[Skip to Content]
[Skip to Content Landing]
November 28, 1990

The Diagnosis of Pulmonary Embolism

Author Affiliations

Stanford (Calif) University School of Medicine

Stanford (Calif) University School of Medicine

JAMA. 1990;264(20):2624. doi:10.1001/jama.1990.03450200031014

To the Editor.—  In the PIOPED study,1 no data are presented that support the clustering of specific outcomes in the five categories presented. Why do the authors not show the distribution of the three outcomes clustered under high probability and indicate whether they all had identical predictive value?When they write "without corresponding ventilation or roentgenographic abnormalities," do they mean that both abnormalities are absent or that either one or the other is absent? What does the term "high-probability scans" mean? Surely not that the authors believe that a scan was performed? Nor, since the authors accept that the predictive value depends on the population under study, can they mean that pulmonary embolism is likely.If a sign (eg, "intermediate probability") is present in 45% of patients without pulmonary embolism and 42% of those with, it seems quaint to define the sensitivity in terms of the presence, rather than