Abstract
OBJECTIVE
We aimed to identify factors associated with adverse outcomes in Jehovah’s Witness patients undergoing complex cardiovascular surgery and to validate our preoperative optimization protocol.
METHODS
We retrospectively reviewed 144 Jehovah’s Witnesses who underwent cardiovascular surgery between 1999 and 2014. We excluded 7 salvage cases. The operative procedures included 56 coronary artery bypass graft surgeries, 43 valve procedures, 13 ventricular assist device implantations, 11 heart transplantations, 9 aortic surgeries, and 5 congenital defect repairs. Our preoperative optimization protocol for Jehovah’s Witnesses includes discontinuing antiplatelets and adding iron/vitamin or erythropoietin to achieve a target hemoglobin greater than 12 g/dL. We evaluated the risk factors for postoperative mortality and composite outcomes (mortality, myocardial infarction, stroke, acute kidney injury, heart failure, sternal wound infection), and compared the outcomes of optimized patients with a preoperative hemoglobin level greater than 12 g/dL (n = 93) versus unoptimized patients with a preoperative hemoglobin level less than 12 g/dL (n = 44).
RESULTS
Preoperative and intraoperative demographics in the optimized and unoptimized groups were similar except for preoperative hemoglobin levels, renal dysfunction (optimized = 25/93 [26.9%], unoptimized = 26/44 [59.1%], P < .001), and emergency/urgent cases (optimized = 20/93 [21.5%], unoptimized = 17/44 [38.6%], P = .035). The mean preoperative, intraoperative nadir, and discharge hemoglobin levels of the entire cohort were 12.7 ± 1.7 g/dL, 9.5 ± 2.6 g/dL, and 9.7 ± 1.8 g/dL, respectively. Hospital mortality was 9 of 137 patients (6.6%) (optimized = 2/93 [2.2%], unoptimized = 7/44 [15.9%], P = .002), and composite outcomes were observed in 44 of 137 patients (32.1%) (optimized = 21/93 [22.6%], unoptimized = 22/44 [50.0%], P = .001). The Youden index identified a cutoff value of the preoperative hemoglobin of 11.7 g/dL for mortality (area under curve, 0.719; sensitivity, 77.8%; specificity, 76.0%). Multivariate analysis identified a suboptimal preoperative hemoglobin (<12 g/dL) as the only important independent factor associated with mortality (odds ratio, 5.64; 95% confidence interval, 1.14-42.18) and composite outcomes (odds ratio, 2.49; 95% confidence interval, 1.06-5.88).
CONCLUSIONS
Complex cardiovascular surgery in Jehovah’s Witnesses was associated with acceptable surgical outcomes, especially if they electively completed optimization. Our Jehovah’s Witnesses’ optimization protocol targeting a hemoglobin level greater than 12 g/dL seemed to be effective in reducing adverse events at The University of Chicago Medicine.