Usefulness of case report literature in determining drugs responsible for toxic epidermal necrolysis.

To evaluate the usefulness of the case report literature concerning toxic epidermal necrolysis, we surveyed English-language reports published from January 1966 to April 1987, using epidermal necrolysis as a key word in the MEDLINE database. Of the 345 articles identified, 80 definitely or possibly contained individual case reports of at least 100 words; 59 of these 82 (72%) were available and reviewed. These 59 reports included a total of 73 separate cases of toxic epidermal necrolysis possibly related to drugs, which represents less than 5% of the case of toxic epidermal necrolysis that occurred in this period. In 62% of these 73 cases the apparent purpose of the report was to show an association between a given drug and the development of toxic epidermal necrolysis. The drugs most frequently associated with toxic epidermal necrolysis are allopurinol, the nonsteroidal anti-inflammatory agents, phenytoin, and the sulfonamide antibiotics. Given that only a small proportion of cases of toxic epidermal necrolysis appear in the case report literature, except to present association with drugs not previously reported, the contribution of case reports in providing an estimate of the relative risk of drug-associated toxic epidermal necrolysis is limited. Furthermore, our analysis suggests that the adoption of specific diagnostic criteria for and more uniform reporting of signs, symptoms, and therapy could greatly increase the usefulness of the case report literature. We propose a minimum set of informational and specific diagnostic criteria.