[Skip to Navigation]
[Skip to Navigation Landing]
December 1986

Record Linkage: A Powerful Tool for Epidemiologic Analysis

Author Affiliations

Department of Dermatology Harvard Medical School Boston, MA 02115

Arch Dermatol. 1986;122(12):1383-1384. doi:10.1001/archderm.1986.01660240047012

In this issue of the Archives, Lindelöf and Eklund1 strongly argue that the risk of malignant skin tumors after grenz-ray treatment for benign skin disorders is small. Their work illustrates the extreme usefulness of data linkage between treatment records and a cancer registry. A follow-up study as comprehensive as this would be extraordinarily expensive and difficult to conduct without such direct record linkage. It is unlikely that the exemplary completeness of this study (99.3%) could be achieved through any other mechanism.

Studies based on record linkage between treatment records and population registries are not without problems. Accurate records of both the treatments used and of the end point are required for the results of a study based on record linkage to be valid. Since this study is based on treatment information from a single center, and the tumors counted are usually confirmed