Impact of Significant Hemoglobin Drop Without Bleeding in Patients Undergoing Transcatheter Aortic Valve Replacement

Abstract

Background Hemoglobin (Hgb) drop without bleeding is common among patients undergoing transcatheter aortic valve replacement; however, the clinical implications of significant Hgb drop have not been fully evaluated. Methods and Results Consecutive patients undergoing transcatheter aortic valve replacement at our institution from 2011 to 2021 were retrospectively reviewed. Three groups were assessed: no Hgb drop and no bleed (NoD‐NoB [reference group]), Hgb drop with bleed, and Hgb drop and no bleed (D‐NoB). Hgb drop was defined as ≥3 g/dL decrease from pre‐ to post‐transcatheter aortic valve replacement. Outcomes of interest were in‐hospital death and 1‐year all‐cause mortality. A total of 1851 cases with complete Hgb data were included: NoD‐NoB: n=1579 (85.3%); D‐NoB: n=49 (2.6%); Hgb drop with bleed: n=223 (12.6%). Compared with NoD‐NoB, the D‐NoB group was older (81.1 versus 78.9 years of age) with higher preprocedure Hgb (12.9 versus 11.7 g/dL). In‐hospital death rate was higher among patients with D‐NoB versus NoD‐NoB (4.5% versus 0.8%, P<0.001) and similar to Hgb drop with bleed (4.5% versus 4.1%, P=0.999). Predictors of in‐hospital death were D‐NoB (odds ratio [OR], 3.45 [95% CI, 1.32–8.69]) and transfusion (OR, 10.6 [95% CI, 4.25–28.2]). Landmark survival analysis found that D‐NoB experienced 1‐year mortality rate comparable to NoD‐NoB, whereas Hgb drop with bleed had higher midterm mortality (hazard ratio [HR], 3.2 [95% CI, 1.83–5.73]), and transfusion continued to impact mortality (HR, 2.5 [95% CI, 1.79–3.63]). Conclusions Hgb drop without bleeding is common among patients undergoing transcatheter aortic valve replacement and may represent a higher risk of periprocedural death. Blood transfusion increases short‐ and midterm mortality risk in patients with and without bleeding, supporting a restrictive transfusion strategy.

Keywords

• bleeding
• risk stratification
• TAVR