Predicting mortality rate in elderly patients operated for hip fracture using red blood cell distribution width, neutrophil-to-lymphocyte ratio, and Nottingham Hip Fracture Score

Abstract

Objectives: In this study, we aimed to compare the neutrophil-tolymphocyte ratio (NLR), red blood cell distribution width (RDW), and Nottingham Hip Fracture Score (NHFS) according to one-year mortality estimation after hip fracture surgery in elderly.

Patients and methods: Between January 2015 and December 2019, a total of 190 elderly patients (63 males, 127 females; mean age: 82.8±6.1 years; range, 70 to 98 years) who were diagnosed with collum femoris fractures treated with hemiarthroplasty were included. The cohort was divided into two groups with NHFS ≤4 and >4 as the low- and high-risk patients, respectively and one-year mortality was assessed for both groups. The RDW was evaluated with blood values sampled on the day of admission. A cut-off of 14.5% was considered for the RDW values. The NLR values calculated on admission (NLR-D0) and postoperative Day 5 (NLR-D5) were considered the primary outcome measures.

Results: A total of 46 patients (24.2%) developed any type of complication. The NLR values higher than 5 on Days 0 and 5 were more frequently seen in the complicated patients (p=0.0016 and p<0.001). There were significantly more patients with higher RDW values (>14.5%) in the complicated group (p<0.001). The median NHFS and the rate of patients with NHFS >4 were significantly higher in the complicated patients (p<0.001 for both). The NHFS value higher than 4 had a sensitivity of 87.7% and specificity of 84.0% in predicting mortality (area under the curve [AUC]=0.910, 95% confidence interval [CI]: 0.860-0.947, p<0.001). Estimation of mortality using an RDW cut-off value of >14.5 showed 87.7% and 80.0% sensitivity and specificity, respectively (95% CI: 0.789-0.904, p<0.001). The AUC of the NLR Day 5 using a cut-off value of >6.8 was 0.953 for the prediction of mortality (95% CI: 0.912-0.978, p<0.001).

Conclusion: Age, NLR Day 5 (>5), RDW (>14.5%) and NHFS (>4) were strongly associated with mortality prediction. The NHSF and RDW values had the highest and similar sensitivity merit, while the highest specificity was in NLR-D5. Therefore, NLR, RDW and NHFS values can be used to classify risk factors in estimating oneyear mortality rates in elderly patients operated for hip fractures. A multidisciplinary approach should be standardized in determining the risk factors before treatment in patients with hip fractures and in planning appropriate treatment for this risk.

Figure 1. Graphical representation of logistic regression analysis showing the independent risk factors for mortality. OR: Odds ratio; CI: Confidence interval; NLR D5: Neutrophil-to-lymphocyte ratio on Day 5, RDW: Red cell distribution width; NHFS: Nottingham Hip Fracture Score.