Failure of simple clinical measurements to predict perfusion status after intravenous thrombolysis

R. M. Califf, W. O'Neil, R. S. Stack, L. Aronson, D. B. Mark, S. Mantell, B. S. George, R. J. Candela, D. J. Kereiakes, C. Abbottsmith, E. J. Topol, W. W. O'Neill, J. A. Walton, E. R. Bates, S. G. Ellis, P. D.V. Bourdillon, M. A. Schork, E. Kline

Research output: Contribution to journalArticlepeer-review

218 Scopus citations


To determine whether coronary patency could be detected early during thrombolytic therapy, commonly used markers of perfusion were recorded in 386 patients with acute myocardial infarction treated with tissue plasminogen activator. Infarct artery angiography 90 minutes after initiation of therapy was used to determine perfusion status. Of patients with complete resolution of ST segment elevation before the angiogram, 96% (95% confidence interval, 79% to 100%) showed perfusion on the angiogram, and among those with partial improvement, 84% (95% confidence interval, 76% to 90%) showed perfusion, but these findings occurred in only 6% and 38% of patients respectively. When complete resolution of chest pain occurred before the angiogram, 84% of patients (95% confidence interval, 75% to 90%) showed perfusion, but this finding occurred in only 29% of patients. Although arrhythmias occurred frequently in the first 90 minutes of therapy, none were associated with a higher patency rate. No other factors predicted coronary patency. A logistic regression model showed 25% of patients with 90% or greater probability of patency, but 56% of patients with no ST segment or symptom resolution had patent arteries.

Original languageEnglish (US)
Pages (from-to)658-662
Number of pages5
JournalAnnals of internal medicine
Issue number5
StatePublished - 1988

ASJC Scopus subject areas

  • Internal Medicine


Dive into the research topics of 'Failure of simple clinical measurements to predict perfusion status after intravenous thrombolysis'. Together they form a unique fingerprint.

Cite this