Is estimated peri-operative glomerular filtration rate associated with post-operative mortality in fractured neck of femur patients?

Abstract

Comparison of renal function in patients who died within 30 days of surgery for hip fractures with surviving patients matched for age, type of surgery, type of anaesthesia and clinical assessment of fitness for surgery. A retrospective case-control study of 80 patients was performed. Pre- and post-operative urea, creatinine, estimated glomerular filtration rate (eGFR), sodium and potassium of 40 patients who died within 30 days post-surgery (cases) were compared with 40 patients who survived matched for age, sex, surgical procedure and pre-operative ASA grade (controls). Statistical analysis involved univariable analysis by paired t-test and logistic conditional regression analysis. Pre- and post-operative mean serum creatinine were significantly higher in patients who died compared with surviving controls (pre-op 108.2 vs. 90.2 micromol/l [p=0.002], post-op 103.9 vs. 87.1 micromol/l [p=0.003]). However, mean creatinine values for both groups were still within normal laboratory reference ranges. Mean serum urea was also higher in patients who died (pre-op 8.81 vs. 6.75 mmol/l [p=0.010] and post-op 9.30 vs. 6.63 mmol/l [p=0.004], respectively). Pre- and post-operative eGFR was significantly lower in those patients who died within 30 days of hip fracture surgery (pre-op 61.75 vs. 68.78 ml/min per 1.73 m(2) [p=0.011], post-op 63.03 vs. 76.03 ml/min per 1.73 m(2) [p=0.009]). The mean eGFR values for survivors and non-survivors fell in the chronic kidney disease mild renal impairment category. There were no significant differences in mean serum sodium or potassium concentration between groups. Stepwise conditional logistic regression analysis, showed that after allowing for creatinine, urea was no longer significantly related to mortality. These results suggest that when allowance is made for demography, fitness for surgery and surgical procedure, subtle reductions in eGFR are still associated with mortality risk. However, these are of questionable clinical use.