FISCHL and co-workers developed an index that they claimed would predict relapse in patients with asthma.1 The index had a 95 per cent sensitivity and a 97 per cent specificity when applied to a sample of emergency room patients discharged or hospitalized within 12 hours. The authors enumerated the advantages of this index, including cost-effectiveness and a shorter period of emergency care before hospitalization. Other investigators have reported that the index is both accurate2 and useful.3 Before this or any other index gains universal acceptance, we should demonstrate that the predictions are valid in a variety of settings.4,5 This validation. © 1984, Massachusetts Medical Society. All rights reserved.