Systolic time intervals revisited: correlations with N-terminal pro-brain natriuretic peptide in a community population.

The usefulness of automated measurements of the systolic time intervals in community screening deserves investigation. The systolic time intervals, including pre-ejection period (PEP), the left ventricular ejection time (LVET), and PEP/LVET were automatically and rapidly determined from signals of electrocardiography, phonocardiography, and pulse volume recording in 1087 residents aged > or = 40 years in a community of homogeneous Chinese. Receiver operating characteristics analysis was performed to determine the cutoff values for PEP, ET, and PEP/LVET in predicting high N-terminal pro-brain natriuretic peptide (NT-proBNP). The prevalence of high NT-proBNP was 12.3% in men and 8.3% in women. NT-proBNP was linearly related to both PEP and PEP/LVET, while there was a U-shaped relationship between NT-proBNP and LVET. In men, the sensitivity and specificity in predicting high NT-proBNP levels were 60% and 60%, respectively, for PEP > or = 89 ms; 60% and 62% for PEP/LVET > or = 0.31; and 63% and 60% for LVET < 272 ms or LVET > or = 310 ms. Various combinations of the criteria further improved either sensitivity or specificity. Women had slightly different cutoff values and performance for the various criteria of abnormal systolic time intervals and their combinations. Automated systolic time intervals appear to be useful in the screening of subjects with high NT-proBNP in a community.