Combined clinical and biomarker model predicts AHF risk
MedWire News: Researchers have developed a new model which accurately predicts risk for acute heart failure (AHF) in patients presenting to an emergency department with undifferentiated dyspnea.
The model combines both clinical values and levels of N-terminal pro–B-type natriuretic peptide (NT-proBNP) and was able to reclassify a substantial number of patients who were initially only assessed clinically.
“The model appears to quickly and reliably redirect the undecided clinician for diagnosing AHF,” said study co-author Brian Steinhart (Saint Michael’s Hospital, Toronto, Ontario, Canada) and colleagues in the Journal of the American Colleague of Cardiology.
Biomarkers are increasingly being used to assist in the diagnosis of AHF in undifferentiated patients, with NT-proBNP and its parent analogue BNP being key players.
Studies have found good sensitivities for detecting AHF in such patients, but relatively lower positive predictive values.
Steinhart et al say part of the problem lies in the way validation studies are conducted.
“Although discrete cut points allow for ease of use, they are often derived from clinical trials that may not have a typical cross section of those patients who are tested in “real life” situations,” they commented.
The researchers hypothesized that a model that uses a continuous measure of NT-proBNP alongside clinical variables may be better. To investigate they recruited 500 patients admitted to an emergency department with undifferentiated dyspnea.
All patients initially underwent chest radiography and electrocardiography. These were used to estimate “pre-test probability” of AHF based on the subgroups: low (20% or less); high (80% or more); and intermediate (21–79%). The researchers subsequently measured NT-proBNP based on immunoassays.
On completion of the study, adjudication for AHF was determined independently by use of the Framingham Heart Score and National Health and Nutrition Examination Survey.
On collation of the data, Steinhart and colleagues found that a model incorporating age, pre-test probability, and log NT-proBNP was a better predictor of confirmed AHF than either of these factors alone.
Indeed, when validated in a separate cohort of 573 emergency dyspnea patients, the model correctly reclassified 44% of patients originally in the intermediate pretest-probability group to either low or high AHF risk.
Commenting on the potential of the model developed in their study, Steinhart et al suggest: “It could ultimately lead to improved health outcomes and streamlined research in this very challenging patient population.”
MedWire (www.medwire-news.md) is an independent clinical news service provided by Current Medicine Group, a trading division of Springer Healthcare Limited. © Springer Healthcare Ltd; 2009
By Andrew Czyzewski