To the Editor.—
The advent of the computerization of medical records has resulted in the publication of many studies attempting to report the coincidence of two diseases, and, thus, by inference, a relationship. Many important new insights have been gained through such research, but a few pitfalls should be mentioned.First, since such studies are retrospective, essential data for the diagnoses are not always available. This was a major problem in the article by Thomas and Su1 in the September Archives. The most important criterion for diagnosing dermatitis herpetiformis (DH) is the granular deposits of IgA in the dermal papillae on direct immunofluorescence microscopy,2 yet, in the article by Thomas and Su, only three patients had direct immunofluorescence microscopy performed. Although the other six patients may have had DH, other possibilities exist, including bullous lupus erythematosus (LE), which can mimic DH both on clinical and light microscopic examination.
Joel L. Schwartz. Coincidence or Substantive Relationship?. Arch Dermatol. 1984;120(6):708. doi:10.1001/archderm.1984.01650420018005