Association between in-hospital haemoglobin drop with or without overt bleeding and long-term clinical outcomes after percutaneous coronary intervention

Abstract Background Periprocedural bleeding complications remain a significant issue in patients undergoing percutaneous coronary intervention (PCI). Although most bleeding complications are defined by the degree of haemoglobin drop, prognostic implications of haemoglobin drop without clinically overt bleeding remain unclear. Purpose We aimed to investigate the association between haemoglobin drop after PCI and long-term prognosis, stratified by the occurrence of clinically overt bleeding. Methods Consecutive patients discharged after PCI between September 2008 to June 2018 from Keio Cardiovascular PCI multicenter prospective registry in Japan were analyzed. Patients were categorized into four groups: those with overt bleeding with haemoglobin drop, those with overt bleeding without haemoglobin drop, those without overt bleeding with haemoglobin drop, and those without overt bleeding nor haemoglobin drop. The pre- and post-PCI haemoglobin were defined as the lowest value within 72 hours before and after PCI. We defined the haemoglobin drop as the value obtained by subtracting the pre-PCI haemoglobin value from the post-PCI haemoglobin value equal to or greater than 3.0 g/dL. The primary outcome was the 2-year composite of major adverse cardiovascular events (MACE): all-cause mortality, acute coronary syndrome, heart failure hospitalization, or stroke. Cox proportional models were used to assess the association between the haemoglobin drop and the clinical outcomes. Results Among the 6,013 patients [mean age 68 years and 78% male], 52 (0.9%) had overt bleeding with haemoglobin drop, 45 (0.7%) had overt bleeding without haemoglobin drop, 497 (8.3%) had haemoglobin drop without overt bleeding, and 5,419 (90.1%) had neither. At the two-year follow-up, patients with overt bleeding, both with and without haemoglobin drop had higher rates of MACE compared to those without overt bleeding, with and without haemoglobin drop (28.8% vs. 20.0% vs. 11.8% vs. 10.5%, respectively). Patients with overt bleeding with haemoglobin drop were associated with increased risks of MACE compare to those without overt bleeding without haemoglobin drop (HR: 2.19, 95% CI: 1.26-3.81, P=0.01). Conversely, patients without overt bleeding with haemoglobin drop and patients with overt bleeding without haemoglobin drop were not associated with increased risk of MACE (HR:1.19, 95% CI: 0.88-1.63, P=0.26, and HR: 0.62, 95% CI: 0.29-1.31, P=0.21, respectively). The spline curve showed a significant correlation between haemoglobin drop and MACE in patients with overt bleeding. However, for patients without overt bleeding, the correlation was attenuated. Conclusions Patients with haemoglobin drop with overt bleeding were significantly associated with long-term prognosis, but no association was observed among those with haemoglobin drop without overt bleeding. Our findings prompt a reevaluation of the current criteria for classifying bleeding complications.Figure