Comparison of true blood loss between short and long cephalomedullary nail fixation of geriatric hip fractures, a retrospective cohort study.

Abstract

Existing literature is discrepant on the differences in blood loss and need for transfusion between short and long cephalomedullary nails used for extracapsular geriatric hip fractures. However, prior studies used the inaccurate estimated rather than the more accurate 'calculated' blood loss based on hematocrit dilution (Gibon in IO 37:735-739, 2013, Mercuriali in CMRO 13:465-478, 1996). This study sought to clarify whether use of short nails is associated with clinically meaningful reductions in calculated blood loss and resultant need for transfusion. A retrospective cohort study using bivariate and propensity score-weighted linear regression analyses was conducted examining 1442 geriatric (ages 60-105) patients undergoing cephalomedullary fixation of extracapsular hip fractures over 10years at two trauma centers. Implant dimensions, pre and postoperative laboratory values, preoperative medications, and comorbidities were recorded. Two groups were compared based on nail length (greater or less than 235mm). Short nails were associated with a 26% reduction in calculated blood loss (95% confidence interval: 17-35%; p < 10-14) and a 24-min (36%) reduction in mean operative time (95% confidence interval: 21-26min; p < 10-71). The absolute reduction in transfusion risk was 21% (95% confidence interval: 16-26%; p < 10-13) yielding a number needed to treat of 4.8 (95% confidence interval: 3.9-6.4) with short nails to prevent one transfusion. No difference in reoperation, periprosthetic fracture, or mortality was noted between groups. Use of short compared to long cephalomedullary nails for geriatric extracapsular hip fractures confers reduced blood loss, need for transfusion, and operative time without a difference in complications.