Background: Autograft harvesting for spine arthrodesis has been associated with longer operative times and increased blood loss. Allograft compared to autograft in spinal fusions has not been studied in a multicenter cohort.
Methods: Patients enrolled in the ACS-NSQIP registry between 2012 and 2013 who underwent cervical or lumbar spinal fusion with either allograft or autograft through a separate incision were included for analysis. The primary outcomes of interest were operative time, length of stay, blood transfusion, and surgical site infection (SSI).
Results: A total of 6790 and 6718 patients received a cervical or lumbar spinal fusion, respectively. On unadjusted analysis in both cervical and lumbar cohorts, autograft was associated with increased rates of blood transfusion (cervical: 2.9% vs. 1.0%, P<0.001; and lumbar: 21.0% vs. 15.7%, P<0.001) and increased operative time (cervical: 167 vs. 128 minutes, P<0.001; and lumbar: 226 vs. 204 minutes, P<0.001) relative to allograft. On multivariable analysis in both the cervical and lumbar cohorts, autograft was associated with increased odds of blood transfusion (cervical: OR=2.3, 95% CI: 1.0-5.1; and lumbar: OR=1.3, 95% CI: 1.1-1.6) and longer operative times (cervical: 27.8 minutes, 95% CI: 20.7-35.0; and lumbar: 25.4 minutes, 95% CI: 17.7-33.1) relative to allograft. Autograft was not associated with either length of stay or SSI.
Conclusions: In a multicenter cohort of patients undergoing cervical or lumbar spinal fusion, autograft was associated with increased rates of blood transfusion and increased operative time relative to allograft.