Pulseless electrical activity with witnessed arrest as a predictor of sudden death from massive pulmonary embolism in outpatients

Background: the objective was to determine clinical characteristics that can quickly distinguish sudden death from massive pulmonary embolism (MPE) from other causes of sudden death. Methods and results: all medical examiner reports from Charlotte, NC from 1992 to 1999 ( n=4926) were hand-searched for cases of sudden death which met the inclusion criteria: non-traumatic death, age 18–65 years, transported to an emergency department (ED), and autopsy performed. Supplemental data from ED and prehospital records were retrieved to complete documentation. Data were analyzed by univariate odds ratios (OR) followed by chi-square ( χ 2) recursive partitioning for decision tree construction. Three hundred eighty four cases met inclusion criteria; MPE was the second most frequent cause of cardiac arrest in this cohort (37/384, 9.6%). The mean age of subjects with MPE (40.2±11.1 years) was significantly lower compared with non-PE subjects (46.5±9.9 years). Pulseless electrical activity was observed as the initial arrest rhythm (primary PEA) in 52/384 (13.5%) subjects. Out of 52 subjects with primary PEA, 28 (53%) died from MPE. Odds ratio data indicated significant association of MPE with female gender, arrest witnessed by medical providers, presence of primary PEA, and return of spontaneous circulation. The most accurate decision rule to recognize MPE consisted of witnessed arrest+primary PEA. This rule generated sensitivity=67.6% and specificity=94.5% and yielded a posttest probability of MPE of 57%. Conclusions: outpatients with witnessed cardiac arrest and primary PEA carry a high probability of MPE. Contexto: O objectivo era determinar as caracterı́sticas clı́nicas que permitissem distinguir rapidamente a morte súbita devida a embolia pulmonar maciça (EPM) de outras causas de morte súbita. Método e resultados: Todos os relatórios de exames médicos provenientes de Charlottte, Carolina do Norte, entre 1992 e 1999 ( n=4926) foram verificados à mão procurando causas de morte súbita que cumpriam os critérios de inclusão: morte de causa não traumática, idade entre 18 e 65 anos, transportados para um departamento de emergência (DE) e autópsia realizada. Os dados foram completados com informação adicional dos registos pré-hospitalares e do DE. Ao dados foram analisados para probabilidades (“ odds ratios”) de variáveis singulares seguidas de analise de partição recursiva pelo qui-quadrado ( χ 2) para construção da árvore de decisão. Trezentos e oitenta e quatro casos cum