The predictive value of RDW in AKI and mortality in patients with traumatic brain injury

Abstract Background Red blood cell distribution width (RDW) has been validated valuable in predicting outcome and acute kidney injury (AKI) in several clinical settings. The aim of this study was to explore whether RDW is associated with outcome and AKI in patients with traumatic brain injury (TBI). Methods Patients admitted to our hospital for TBI from January 2015 to August 2018 were included in this study. Multivariate logistic regression analysis was performed to identify risk factors of AKI and outcome in patients with TBI. The value of RDW in predicting AKI and outcome was evaluated by receiver operating characteristic (ROC) curve. Results Three hundred and eighteen patients were included in this study. The median of RDW was 14.25%. We divided subjects into two groups based on the median and compared difference of variables between two groups. The incidence of AKI and mortality was higher in high RDW (RDW > 14.25) group (31.45% vs 9.43%, P < .001; 69.81% vs 29.56%, P < .001). Spearman's method showed RDW was moderately associated with 90‐day Glasgow Outcome Scale (GOS) (P < .001). In multivariate logistic regression analysis, RDW, lymphocyte, chlorine, and serum creatinine were risk factors of AKI. And Glasgow Coma Scale (GCS), glucose, chlorine, AKI, and RDW were risk factors of mortality. The area under the ROC curve (AUC) of RDW for predicting AKI and mortality was 0.724 (0.662‐0.786) and 0.754 (0.701‐0.807), respectively. Patients with higher RDW were likely to have shorter median survival time (58 vs 70, P < .001). Conclusions Red blood cell distribution width is an independent risk factor of AKI and mortality in patients with TBI.


| INTRODUC TI ON
Traumatic brain injury (TBI) has attracted much attention for the severe burden it brings to the public health and social economy.
More than 50 million people suffer TBI each year in the world. 1 And it has become the primary cause of mortality in people under 40 years. 2 Even if patients suffered from TBI survive, their daily activities would be strictly limited and quality of life would be severely impaired. 3 It has been demonstrated that not only the primary and secondary brain injury, but also the non-neurologic organ dysfunction following TBI would contribute to the increased risk of mortality. 4 Acute kidney injury (AKI), one of non-neurologic complications following TBI, has attracted much attention because of its effect on the increased mortality. [5][6][7][8] Therefore, identifying patients who are at high risk of developing AKI and taking suitable treatment strategies may be beneficial for recovery and prognosis of patients.
Red blood cell distribution width (RDW) usually plays an important role in the diagnosis and classification of anemia. However, it has been regarded as an indicator of inflammation, recently. A growing number of studies evaluate the prognostic value of RDW in various clinical settings including critical illness, acute pancreatitis, cancer, diabetes mellitus, and stroke. [9][10][11][12][13] There are two studies demonstrating the association between RDW and mortality in patients with TBI. However, in consideration of the value of RDW in predicting the prognosis of TBI, these two studies have shown conflicting conclusions. 14,15 It is unknown whether RDW can really act as a valuable marker in predicting outcome of patients with TBI.
In addition to the prognostic value, RDW also has been confirmed valuable in predicting AKI in many clinical settings. Recent studies report that RDW is associated with AKI in patients with sepsis, after cardiac surgery, and those hospitalized in coronary care unit, respectively. [16][17][18] However, the predictive value of RDW in AKI following TBI has not been validated. We design this study F I G U R E 1 Flowchart of participant recruitment to confirm whether RDW is helpful in predicting the outcome of patients with TBI and discover the predicting value of RDW in AKI following TBI.

| Patients
We conducted this retrospective study in West China Hospital affiliated to Sichuan University. A total of 318 patients admitted for TBI between January 2015 and August 2018 were included in our study. The inclusion criteria of participants were showed as follows: (a) patients diagnosed with TBI and transferred to our hospital within 24 hours after injury; (b) patients hospitalized in the neurosurgery department or neuro-intensive care unit (NICU) for more than 48 hours.
These patients were excluded from this study: (a) patients had a history of cardiovascular diseases, hematological system diseases, metabolic diseases, chronic hepatorenal diseases, and cancer; (b) patients whose clinical records and laboratory tests were not complete and who did not follow treatment strategies (Figure 1). And we considered it was unnecessary to acquire informed consent because of this study design. Note: All values are expressed as n (%) or median (first to third quartile).

| Data collection
Blood samples of patients were regularly collected once they admitted to our hospital. All laboratory and clinical data were collected from our electronic medical record (EMR) system. We selected results of laboratory test on admission as included variables. The GCS score on admis- increase of SCr of ≥1.5 times over baseline (which is known or presumed within the prior 7 days), and (c) urine volume < 0.5 mL/kg/h for 6 hours.
The survival outcome and GOS of patients were obtained through medical record or telephone following up until 3 months after initial injury.

| Statistical analysis
Kolmogorov-Smirnov test was used to testify the normality of included variables. Categorical data were showed as numbers (percentage). Normal distribution data were showed as mean ± standard deviation while non-normal distribution data were showed as median (interquartile range). The difference between two normal distribution variables was testified using independent Student's t test.
And the difference between non-normal distribution variables was compared using Mann-Whitney U test. Chi-square test was used to analyze the difference between categorical variables. We used univariate and multivariate logistic regression analysis to discover potential risk factors of AKI and mortality. The odds ratio (OR) and 95% confidence intervals (CI) of all risk factors were calculated. We also calculated Spearman's correlation coefficients for the relationships between RDW and other clinical or laboratory factors. ROC curves were generated to evaluate predictive value of RDW in AKI and mortality. Meanwhile, the sensitivity, specificity, and cutoff value were calculated. In addition, we divided patients into two groups based on the cutoff value and then draw cumulative survival curves of two groups using Kaplan-Meier analysis. The difference of survival curves between two groups was compared by log-rank test.
A P value < .05 was considered statistically significant. SPSS 22.0 Windows software (SPSS, Inc) was used for all statistical analyses.

| Baseline characteristics of patients
The median of RDW was 14.25%. We divided patients into two groups including high RDW (RDW > 14.25%) group and low RDW (RDW < 14.25%) group. As we can see in Table 1, the age of high RDW group was higher than low RDW group (47 vs 41, P < .001).

| Factors associated with AKI and mortality following TBI
In univariate logistic regression analysis, many variables were found associated with AKI and mortality including shock, GCS, chlorine, and RDW.

| Predictive values of RDW in AKI and mortality
By drawing ROC curve, we calculated that the AUC of GCS to predict AKI was 0.672 ( Figure 2). The sensitivity and specificity of GCS were 0.52 and 0.773, respectively (  Figure 3). The AUC of model 2 was 0.891 with sensitivity of 0.848 and specificity of 0.8.

| Association of RDW and survival time
We divided patients into two groups including whose RDW ≥ 14% and whose RDW < 14% according to the best cutoff value. Abbreviations: CI, confidence interval; GCS, Glasgow Coma Scale; MAP, mean arterial pressure; OR, odds ratio; RDW, red blood cell distribution width; WBC, white blood cell.

TA B L E 3
Univariate and multivariate logistic regression analysis of factors associated with AKI following TBI group was significantly shorter than low RDW group (P < .001) ( Figure 4).

| D ISCUSS I ON
Acute kidney injury following TBI has been paid much attention for the harmful effects it brings to the outcome. Several studies have investigated the incidence of AKI following TBI, ranged from 3.5% to 40%. [5][6][7]19,20 The incidence of AKI was 20.44% in our study. This difference may be caused by the different diagnostic criteria of AKI over the years and respective medical level in different hospitals.
It is now generally believed that non-neurologic complications including AKI following TBI may be due to the massive catecholamine release, neuroinflammation, and side effects of therapies aimed at neuroprotection. 21,22 There were some studies exploring the association between elevated RDW and the development of AKI in several patient groups. 17 Abbreviations: AKI, acute kidney injury; CI, confidence interval; GCS, Glasgow Coma Scale; MAP, mean arterial pressure; OR, odds ratio; RDW, red blood cell distribution width; WBC, white blood cell.

TA B L E 4
Univariate and multivariate logistic regression analysis of factors associated with in-hospital mortality after TBI study has validated the predictive value of serum CRP, a common inflammatory marker, in AKI following aneurysmal subarachnoid hemorrhage. 30 Therefore, the RDW might be associated with AKI following TBI through the mediation of systemic inflammation.
Secondly, oxidative stress may contribute to anisocytosis through inhibiting erythropoiesis and impairing red blood cell membrane deformability which in turn shorten the circulation half-life of red blood cell and result in an increase in RDW value. 31,32 Furthermore, the triangle among reactive oxygen species (ROS) and nitric oxide (NO), and oxygen was involved in the oxidative stress system and play an important role in the pathogenesis of AKI. 33,34 Thus, RDW may be valuable in evaluating the oxidative stress status in kidney. Thirdly, it has been illustrated that the sympathetic system and the renin-angiotensin system (RAAS) may promote the release of erythropoietin.
Consequently, erythropoiesis is accelerated and the heterogeneity of red blood cell increased. 35 The catecholamine surge after TBI may contribute to non-neurologic complications including AKI. 21 We make a reasonable conjecture that the increase of RDW after including catecholamine, and albumin. [44][45][46] The association between increased RDW and unfavorable outcome may be mediated through these mechanisms. In addition, increased RDW means that an increasing number of red blood cells with partially saturated hemoglobin.
One study inferred that elevated RDW reflected decreased oxygen transport capacity. 47 It is well known that adequate oxygen delivery is essential for cerebral metabolism and recovery of patients with TBI.
Excessively high value of RDW can cause the insufficiency of cerebral oxygen and then worsen the outcome.
In addition to RDW, we found that lymphocyte, chlorine, and serum creatinine were also risk factors of AKI in model 1. Recent evidence demonstrates that both innate and adaptive immune cells play an important role in initiating and promoting damage to renal tubular in the development of AKI. 48 In addition, there are several studies confirming that high neutrophil lymphocyte ratio is independently associated with the development of AKI in some clinical settings. [49][50][51] The OR value of lymphocyte in multivariate logistic regression analysis is 0.245, which indicates that increased lymphocyte is a protective factor in the development of AKI. However, this result is worth further exploring in future study because of diverse role of different immune cells in AKI including monocytes, neutrophils, T lymphocytes, and B lymphocytes. 52 Hyperchloremia has been validated valuable in predicting AKI in patients undergoing craniotomy for brain tumor resection and those diagnosed with subarachnoid hemorrhage. 53,[53][54][55][56] This fact emphasized that physicians should pay attention to AKI when regularly using 0.9% normal saline to reduce intracranial pressure.

| Study limitations
There are several limitations in our study. First, the retrospective and single-center study design leads to the fact that selection bias was unavoidable. Second, hematopoietic raw material including iron, vitamin B12, and folate was not measured and history of hemolysis or blood transfusion was not recorded. These factors might influence the value of RDW. Third, as mentioned above, inflammation and activation of sympathetic system and RAAS might contribute to the development of AKI and unfavorable outcome after TBI. However, we did not measure well-known markers of these systems including C-reactive protein, norepinephrine, and angiotensin II. Finally, stages of AKI were not recorded by us so that we could not analyze the exact association between RDW and the severity of AKI.

| CON CLUS ION
This study suggests that RDW is valuable in predicting AKI following TBI and is an efficient and economical predictor of in-hospital outcome in patients with TBI.

CO N FLI C T O F I NTE R E S T
The authors declared that they have no conflicts of interest to this work.