$50 Blood Test Could Predict CHF Readmission Risk

Christine Leccese, for HealthLeaders Media, March 4, 2011

A simple blood test may be able to predict the likelihood that a patient with congestive heart failure will be readmitted to the hospital or die within a year of a hospital discharge, says a study published in the online issue of the American Journal of Cardiology. The inexpensive blood test, performed at admission and again at discharge, can help doctors make important decisions about care, and reduce the risk of costly readmissions.

Researchers examined heart failure patients on admission and discharge to determine levels of a certain protein (NT-proBNP) that's a marker for heart stress. They found that if a patient's levels decreased by less than 50% during their hospital stay, they were 57% more likely to be readmitted or to die within a year than those who experienced a larger drop.

The study group consisted of 241 patients admitted to Johns Hopkins Hospital in Baltimore for congestive heart failure between June 2006 and April 2007. They underwent the NT-proBNP test at admission, and the patients' doctors then treated their symptoms throughout the hospital stay. Researchers re-tested the protein level at discharge, and patients' treating physicians were not made aware of the level change. Researchers then followed up with hospital records, patient family interviews, and death records over the next 12 months.

Hospitals typically test CHF patients for this protein when they are admitted, but not when they are discharged. Instead, other clinical factors are relied upon to determine whether a patient is healthy enough for discharge. Physicians typically consider a patient's function, heart and lung sounds, and weight loss, among other factors.

Dr. Henry Michtalik, M.D., M.P.H. a clinical fellow in internal medicine and a hospitalist at Johns Hopkins Hospital, and lead researcher on the study, explained that physicians don't typically test NT-proBNP levels at discharge since the number alone may not be significant. A number that could be suspicious for one patient might be insignificant in another.

Facebook icon
LinkedIn icon
Twitter icon