Toxic shock syndrome. I. Clinical exclusion of other syndromes by strict and screening definitions

Several clinical definitions of toxic shock syndrome have been proposed and used in extensive epidemiologic and clinical studies. Most of these definitions suggest (but usually do not require) that there be sufficient laboratory studies to exclude other potentially similar syndromes. Simplified definitions which broaden the spectrum of toxic shock syndrome illness have also been proposed but not validated. In this study, clinical findings of consecutive hospitalized patients were compared: nine with toxic shock syndrome (confirmed by a modification of the collaborative strict case definition) and 120 with potentially similar diagnoses (bacteremia with shock, meningococcemia, Staphylococcus aureus bacteremia, scarlet fever, toxic epidermal necrolysis, acute rheumatic fever, leptospirosis, Rocky Mountain spotted fever, rubeola, Kawasaki syndrome, erythema multiforme, and Stevens-Johnson syndrome). None of the 120 controls satisfied the clinical criteria of the modified strict definition of toxic shock syndrome, demonstrating its exclusionary properties even in the absence of additional laboratory data. A "simplified" screening definition was constructed which might be applied early in illness (i.e., at admission) and this definition distinguished all the patients with toxic shock syndrome from all but three (2.5%) of the 117 analyzable patients with other mucocutaneous or potential infectious shock syndromes. Applied prospectively in the state of Colorado passive/active reporting system, the screening definition identified 24 potential toxic shock syndrome cases of which 19 (76%) eventually were confirmed as toxic shock syndrome. Before being adopted and widely used, clinical syndrome definitions should be documented to exclude other potentially overlapping syndromes or should require additional mandatory exclusionary laboratory data.