, approximately 7.5/100,000/year. The incidence was low in children under 15 years of age, 0.4/100,000/year, and unexpectedly so in the 15-24 decade, 1.2/100,000/year. Patient retrieval was based on four separate sources: a computerized diagnosis register; referrals from both public health care physicians and private practitioners; and records of ANA positive individuals from the single laboratory serving the area. All incident patients were present in at least two of these sources and 11 patients were present in three. We interpret the high amount of overlap between retrieval sources, without patients detected in only one source, as evidence for a high degree of completeness in our retrieval of SLE patients.