Abstract
BACKGROUND
Hip fractures (HFs), which are common in the geriatric population and have a high mortality rate, are among the pathologies frequently encountered in emergency departments (EDs). The aim of this study was to investigate the prognostic value of red cell distribution width (RDW) and modified 5-item frailty index (mFI-5) scores in the prediction of short-term mortality and intensive care unit (ICU) admission in patients with HFs.
METHODS
Patients who presented to the ED due to HF between January 01, 2019, and December 31, 2023, were evaluated. Patients who underwent surgery for femoral neck, intertrochanteric, or subtrochanteric fractures were included in the study. Baseline demographics (age and sex), comorbidities, ICU admission status, and in-hospital and 30-day mortality rates were recorded. Univariate and multivariable logistic regression analyses were performed to identify independent predictors of ICU admission and 30-day mortality.
RESULTS
A total of 413 patients undergoing surgical treatment for HFs were included. The mean age of the patients was 80±11 years, and 271 (65.6%) were female. ICU admission was required for 43 patients (10%), while in-hospital and 30-day mortality rates were 7% (n=30) and 9.9% (n=41), respectively. Multivariable logistic regression analysis identified RDW as an independent predictor of both ICU admission (odds ratio [OR] 1.22, 95% confidence interval [CI] 1.05–1.41; p=0.009) and 30-day mortality (OR 1.20, 95% CI 1.03–1.39; p=0.016). Atrial fibrillation (AF) and Alzheimer’s disease were also identified as independent predictors of ICU admission, while AF and age independently predicted 30-day mortality. The mFI-5 score did not reach statistical significance in univariate analysis for either outcome. A predictive model incorporating RDW and other covariates achieved area under the curve values of 0.74 for ICU admission and 0.70 for 30-day mortality.
CONCLUSION
RDW emerged as an independent predictor of ICU admission and 30-day mortality in surgically managed HF patients, with improved discriminatory performance when integrated into a multiparametric model. Advanced age, Alzheimer’s disease, and AF further contribute to increased risk and should be closely monitored in clinical management. These findings support the incorporation of RDW into routine emergency risk stratification for this vulnerable population.
Keywords: 30-day mortality, comorbidities, intensive care unit admission, monocyte‒lymphocyte ratio, neutrophil‒lymphocyte ratio, systemic immune‒inflammation index, systemic inflammation response index.ß
Abstract
AMAÇ
Yaşlı bireylerde yaygın olarak görülen ve yüksek mortaliteye sahip kalça kırıkları, acil servislerde sıklıkla karşılaşılan patolojiler arasındadır. Bu çalışmanın amacı, kalça kırığı olan hastalarda kısa dönem mortalite ve yoğun bakım ünitesine (YBÜ) yatış gereksinimlerini tahmin etmede eritrosit dağılım genişliği (RDW) ve modifiye 5-maddelik kırılganlık indeksi (mFI-5) skorlarının prognostik değerini araştırmaktır.
GEREÇ VE YÖNTEM
1 Ocak 2019 ile 31 Aralık 2023 tarihleri arasında kalça kırığı nedeniyle acil servise başvuran hastalar değerlendirildi. Femur boyun, intertrokanterik veya subtrokanterik kırıklar nedeniyle ameliyat geçiren hastalar çalışmaya dahil edildi. Temel demografik veriler (yaş ve cinsiyet), eşlik eden hastalıklar, YBÜ’ye yatış durumu, hastane-içi ve 30 günlük mortalite oranları kaydedildi. YBÜ’ye yatış ve 30-günlük mortalitenin bağımsız belirleyicilerini tespit etmek için tek değişkenli ve çok değişkenli lojistik regresyon analizleri yapıldı.
BULGULAR
Kalça kırığı nedeniyle ameliyat yapılan toplam 413 hasta incelenmiştir. Hastaların ortalama yaşı 80±11 yıl olup, 271'i (%65.6) kadındı. Kırk üç hasta (%10) YBÜ’ye yatırılırken, hastane-içi ve 30-günlük mortalite oranları sırasıyla %7 (n=30) ve %9.9 (n=41) idi. Çok değişkenli lojistik regresyon analizi sonucunda RDW hem YBÜ’ye yatış (OR 1.22, %95 CI 1.05–1.41; p=0.009) hem de 30-günlük mortalite oranı (OR 1.20, %95 CI 1.03–1.39; p=0.016) için bağımsız bir öngörücü olarak belirlendi. Atriyal fibrilasyon ve Alzheimer hastalığı YBÜ’ye yatışın bağımsız öngörücüleri olarak belirlenirken, atriyal fibrilasyon ve yaş ise 30-günlük mortaliteyi bağımsız olarak öngördü. mFI-5 skoru, tek değişkenli analizde her iki sonuç için de istatistiksel olarak anlamlılığa ulaşmadı. RDW ve diğer kovaryatları içeren bir tahmin modeli, YBÜ’ye yatış için 0.74 ve 30 günlük mortalite için 0,70 AUC değerlerine ulaştı.
SONUÇ
RDW cerrahi olarak tedavi edilen kalça kırığı hastalarında YBÜ’ye yatış ve 30 günlük mortalitenin bağımsız olarak öngörücüsü olarak belirlendi ve çok parametreli bir modele entegre edildiğinde ayırt edici performansı iyileşti. İleri yaş, Alzheimer hastalığı ve atriyal fibrilasyon risk artışına daha fazla katkıda bulunur ve klinik yönetim sürecinde yakından takip edilmelidir. Bu bulgular, bu hasta grubu için RDW'nin rutin acil risk sınıflandırmasına dahil edilmesini desteklemektedir.
Anahtar sözcükler: 30 günlük mortalite, eşlik eden hastalıklar, monosit-lenfosit oranı, nötrofil-lenfosit oranı, sistemik immün-enflamasyon indeksi, sistemik inflamasyon yanıt indeksi, yoğun bakım yatışı
INTRODUCTION
Hip fractures (HFs) are among the most common causes of mortality and morbidity following trauma, particularly in geriatric individuals. A meta-analysis reported a median (Q1‒Q3) 1-month mortality rate of 8% (6.5‒9.6%) in HF patients.[1] This high early mortality rate is due to major surgical interventions, comorbidities, and diminished physiological reserves in geriatric patients rather than the severity of the trauma.[2] Such high risk levels emphasize the necessity of biomarkers and scoring systems that can optimize patient care by identifying intensive care unit (ICU) requirements and mortality risk at an early stage.
Numerous studies have evaluated various biomarkers and indices aimed at predicting mortality and morbidity in patients with HF.[3-5] Among these, red cell distribution width (RDW), a standardized component of the complete blood count (CBC), reflects the variability in erythrocyte volume. RDW levels tend to rise under conditions of inflammation and oxidative stress, and have been established as prognostic markers in many conditions, such as sepsis, congestive heart failure (CHF), chronic obstructive pulmonary disease (COPD), stroke, and oncological diseases. Beyond RDW, the literature extensively discusses the neutrophil/lymphocyte ratio (NLR), monocyte/lymphocyte ratio (MLR), systemic ımmuno-ınflammation ındex (SII), and systemic ınflammatory response ındex (SIRI), all of which are derived from platelet, neutrophil, lymphocyte, and monocyte values.[4-6]
Frailty is defined as a state of increased vulnerability resulting from reduced physiological reserve and an impaired ability to adapt to stressors with advancing age. The 5-factor modified frailty index (mFI-5) is a simplified scoring tool derived from a patient’s comorbidities, developed to predict post-operative complications and mortality across various surgical populations.[7] Despite its simplicity, its high predictive utility has led to its widespread adoption in clinical and research settings.
This study aimed to evaluate the prognostic value of RDW, CBC-derived inflammatory indices, and the mFI-5 score in predicting short-term mortality and ICU admission requirements in patients who presented to the emergency department (ED) and underwent surgery for HF.
MATERIALS AND METHODS
In this single-center, retrospective observational study, patients presenting to the ED with HF and undergoing surgery between January 01, 2019, and December 31, 2023, were evaluated. A total of 413 patients aged 18 years and older who underwent surgery for femoral neck, intertrochanteric, or subtrochanteric fractures, and for whom complete medical records were available, were included in the study. The exclusion criteria were as follows: Hematological malignancy, active infection, or known inflammatory disorders. The patient selection process is summarized in Figure 1.
Figure 1.
Flow diagram of patient selection. Of 434 patients assessed for eligibility, 21 were excluded due to incomplete medical or radiological records (n=7), active infection or inflammatory disease (n=8), or known hematological malignancy (n=6). A total of 413 patients were included in the final analysis.
During the data collection process, patient demographics, such as age and sex, comorbidities, ICU admission status, and in-hospital and 30-day mortality rates, were recorded. ICU admission decisions were made by attending anesthesiologists based on standard clinical criteria, independent of RDW or mFI-5 values, which were not utilized as ICU admission indicators at our institution. Admission to the ICU was based on the patient’s clinical status and could occur either pre-operatively or post-operatively. Intertrochanteric femoral fractures were categorized as either stable or unstable according to the Evans classification.[8] For analytical purposes, surgical techniques were grouped into three categories: İntramedullary nailing (ıntramedullary nailing, including proximal femoral nailing [PFN]), open reduction and internal fixation (including cannulated screw fixation and plate fixation), and arthroplasty (including partial and total hip arthroplasty). Laboratory parameters were obtained from blood samples collected during initial ED presentation. As this was a retrospective study, the precise timing of sampling could not be systematically verified; all values were derived from the first recorded CBC at admission. All CBC analyses were performed using the Mindray BC-760 automated hematology analyzer (Mindray, Shenzhen, China) with an institutional reference range for RDW of 11–16%.
This study received ethical approval from the Non-Interventional Clinical Research Ethics Committee of the Bilecik Şeyh Edebali University Faculty of Medicine (approval no. 9/4, dated November 06, 2024).
Calculation of Indices
Laboratory parameters
The patient’s white blood cell, RDW, platelet, neutrophil, lymphocyte, and monocyte values were obtained from the CBC. Using these values, the NLR, MLR, SII, and SIRI were calculated as follows:
mFI-5
The mFI-5 score was calculated by assigning one point for each of the following five comorbidities: Hypertension requiring pharmacological treatment, diabetes mellitus, COPD, CHF, and non-independent (partially or completely dependent) functional status.[7] These points were summed to obtain a total mFI-5 score ranging from 0 to 5, with higher scores indicating greater frailty.
Statistical Analysis
Patient characteristics were summarized using descriptive statistics. Differences in categorical variables (such as fracture side, fracture type, mFI-5 score group, admission status, mortality variables presented in Table 1, and comorbidities in Table 2) between groups were assessed using Chi-square tests, and reported as counts (percentages). The normality of distribution for continuous variables (such as age and laboratory test measurements) was evaluated using the Shapiro–Wilk test. The Wilcoxon rank-sum test was used for non-normally distributed variables, while the student t-test was used to compare continuous variables with a normal distribution. Continuous numerical variables were expressed as the means±standard deviations for normally distributed data, and as medians (interquartile range) for non-normally distributed data. Univariate and multivariable logistic regression analyses were performed to identify independent factors associated with ICU admission and 30-day mortality. Variables that did not reach statistical significance in univariate analysis were not entered into multivariable modeling. Odds ratios (OR) and corresponding 95% confidence intervals were presented. Receiver operating characteristic (ROC) analyses derived from the logistic regression models were performed using the pROC package.[9] Statistical significance was set at P<0.05. Statistical analyses were conducted with the IBM Statistical Package for the Social Sciences Statistics for Windows, Version 22.0 (IBM Corp., Armonk, NY, USA), and R version 3.6.1.[10]
Table 1.
Baseline demographic and clinical characteristics of the study population by sex
| Female (n=271) | Male (n=142) | Total (n=413) | p | |
|---|---|---|---|---|
| Age | 81±10 | 78±12 | 80±11 | NS |
| Age ≥65 years | 261 (96) | 123 (87) | 384 (93) | 0.001 |
| Fracture side | ||||
| Left | 135 (50) | 73 (51) | 208 (50.4) | NS |
| Right | 136 (50) | 69 (49) | 205 (49.6) | |
| Fracture type | ||||
| Femoral neck | 88 (32) | 44 (31) | 132 (32) | NS |
| Unstable intertrochanteric | 107 (40) | 50 (35) | 157 (38) | |
| Stable intertrochanteric | 62 (23) | 39 (27) | 101 (24.5) | |
| Subtrochanteric | 11 (4) | 8 (6) | 19 (4.5) | |
| Reverse obliquity intertro-chanteric | 3 (1) | 1 (1) | 4 (1) | |
| Surgery | ||||
| Intramedullary nailing | 1 (0.4) | 2 (1) | 3 (0.7) | NS |
| Cannulated screw fixation | 3 (1) | 3 (2) | 6 (1.4) | |
| Partial hip arthroplasty | 100 (37) | 48 (34) | 148 (36) | |
| Proximal femoral nailing | 162 (60) | 85 (60) | 247 (60) | |
| Plate fixation | 1 (0.4) | - | 1 (0.2) | |
| Total hip arthroplasty | 4 (1.2) | 4 (3) | 8 (1.9) | |
| ASA score* | ||||
| 1 | 10 (3.7) | 5 (3.5) | 15 (3.6) | NS |
| 2 | 55 (20.4) | 27 (19) | 82 (19.9) | |
| 3 | 180 (66.7) | 90 (63.4) | 270 (65.5) | |
| 4 | 25 (9.3) | 20 (14.1) | 45 (10.9) | |
| mFI-5 score | ||||
| 0 | 38 (14) | 27 (19) | 65 (15.7) | NS |
| 1 | 85 (31.4) | 51 (36) | 136 (32.9) | |
| 2 | 102 (37.6) | 47 (33) | 149 (36.1) | |
| 3 | 39 (14.4) | 14 (10) | 53 (12.8) | |
| 4 | 7 (2.6) | 3 (2) | 10 (2.4) | |
| 5 | 0 (0) | 0 (0) | 0 (0) | |
| Admission status | ||||
| Ward | 243 (90) | 127 (89) | 370 (90) | NS |
| ICU | 28 (10) | 15 (11) | 43 (10) | |
| In-hospital mortality | ||||
| Discharged | 253 (93) | 130 (92) | 383 (93) | NS |
| Deceased | 18 (7) | 12 (8) | 30 (7) | |
| 7-day mortality | ||||
| Alive | 267(98.5) | 130 (91.5) | 397 (96.1) | 0.001 |
| Deceased | 4 (1.5) | 12 (8.5) | 16 (3.9) | |
| 30-day mortality | ||||
| Alive | 249 (91.9) | 123 (86.6) | 372 (90.1) | NS |
| Deceased | 22 (8.1) | 19 (13.4) | 41 (9.9) |
Data reported as number (%) or mean±standard deviation. Statistical significance was defined as p≤0.05. *ASA score was not available for one patient. mFI-5: Modified 5-item frailty index,. ICU: Intensive care unit, ASA: American society of anesthesiologists, NS: Not significant (p>0.05)
Table 2.
Associations between comorbidities and ICU admission and 30-day mortality
| Total n (%) | ICU admission | p | 30-day mortality | p | |||
|---|---|---|---|---|---|---|---|
| Ward (n=370) | ICU (n=43) | Alive (n=372) | Deceased (n=41) | ||||
| Presence of comorbidity | 376 (91) | 333 (90) | 43 (100) | 0.02 | 337 (91) | 39 (95) | NS |
| Hypertension | 265 (64) | 240 (65) | 25 (58) | NS | 242 (65) | 23 (56) | NS |
| Diabetes melli-tus | 117 (28) | 107 (29) | 10 (23) | NS | 106 (29) | 11 (27) | NS |
| Congestive heart failure | 46 (11) | 37 (10) | 9 (21) | 0.04 | 42 (11) | 4 (10) | NS |
| COPD | 75 (18) | 62 (17) | 13 (30) | 0.03 | 63 (17) | 12 (29) | NS |
| Alzheimer’s disease | 90 (22) | 75 (20) | 15 (35) | 0.03 | 78 (21) | 12 (29) | NS |
| History of stroke | 45 (11) | 40 (11) | 5 (12) | NS | 38 (10) | 7 (17) | NS |
| Chronic kid-ney disease | 29 (7) | 23 (6) | 6 (14) | NS | 24 (7) | 5 (12) | NS |
| Atrial fibrilla-tion | 55 (13) | 42 (11) | 13 (30) | 0.001 | 43 (12) | 12 (29) | 0.002 |
| Coronary ar-tery disease | 81 (20) | 71 (19) | 10 (23) | NS | 70 (19) | 11 (27) | NS |
Data reported as number (%). Statistical significance was defined as p≤0.05. ICU: Inten-sive care unit, COPD: Chronic obstructive pulmonary disease, NS: Not significant (p>0.05).
RESULTS
In this study, of the 434 patients undergoing surgery for HF over a 5-year period, 413 met the inclusion criteria and were included in the final analysis. The mean age of the patients was 80±11 years, 384 (93%) were aged 65 years or older, and 271 (65.6%) were female. The most common fracture type was intertrochanteric fracture (n=258, 62.5%), comprising 157 (38%) unstable and 101 (24.5%) stable patterns. Femoral neck fractures accounted for 32% (n=132) of the total cases.
Regarding surgical procedures, PFN was the most frequently performed technique (n=247, 60%), followed by partial hip replacement in 148 patients (36%). A total of 43 patients (10%) required ICU admission, either pre-operatively or post-operatively. The mortality rates were as follows: 3.9% (n=16) within the first 7 days, 7% (n=30) during hospitalization, and 9.9% (n=41) at the 30-day follow-up (Table 1).
Analysis of comorbidities in relation to ICU admission and 30-day mortality revealed that all patients admitted to the ICU, and 95% of those who died within 30 days had at least one comorbidity. When evaluating the prognostic impact of specific conditions, significantly higher rates of ICU admission and 30-day mortality were observed in patients with atrial fibrillation (AF) (Table 2).
Among the clinical and perioperative variables assessed, including fracture type, surgical technique, American Society of Anesthesiologists score, and anesthesia type, none were significantly associated with ICU admission or 30-day mortality. Age was significantly associated with 30-day mortality (84±7 vs. 79±11 years; p=0.003) but not with ICU admission. Of the laboratory parameters evaluated, RDW was the only parameter to reach statistical significance, with higher values observed in patients requiring ICU admission (15.6±2.4 vs. 14.5±1.8; p=0.003) and in non-survivors at 30 days (15.5±2.7 vs. 14.5±1.8; p=0.03). No other blood values or derived inflammatory indices reached statistical significance for either outcome (Table 3).
Table 3.
Age, fracture type, surgical technique, ASA score, anesthesia, and laboratory findings in relation to ICU admission and 30-day mortality
| ICU admission | p | 30-day mortality | p | |||
|---|---|---|---|---|---|---|
| Ward (n=370) | ICU (n=43) | Alive (n=372) | Deceased (n=41) | |||
| Age | 79±11 | 83±9 | NS | 79±11 | 84±7 | 0.003 |
| Fracture type (%) | ||||||
| Femoral neck | 120 (32) | 12 (28) | NS | 120 (32) | 12 (29) | NS |
| Unstable intertro-chanteric | 141 (38) | 16 (37) | 139 (37) | 18 (44) | ||
| Stable intertrochan-teric | 88 (24) | 13 (30) | 94 (25) | 7 (17) | ||
| Subtrochanteric | 18 (5) | 1 (2) | 16 (4) | 3 (7) | ||
| Reverse obliquity intertrochanteric | 3 (1) | 1 (2) | 3 (1) | 1 (2) | ||
| Surgery (%) | ||||||
| IMN | 222 (60) | 28 (65) | NS | 228 (61) | 22 (54) | NS |
| ORIF | 7 (2) | 0 (0) | 7 (2) | 0 (0) | ||
| Arthroplasty | 141 (38) | 15 (35) | 137 (37) | 19 (46) | ||
| ASA score* (%) | ||||||
| 1 | 14 (4) | 1 (2) | NS | 13 (4) | 2 (5) | NS |
| 2 | 77(21) | 5 (12) | 76 (20) | 6 (15) | ||
| 3 | 242(66) | 28 (65) | 243 (65) | 27 (66) | ||
| 4 | 36(10) | 9 (21) | 39 (11) | 6 (15) | ||
| Anesthesia (%) | ||||||
| Femoral block | 2 (1) | 1 (2) | NS | 2 (1) | 1 (2) | NS |
| General anesthesia | 26(7) | 4 (9) | 27 (7) | 3 (7) | ||
| Spinal anesthesia | 342 (92) | 38 (88) | 343 (92) | 37 (90) | ||
| Laboratory tests | ||||||
| WBC (×109/L) | 11±3.8 | 11±4.2 | NS | 11±3.7 | 11.1±4.5 | NS |
| Plt (×109/L) | 232±76 | 237±97 | NS | 233±76 | 233±96 | NS |
| RDW (%) | 14.5±1.8 | 15.6±2.4 | 0.003 | 14.5±1.8 | 15.5±2.7 | 0.03 |
| Lymphocyte count | 1.3±0.7 | 1.3±0.9 | NS | 1.3±0.7 | 1.4±1 | NS |
| Neutrophil count | 9±3.7 | 9±4.2 | NS | 9±3.7 | 9±4.4 | NS |
| Monocyte count | 0.57±0.22 | 0.62±0.36 | NS | 0.57±0.24 | 0.60±0.26 | NS |
| NLR | 9.2±7.7 | 9.6±9 | NS | 9.3±7.9 | 9±7.5 | NS |
| MLR | 0.54±0.37 | 0.59±0.43 | NS | 0.55±0.38 | 0.57±0.39 | NS |
| SII | 2185±2120 | 2280±2211 | NS | 2216±2171 | 2004±1684 | NS |
| SIRI | 5.6±6.2 | 6±5.7 | NS | 5.6±6.2 | 5.8±6 | NS |
Data reported as number (%) or mean±standard deviation. Statistical significance was defined as p≤0.05. *ASA score was not available for one patient. ICU: Intensive care unit, IMN: Intramedullary nailing, ORIF: Open reduction and internal fixation, ASA: American society of anesthesiologists, WBC: White blood cell, Plt: Platelet, RDW: Red cell distribution width, NLR: Neutrophil/lymphocyte ratio, MLR: Mono-cyte/lymphocyte ratio, SII: Systemic immuno-inflammation index, SIRI: Systemic inflammatory response index, NS: Not significant (p>0.05)
Univariate and multivariable logistic regression analyses were performed to evaluate the clinical and laboratory predictors of outcomes. In the multivariable models, RDW, AF, and Alzheimer’s disease emerged as independent risk factors for ICU admission, while RDW, age, and AF were independently associated with 30-day mortality (Table 4).
Table 4.
Univariate and multivariable logistic regression models examining the relationship between demographic and clinical variables, ICU admission and 30-day mortality
| ICU admission | 30-day mortality | |||||||
|---|---|---|---|---|---|---|---|---|
| OR (95%) | p | ORadj (95%) | Padj | OR (95%) | p | ORadj (95%) | Padj | |
| Age | 1.05 (1.01–1.09) | 0.02 | 1.03 (0.99–1.07) | 0.16 | 1.06 (1.02–1.11) | 0.004 | 1.06 (1.02–1.11) | 0.009 |
| Sex, Male | 1.02 (0.52–1.96) | 0.94 | − | − | 1.74 (0.90–3.35) | 0.09 | − | − |
| Comorbidi-ties | ||||||||
| CHF | 2.39 (1.01–5.18) | 0.04 | 1.20 (0.43–3.10) | 0.72 | 0.85 (0.25–2.25) | 0.77 | − | − |
| COPD | 2.15 (1.03–4.28) | 0.03 | 1.93 (0.87–4.11) | 0.09 | 2.02 (0.95–4.10) | 0.06 | − | − |
| Alzheimer’s disease | 2.11 (1.05–4.10) | 0.03 | 2.43 (1.13–5.14) | 0.03 | 1.56 (0.74–3.12) | 0.23 | − | − |
| AF | 3.38 (1.60–6.88) | 0.001 | 2.82 (1.14–6.70) | 0.02 | 3.16 (1.46–6.54) | 0.002 | 2.62 (1.19–5.53) | 0.01 |
| RDW | 1.27 (1.19–1.45) | 0.001 | 1.22 (1.05–1.41) | 0.009 | 1.23 (1.07–1.42) | 0.003 | 1.20 (1.03–1.39) | 0.016 |
| mFI-5 score | 1.34 (0.98–1.85) | 0.07 | − | − | 1.15 (0.83–1.60) | 0.39 | − | − |
OR and P report results of univariate models. ORadj and Padj report results of multivariable models that include variables that were significant in univariate model analyses. ICU: Intensive care unit, CHF: Congestive heart failure, COPD: Chronic obstructive pulmonary disease, AF: Atrial fibrillation, RDW: Red cell distribution width, mFI-5: Modified 5-item frailty index. Age, RDW, and mFI-5 are modelled as numeric, whereas comorbidities are modelled as categorical variables.
ROC curve analysis revealed that RDW achieved area under the curve (AUC) values of 0.640 (p=0.003) and 0.606 (p=0.03) for ICU admission and 30-day mortality, respectively (Fig. 2). The integration of AF and Alzheimer’s disease with RDW in the predictive model for ICU admission increased the AUC to 0.74. Similarly, incorporating age and AF alongside RDW in the 30-day mortality model improved the AUC to 0.70 (Fig. 3).
Figure 2.
Receiver operating characteristic analysis for admission red cell distribution width values in predicting clinical outcomes. (a) Intensive care unit admission: Area under the curve (AUC) = 0.640, standard error (SE) = 0.048, 95% confidence interval [CI]: 0.546–0.733, p=0.003. (b) 30-day mortality: AUC = 0.606, SE = 0.046, 95% CI: 0.515–0.696, p=0.03
Figure 3.
Receiver operating characteristic (ROC) analysis for the combined predictive models. (a) Intensive care unit admission model (red cell distribution width [RDW] + atrial fibrillation [AF] + Alzheimer’s disease): Area under the curve (AUC) = 0.74, 95% confidence interval [CI]: 0.65–0.81, p<0.001. (b) 30-day mortality model (RDW + AF + age): AUC = 0.70, 95% CI: 0.62–0.79, p<0.001.
DISCUSSION
This study evaluated the predictive value of RDW and mFI-5 for ICU admission and short-term mortality in patients undergoing surgery for HF. Multivariable logistic regression analysis demonstrated that RDW independently predicts both ICU admission and 30-day mortality after adjusting for age, sex, and relevant comorbidities. The mFI-5 score did not reach statistical significance in univariate analysis for either ICU admission or 30-day mortality, and was therefore excluded from multivariable modeling. When considered in isolation, RDW demonstrated modest discriminatory performance, with AUC values of 0.640 for ICU admission and 0.606 for 30-day mortality. However, when combined with other independent clinical variables – AF and Alzheimer’s disease for ICU admission, and AF and age for 30-day mortality – the combined models achieved AUC values of 0.74 and 0.70, respectively, representing a clinically meaningful improvement in risk stratification. These findings suggest that RDW is best utilized not as a standalone screening tool, but as a valuable adjunct marker within a broader multifactorial risk assessment framework.
Intertrochanteric fractures are among the most prevalent HF types in elderly and osteoporotic populations,[11] a pattern reflected in our cohort (62.5%), consistent with previously reported rates.[12] The 30-day mortality rate of 9.9% in our study falls within the range reported in the literature (6–15.4%).[13-16] A consistent sex-based disparity in mortality has been documented across studies, with men facing disproportionately higher early mortality despite lower fracture incidence.[2,12,14,17] Although male sex did not emerge as an independent predictor in our multivariable analysis, the consistently higher observed mortality rates in men suggest that sex-based differences in outcomes warrant further investigation.
Comorbidity burden is well established as a key determinant of post-operative outcomes in HF patients. In a study conducted by Hamdan et al.,[12] cardiovascular disease was identified as the only comorbidity independently associated with mortality, whereas Orhan et al.[18] demonstrated that AF was specifically linked to higher long-term mortality following HF surgery. Our findings corroborate and extend these observations: In our multivariable analysis, AF emerged as an independent predictor of both ICU admission (OR 2.82, 95% CI 1.14–6.70; p=0.02) and 30-day mortality (OR 2.62, 95% CI 1.19–5.53; p=0.01), confirming its role as a robust perioperative risk factor. In addition, Alzheimer’s disease was identified as an independent predictor of ICU admission (OR 2.43, 95% CI 1.13–5.14; p=0.03). This finding likely reflects the compounded vulnerability of cognitively impaired patients in the post-operative period, stemming from impaired physiological reserve, limited capacity for rehabilitation, and increased susceptibility to delirium and respiratory complications.
Several CBC-derived inflammatory indices, including NLR, MLR, SII, and SIRI, have been proposed as prognostic markers in HF patients. A study examining HF in patients over 65 years of age demonstrated that NLR and MLR values were significantly higher among those who died at 30 days and 1 year compared with survivors.[6] In contrast, none of these indices – including SII and SIRI – reached statistical significance in predicting ICU admission or 30-day mortality in our cohort, either in univariate comparisons or in multivariable regression modeling. This discrepancy may be attributable to the chronic low-grade inflammation (often termed “inflammaging”) characteristic of the geriatric population, which can reduce the discriminatory capacity of inflammation-based markers by diminishing the margin between high- and low-risk individuals. Furthermore, the dynamic and fluctuating nature of these indices renders single-point measurements at admission susceptible to confounding by acute physiological perturbations, such as the surgical stress response, pain, and fluid shifts occurring in the perioperative setting.
The association between RDW and adverse clinical outcomes reflects the broader role of this parameter as an integrative marker of systemic physiological dysregulation. Elevated RDW is mechanistically linked to chronic low-grade inflammation, oxidative stress, ineffective erythropoiesis, and nutritional deficiencies – pathways that are not only interrelated but also independently associated with increased mortality risk, particularly in geriatric populations with a high comorbidity burden. Several studies have investigated this relationship specifically in HF patients. Garbharran et al.[19] demonstrated that RDW independently predicted both 4-month and 1-year mortality, while higher pre-operative RDW values were significantly associated with increased 30-day mortality in surgically managed HF patients.[13] Dülgeroğlu et al.[14] observed higher RDW values in non-survivors compared with survivors, and Hamdan et al.,[12] reported that elevated RDW was associated with increased 6-month mortality risk, even in the absence of a statistically significant intergroup difference.[12] In a cohort restricted to patients over 70 years of age, RDW was reported as a significant predictor of mortality, suggesting that its prognostic utility may be particularly pronounced in the oldest patient subgroups.[20] Our findings are consistent with this body of evidence: RDW remained an independent predictor of both ICU admission (OR 1.22, 95% CI 1.05–1.41; p=0.009) and 30-day mortality (OR 1.20, 95% CI 1.03–1.39; p=0.016) in multivariable analysis. While RDW alone demonstrated modest discriminatory performance (AUC 0.640 and 0.606, respectively), its integration into combined clinical models substantially improved predictive accuracy, underscoring its value as an adjunct rather than a standalone risk marker.
The mFI-5 is a practical comorbidity-based scoring tool that has been associated with post-operative complications and prolonged length of stay, though its ability to independently predict mortality has yielded conflicting results across the literature. In patients undergoing surgical fixation for thoracolumbar fractures, mFI-5 scores of 2 or higher were associated with increased 30-day mortality risk, whereas a score of 1 conferred no significant excess risk.[21] In HF patients over 65 years of age, both 30-day and 1-year mortality rates increased progressively with rising mFI-5 scores,[6] a pattern also observed by Traven et al.,[22] who reported an overall mortality rate of 6.6% with a notably lower rate of 3.6% in patients with an mFI-5 score of 0. In our cohort, mFI-5 did not reach statistical significance in univariate analysis for either ICU admission (OR 1.34, 95% CI 0.98–1.85; p=0.07) or 30-day mortality (OR 1.15, 95% CI 0.83–1.60; p=0.39), and was therefore not entered into multivariable modeling. This is likely attributable to the limited number of patients at the higher end of the score distribution – only 10 patients had an mFI-5 score of ≥4 – which substantially constrains statistical power for detecting an independent effect.
These findings have practical implications for patient management and clinical follow-up. RDW can be integrated with clinical risk factors, such as AF, Alzheimer’s disease, and age to identify high-risk patients early and support timely multidisciplinary intervention.
This study has several strengths worth acknowledging. The relatively large patient cohort accumulated over a 5-year period enhances the robustness of the findings, while the use of multivariable logistic regression analysis strengthens the validity of RDW and comorbidities as independent predictors. Nevertheless, several limitations should be considered when interpreting these results. The retrospective single-center design may introduce heterogeneity in data quality and limits the generalizability of the results to broader patient populations. In addition, the inability to determine specific causes of death precludes a definitive analysis of the underlying pathophysiological mechanisms, particularly for early mortality. Furthermore, the limited number of patients in the higher mFI-5 score categories may have constrained statistical power to detect an independent predictive effect of this index. These limitations underscore the need for prospective, multicenter studies with standardized outcome definitions to validate and extend the present findings.
CONCLUSION
Early identification of patients at elevated risk for adverse outcomes remains a critical challenge in the perioperative management of HF. This study demonstrated that RDW, measured at ED admission, independently predicts both ICU admission and 30-day mortality in surgically managed HF patients, and that its predictive performance improves substantially when integrated with other clinical variables, including AF, Alzheimer’s disease, and age. Advanced age, Alzheimer’s disease, and AF further contribute to increased risk and should be closely monitored in clinical management. While mFI-5 did not reach statistical significance as a predictor in this cohort, its potential value in larger prospective studies cannot be excluded. Given its routine availability and ease of calculation, RDW represents a practical adjunct marker for rapid risk stratification in the emergency setting. Future multicenter, prospective studies are warranted to validate and further define the clinical utility of these parameters.
Footnotes
Cite this article as: Değirmenci S, Tirin H, Arikan E, Andiç K. Red cell distribution width and modified 5-item frailty index to predict short-term mortality in patients undergoing surgery for hip fracture. Ulus Travma Acil Cerrahi Derg 2026;32:998-1007.
Ethics Committee Approval
This study received ethical approval from the Non-Interventional Clinical Research Ethics Committee of the Bilecik Şeyh Edebali University Faculty of Medicine (Date: 06.11.2025, Decision No: 9/4).
Peer-review
Externally peer-reviewed.
Authorship Contributions
Concept: S.D., H.T.; Design: S.D., E.S.; Supervision: E.A., K.A.; Resource: S.D., E.A.; Materials: H.T., K.A.; Data collection and/or processing: H.T., E.A., K.A.; Analysis and/or interpretation: S.D., E.S.; Literature review: E.A., K.A.; Writing: S.D., H.T.; Critical review: E.S.
Conflict of Interest
None declared.
Financial Disclosure
The author declared that this study has received no financial support.
References
- 1.Bui M, Nijmeijer WS, Hegeman JH, Witteveen A, Groothuis-Oudshoorn CGM. Systematic review and meta-analysis of preoperative predictors for early mortality following hip fracture surgery. Osteoporos Int. 2024;35:561–74. doi: 10.1007/s00198-023-06942-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Forget P, Moreau N, Engel H, Cornu O, Boland B, De Kock M, et al. The neutrophil-to-lymphocyte ratio (NLR) after surgery for hip fracture (HF) Arch Gerontol Geriatr. 2015;60:366–71. doi: 10.1016/j.archger.2014.11.008. [DOI] [PubMed] [Google Scholar]
- 3.Zhu XF, Weng HY, Huang SF. Red cell distribution width as a predictor of mortality after hip fracture: a systematic review and meta-analysis. Eur Rev Med Pharmacol Sci. 2023;27:6996–7004. doi: 10.26355/eurrev_202308_33271. [DOI] [PubMed] [Google Scholar]
- 4.Pan WG, Chou YC, Wu JL, Yeh TT. Impact of hematologic inflammatory markers on the prognosis of geriatric hip fracture: a systematic review and meta-analysis. Eur J Med Res. 2024;29:609. doi: 10.1186/s40001-024-02211-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Fang Z, Gao B, Wang Z, Chen X, Liu M. Association of systemic inflammation response index with mortality risk in older patients with hip fracture: a 10-year retrospective cohort study. Front Med (Lausanne) 2024;11:1401443. doi: 10.3389/fmed.2024.1401443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bingol O, Ozdemir G, Kulakoglu B, Keskin OH, Korkmaz I, Kilic E. Admission neutrophil-to-lymphocyte ratio and monocyte-to-lymphocyte ratio to predict 30-day and 1-year mortality in geriatric hip fractures. Injury. 2020;51:2663–7. doi: 10.1016/j.injury.2020.07.048. [DOI] [PubMed] [Google Scholar]
- 7.Subramaniam S, Aalberg JJ, Soriano RP, Divino CM. New 5-factor modified frailty ındex using american college of surgeons NSQIP data. J Am Coll Surg. 2018;226:173–81.e8. doi: 10.1016/j.jamcollsurg.2017.11.005. [DOI] [PubMed] [Google Scholar]
- 8.Evans EM. The treatment of trochanteric fractures of the femur. J Bone Joint Surg Br. 1949;31(B):190–203. [PubMed] [Google Scholar]
- 9.Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez JC, et al. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics. 2011;12:77. doi: 10.1186/1471-2105-12-77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.R Core Team . Vienna, Austria: R Foundation for Statistical Computing; 2019. R: A Language and environment for statistical computing. Available at: https://www.r-project.org/ Accessed 02 July 2026. [Google Scholar]
- 11.Adeyemi A, Delhougne G. Incidence and economic burden of intertrochanteric fracture: a medicare claims database analysis. JB JS Open Access. 2019;4:e0045. doi: 10.2106/JBJS.OA.18.00045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Hamdan M, Haddad BI, Jabaiti M, Alryalat SA, Abdulelah AA, Alabed SH, et al. Does red cell distribution width predict hip fracture mortality among the arab population? A single-center retrospective cohort study. Int J Gen Med. 2021;14:10195–202. doi: 10.2147/IJGM.S343538. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Wei-Hsiang H, Zhu Y, Zhang J, Zhang Y. Pretreatment red blood cell distribution width as an efficient predictor of survival in older patients undergoing hip fracture surgery. Int J Clin Pract. 2021;75:e14791. doi: 10.1111/ijcp.14791. [DOI] [PubMed] [Google Scholar]
- 14.Dülgeroğlu TC, Kurt M, Üzümcigil AO, Yilmaz S, Karaaslan F. Multivariate analysis of blood parameters for predicting mortality in patients with hip fractures. Exp Ther Med. 2024;28:414. doi: 10.3892/etm.2024.12703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Blanco JF, da Casa C, Pablos-Hernández C, González-Ramírez A, Julián-Enríquez JM, Díaz-Álvarez A. 30-day mortality after hip fracture surgery: Influence of postoperative factors. PLoS One. 2021;16:e0246963. doi: 10.1371/journal.pone.0246963. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Khan MA, Hossain FS, Ahmed I, Muthukumar N, Mohsen A. Predictors of early mortality after hip fracture surgery. Int Orthop. 2013;37:2119–24. doi: 10.1007/s00264-013-2068-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bajracharya R, Guralnik JM, Shardell MD, Rathbun AM, Yamashita T, Hochberg MC, et al. Long-term sex differences in all-cause and infection-specific mortality post hip fracture. J Am Geriatr Soc. 2022;70:2107–14. doi: 10.1111/jgs.17800. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Orhan AL, Çınar T, Hayıroğlu Mİ, Çiçek V, Selçuk M, Doğan S, Asal S, Yavuz S, Orhan S, Keser N. Atrial fibrillation as a preoperative risk factor predicts long-term mortality in elderly patients without heart failure and undergoing hip fracture surgery. Rev Assoc Med Bras 1992. 2021;67:1633–8. doi: 10.1590/1806-9282.20210686. [DOI] [PubMed] [Google Scholar]
- 19.Garbharran U, Chinthapalli S, Hopper I, George M, Back DL, Dockery F. Red cell distribution width is an independent predictor of mortality in hip fracture. Age Ageing. 2013;42:258–61. doi: 10.1093/ageing/afs176. [DOI] [PubMed] [Google Scholar]
- 20.Karadeniz S, Yurtbay A. 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. Jt Dis Relat Surg. 2022;33:538–46. doi: 10.52312/jdrs.2022.683. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Taghlabi KM, Quaddoura AA, Nisar A, Gupta P, Bhenderu LS, Guerrero JR, et al. Modified 5-ıtem frailty ındex: a useful tool for assessing the ımpact of frailty on postoperative morbidity and mortality following surgical fixation of thoracolumbar fractures. World Neurosurg. 2024;187:e1062–e71. doi: 10.1016/j.wneu.2024.05.045. [DOI] [PubMed] [Google Scholar]
- 22.Traven SA, Reeves RA, Althoff AD, Slone HS, Walton ZJ. New five-factor modified frailty ındex predicts morbidity and mortality in geriatric hip fractures. J Orthop Trauma. 2019;33:319–23. doi: 10.1097/BOT.0000000000001455. [DOI] [PubMed] [Google Scholar]