Usefulness of Estimated Plasma Volume at Postdischarge Follow-Up to Predict Recurrent Events in Patients With Heart Failure

Abstract

Hospital readmission for patients admitted with heart failure is a persistent problem. Better identification of patients at high risk of readmission for volume overload could have clinical implications. We evaluated estimated plasma volume (ePV), a marker of congestion, to predict readmission for patients seen early after discharge for heart failure. We identified patients hospitalized with a primary heart failure diagnosis and were then seen in a postdischarge clinic. We assessed clinical factors, ePV (derived from hemoglobin and hematocrit), and B-type natriuretic peptide (BNP). The primary outcome was death or readmission for heart failure within 90 days of discharge. We identified 218 patients, of whom 23% experienced the primary outcome. No clinical variables at time of admission were different between those who did and did not experience the primary outcome, nor were BNP (1,581 vs 1,267 pg/ml, p = 0.33) or ePV (6.00 vs 5.80, p = 0.36). At clinic follow-up, both BNP (1,164 vs 636, p = 0.002) and ePV (6.18 vs 5.58, p = 0.02) were higher in those with subsequent events. Kaplan-Meier survival analysis showed that the lowest tertile of ePV had significantly lower incidence of the primary outcome than the other 2 tertiles (12% vs 29% and 27%, p = 0.02). Estimated plasma volume remained independently predictive of outcomes after controlling for BNP (p <0.05). In conclusion, EPV may be predictive of death or hospital readmission in heart failure patients seen soon after discharge, independent of BNP. Its potential warrants future prospective research evaluating its utility in larger heart failure cohorts.