Should asymptomatic patients discharged with lower hemoglobin expect worse outcomes after valve surgery?

Abstract

Blood transfusion in cardiac surgery patients is associated with increased morbidity and cost. The decision to transfuse patients after surgery varies but is often based on low hemoglobin (Hgb) levels, regardless of symptom status. This study examined whether asymptomatic patients discharged with lower Hgb levels had increased risk for perioperative complications and 1-year mortality. Between 2008 and mid-2014, a total of 1107 valve-only procedures were performed. Patients discharged alive with complete data (N = 1044) were divided into 2 groups with discharge Hgb levels of ≤8 g/dL (n = 153) or >8 g/dL (n = 891). Propensity score matching was conducted between Hgb groups, resulting in 152 patient pairs. In multivariate analyses, discharge Hgb level did not predict 30-day mortality (odds ratio [OR] = 1.01, P = .991), 1-year survival (hazard ratio [HR] = 0.87, P = .34), or readmission <30 days (OR = 0.92, P = .31). Furthermore, after propensity score matching, no differences were found between groups with Hgb levels ≤8 versus >8 g/dL in 30-day mortality (0% vs 0.7%, P > .99) or readmissions (14% vs 16%, P = .52). Cumulative 1-year survival was similar between matched groups with discharge Hgb level of ≤8 versus >8 g/dL (89.3% vs 91.4%, P = .67). Matched groups with Hgb level ≤8 versus >8 g/dL had similar physical (28% vs 18% increase; P = .27) and mental (7% vs 6% increase; P = .94) health-related quality of life (HRQL) improvements at 6 months. Asymptomatic patients discharged with lower Hgb levels did not manifest inferior outcomes, including perioperative morbidity/mortality, readmission <30 days, HRQL, and 1-year survival. The practice of blood transfusion to correct lower Hgb levels in asymptomatic patients should be eliminated, as it may be associated with increased morbidity without apparent clinical benefits after valve surgery.