I read with great interest the article entitled “Uncovering Histologic Criteria With Prognostic Significance in Toxic Epidermal Necrolysis” wherein the authors suggest that “the degree of inflammation predicts the clinical outcome approximately. . . . ”1 Strangely, 2 different statistical tests were performed to show ultimately that histologic grade was associated with survival. Rather than directly comparing the histologic grade with survival, the authors employed the transitive property and first showed that histologic grade was associated with total body surface area sloughing and then they showed that total body surface area sloughing was associated with survival. Thus, by the transitive property, they suggest that since A was associated with B and B with C, then A must be associated with C. This is not necessarily true.
Adams BB. A Consideration of Statistical Association. Arch Dermatol. 2005;141(12):1604–1605. doi:10.1001/archderm.141.12.1604-b
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: