Back to Annual Meeting Posters

OBJECTIVE: The aim of this study was to evaluate outcomes associated with red blood transfusion in patients undergoing lower extremity amputation.
METHODS: Using National Surgical Quality Improvement (NSQIP) database from 2005-2007, we identified patients undergoing non-traumatic lower extremity amputation. Patients were categorized into two groups based on those who received intraoperative transfusion and those who did not. Patient demographics, co-morbidities, baseline hemoglobin, perioperative morbidity and mortality were compared between groups using chi-square and t-test as appropriate. Logistic regression analysis was performed to analyze factor independently related to postoperative mortality.
RESULTS: 9,275 patients were identified in the study, 971 (10.4%) had one or more unit packed or whole red blood cell transfusion intraoperative, 8304 (89.6%) patients had no transfusion. Among 971 patients who received transfusion, 143 (15%) had preoperative hemoglobin (hgb)< 8mg/dl, 613 (63%) had hgb 8-10 mg/dl and 215 (22%) had hgb > 10 mg/dl. Mean number of unit transfused was 1.7 (range 1-9 units). Comparing to no-transfusion group, those who received transfusion had lower baseline hemoglobin (9.0 ± 3.3 vs. 10.0 ± 5.7 mg/dl, p <.001), higher number of CHF patients (10.5% vs. 7.5%, p<.001), and patients at increase bleeding risk (32.6% vs. 26.7%, p<.001). After adjusting for baseline risk profile and trigger hemoglobin level, intraoperative transfusion (RR 1.36), congestive heart failure (RR, 3.09), bleeding disorder (RR, 1.26) and having end-stage renal disease (RR, 2.79) were independently related to significant increase in 30-day mortality (Table1). Patient who received transfusion had higher incidence of postoperative complications including sepsis, pneumonia, acute renal failure, incidence of pulmonary embolism and deep venous thrombosis (Table2)
CONCLUSION: Regardless of preoperative hemoglobin levels, intraoperative blood transfusion in patients undergoing lower extremity amputation is related to increase risk of postoperative mortality and morbidity. Further studies are required to identify criteria for transfusion and alternative treatment modality.
Table 1. Independent predictors of postoperative mortality

Variables	Relative risk	Confidence interval (95%)	P- value
Transfusion	1.36	1.07-1.72	.011
Smoking	0.98	.80-1.21	.887
CHF	3.09	2.50-3.81	<.001
ESRD	2.79	2.36-3.22	<.001
Bleeding disorder	1.26	1.06-1.49	.007

Table 2. Postoperative complications

Complications	No Transfusion N=8304	Transfusion N=971	P Value
Superficial surgical site infection	415 (5%)	37 (3.8%)	0.115
Deep Incisional SSI	255 (3.1%)	40 (4.1%)	0.082
Wound disruption	147 (1.85)	14 (1.4%)	0.518
Pneumonia	301 (3.6%)	64 (6.6%)	<.001
Reintubation	315 (3.8%)	52 (5.4%)	0.023
Pulmonary embolism	39 (0.5%)	16 (1.6%)	<.001
Failure to wean	317 (3.8%)	94 (9.7%)	<.001
ARF	121 (1.5%)	29 (3.0%)	0.001
Progressive renal failure	78 (0.9%)	10 (1.0%)	0.727
UTI	369 (4.4%)	55 (5.7%)	0.088
Cardiac arrest	187 (2.3%)	29 (3%)	0.176
MI	99 (1.2%)	13 (1.3%)	0.642
DVT	104 (1.3)	25 (2.6%)	0.002
Return to OR	1630 (19.6%)	207 (21.3%)	0.217
Sepsis	424 (5.1%)	70 (7.2%)	0.008
Septic Shock	279 (3.4%)	54 (5.6%)	0.001
Mortality	681 (8.2%)	100 (10.3%)	0.028