Abstract 14653: Periprocedural Transfusion is Associated With a Dramatic Increase in Morbidity and Mortality in Patients Undergoing Percutaneous Vascular Interventions: Insights From the The Bmc2-vic
Objective: Blood transfusions are common among patients undergoing vascular interventions. The impact of transfusion on patients undergoing vascular percutaneous interventions has been poorly defined.
Methods: We examined data from a large multicenter (n=44 hospitals) quality improvement (QI) vascular interventional registry of all patients undergoing percutaneous vascular interventional (PVI) procedures for peripheral arterial disease (PAD) between 2010 and 2013. Multiple logistic regression modeling was used to identify predictors of, and effect of transfusion on outcomes.
Results: Of the 18,127 patients enrolled in the registry, 4.1% received a blood transfusion. The median hemoglobin of those transfused was 7.7 mg/dL (IQR =7.2 - 8.4). Transfusion rates varied from 0 to 14% amongst the hospitals in the registry. From a logistic regression model for transfusion (C-statistic = 0.84 and Hosmer-Lemeshow P-value = 0.653), significant factors (all p<.01) that were independently associated with transfusion included low creatinine clearance (OR= 1.7; 95% CI:1.5 - 2.1), pre-procedural anemia (4.7; 3.8 - 5.9), CHF (1.5; 1.3 - 1.8), COPD (1.3; 1.1 - 1.5), CVD or TIA (1.3; 1.1 - 1.5), warfarin use (1.4; 1.1 - 1.8), critical limb ischemia (1.9; 1.5 - 2.3), urgent procedure (2.7; 2.3 - 3.3), and emergent procedure (8.3; 5.7 - 12.1). As compared with non-transfused patients, MI, death, major cardiac adverse events (MACE), and vascular complications were all significantly elevated (p<.001). Adjusting for patient demographics, disease states, and peri-procedural medications, transfusion was a significant risk factor for increased risk with MI, (OR = 25; 95% CI = 13.7 - 45.8), death (12.7; 7.9 - 20.4), MACE (15.3; 10.6 - 22.1), and vascular access complications (49; 37 - 67), respectively (all p < .001). Over 4 years, a focused performance improvement program was associated with a drop from observed to expected transfusion ratio from 1.2 (2010) to .89 (2013) (p < .001).
Conclusion: In a large PVI registry, post procedure transfusion was common and highly correlated with a specific set of clinical risk factors and with major morbidity and mortality. However, using a focused QI program, a significant reduction in transfusion is possible.
Author Disclosures: P. Henke: None. Y. June Park: None. P. Bove: None. R. Cuff: None. A. Kazmers: None. S. Hans: None. H. Gurm: None. P.M. Grossman: None.
- © 2014 by American Heart Association, Inc.