Article, Critical Care

Neutrophil-to-lymphocyte ratio as a prognostic marker in critically-ill septic patients

a b s t r a c t

Background: We evaluated the associations between the neutrophil-to-lymphocyte ratio or changes in NLR and outcomes in septic patients.

Methods: Patients who met the criteria for severe sepsis or septic shock were categorized into five groups accord- ing to the quintile of initial NLR value. We defined two risk groups according to NLR value and changes in NLR during the first two days (defined as the persistently low NLR group and the persistently high NLR group). The primary outcome was 28-day mortality.

Results: A total of 1395 patients were included. The median initial NLR values from Quintile 1 to Quintile 5 were as follows: 0.2 (IQR [interquartile range], 0.1-0.7), 3.4 (IQR, 2.6-4.7), 8.6 (IQR, 7.1-9.9), 15.4 (IQR, 13.3-17.8), and

31.0 (IQR, 24.6-46.8), respectively. The 28-day mortality values for the same groups were as follows: 24.4%, 12.2%, 11.1%, 11.8%, and 16.1% (P b .01). Cox regression analysis showed that inclusion in Quintile 1 or Quintile 5 was a significant risk factor predicting 28-day mortality compared to Quintile 3 (adjusted hazard ratio [HR]:

1.79 (95% confidence interval [CI], 1.15-2.78) in Quintile 1; 1.67 (95% CI, 1.04-2.66) in Quintile 5). The analysis

indicated that persistently low NLR (adjusted HR: 2.25, 95% CI, 1.63-3.11) and persistently high NLR (adjusted HR: 2.65, 95% CI, 1.64-4.29) were significant risk factors.

Conclusions: In summary, the initial NLR measured at ED admission was independently associated with 28-day mortality in patients with severe sepsis and septic shock. In addition, change in NLR may prove to be a valuable prognostic marker.

Introduction

Sepsis is a major cause of morbidity and mortality resulting from a devastating host response to infection, and it affects millions of people worldwide each year [1,2]. Despite recent advances in knowledge about the disease and critical care modalities, the short-term mortality rate in patients with severe sepsis and septic shock remains high, ac- counting for approximately 30% of all cases [3-5].

Identifying patients that are at high risk of mortality and accurately anticipating outcomes in the early phase of sepsis is imperative for providing timely and adequate interventions for these patients [6]. However, the complexity and heterogeneity of the disease make this

? The authors declare they have no potential conflicts of interest or funding sources.

* Corresponding author at: Department of Emergency Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81 Irwon-ro, Gangnam-gu, Seoul 06351, Republic of Korea.

E-mail address: [email protected] (T.G. Shin).

very challenging. In 2001, Zahorec et al. reported a correlation between neutrophilia/lymphocytopenia and the severity of clinical course in oncological intensive care unit (ICU) patients and suggested the Neutrophil-to-lymphocyte ratio as an indicator of disease severity [7]. Additionally, the NLR is advantageous in regard to simplicity, low cost, and availability compared to many other previously proposed bio- markers, which makes it promising for diagnostic clinicians. Several studies have reported that the NLR is useful in various clinical situations, including cardiovascular disease, malignancies, and infectious diseases [8-11].

Although the association between the NLR and mortality in patients with sepsis has been investigated, the predictive value of the initial NLR and how it changes during the early resuscitation period are not well understood [12,13]. In this study, we evaluated the association between the initial NLR and 28-day mortality in patients that presented to the emergency department (ED) with severe sepsis or septic shock. We also investigated the association between change in NLR after ED ad- mission and 28-day mortality.

http://dx.doi.org/10.1016/j.ajem.2016.10.055

Methods

Study Design

This was a single-center, retrospective cohort study of patients with severe sepsis or septic shock who presented to the ED of Samsung Medical Center, a university-affiliated Tertiary teaching hospital located in a metropolitan city. The Samsung Medical Center Institutional Review Board (IRB file number: 2016-01-101) approved this study, and in- formed consent was waived because the study was conducted retro- spectively and no interventions were applied.

Study Population and Definitions

Patients who fulfilled the following requirements were eligible for inclusion: 1) 18 years of age or older and 2) initially presented to the ED with septic shock or severe sepsis with elevated serum lactate concentration (>= 4 mmol/L) or organ dysfunction. The exclusion criteria were: 1) terminal malignancy, 2) previously signed a “Do Not Resusci- tate” order, or 3) refused invasive procedures such as central line inser- tion or endotracheal intubation.

Severe sepsis was defined as sepsis that was associated with acute organ dysfunction. Septic shock was defined as sepsis with acute circu- latory failure, characterized by persistent hypotension (systolic arterial pressure b 90 mm Hg, mean arterial pressure (MAP) b 60 mm Hg), or a reduction in systolic blood pressure of more than 40 mm Hg from baseline despite adequate fluid resuscitation [14,15]. The NLR was de- termined by dividing the absolute Neutrophil count by the absolute lymphocyte count [7]. Patients were categorized into 5 groups (from Quintile 1 to Quintile 5) according to the quintile of the initial NLR value measured at ED presentation in order to compare baseline charac- teristics and outcomes. We also evaluated changes in the NLR on day 2 compared with that at initial ED presentation (delta NLR, initial NLR – NLR on day 2).

Data Collection and Outcome Measurements

Patients who presented to the ED with severe sepsis or septic shock were prospectively registered in the Samsung Medical Center sepsis registry from August 2008 to September 2014. Patients in our cohort were managed by a specific protocol, based on international guidelines [16,17]. The data from this registry were previously used in other stud- ies on sepsis care [18-23]. Demographic and clinical data of the regis- tered patients were collected by research coordinators and board- certified emergency physicians using a case report form. They reviewed the electronic medical records and verified the final data. The following data were extracted: general patient characteristics, including age, gen- der and comorbidities; focus of infection; presence of bacteremia; and initial sepsis management. Vital signs, laboratory tests, length of stay, and survival data were electronically extracted from the hospital data- base. Circulating neutrophil and lymphocyte counts measured at ED ar- rival and on day 2 after ED admission were collected to calculate the NLR. To estimate disease severity, the Sequential Organ Failure Assess- ment (SOFA) score and Acute physiology and chronic health evaluation II score were calculated at baseline from the registry at the time of diagnosis of severe sepsis or septic shock [24,25]. The primary endpoint was 28-day mortality, and the secondary endpoints were length of ICU stay (LOS) and in-hospital mortality.

Statistical Analysis

The results are presented as median values with interquartile ranges (IQR) for continuous variables and number of patients with percentages for categorical data. Continuous and categorical variables were analyzed by the Kruskal-Wallis test and Chi-square test, respectively. Multivari- able Cox regression models were developed to evaluate the predictive

values of initial NLR and the change in NLR for 28-day mortality. Vari- ables were selected by a backward stepwise method with significance at P b 0.10. The following variables were used for the models: age, gen- der, initial SOFA score, comorbidities (liver cirrhosis, metastatic cancer, and chronic heart failure), focus of infection, mechanical ventilation, and laboratory data (lactate and C-reactive protein). The results were described as the hazard ratio (HR) and 95% confidence interval (CI). Kaplan-Meier curves were plotted to show survival trends across quin- tiles of initial NLR and among groups, according to the change in NLR. The log-rank test was used to compare the survival curves. A P-value less than 0.05 was considered significant. STATA 13.0 software (STATA Corporation, College Station, TX) was used for statistical analysis.

Results

Baseline Characteristics

During the study period, a total of 1728 patients with severe sepsis or septic shock were registered in the sepsis registry. Of these, 333 pa- tients were excluded due to terminal malignancy (n = 160), signed DNR orders, or refusal of invasive procedures (n = 173). A total of 1395 patients were included in the final analysis.

Baseline characteristics of patients across the quintiles are described

in Table 1. The median age of the population was 65 years (IQR, 55-73) and 56.4% (n = 787) were male. The median value of the initial NLR was

8.6 (IQR, 2.6-17.8) in the overall group and, in ascending order from Quintile 1 to Quintile 5: 0.2 (IQR, 0.1-0.7), 3.4 (IQR, 2.6-4.7), 8.6 (IQR,

7.1-9.9), 15.4 (IQR, 13.3-17.8) and 31.0 (IQR, 24.6-46.8). Baseline

SOFA and APACHE II scores were 7 (IQR, 4-9) and 15 (IQR, 11-20), re- spectively, and there were significant differences across the initial NLR quintiles. There were also significant differences in the comorbidities, suspected focus of infection, bacteremia, initial presentation of septic shock, and laboratory tests across the NLR quintiles.

Comparison of Outcomes

The overall 28-day mortality was 15.1% (n = 211) (Table 2). When the 28-day mortality was compared across the initial NLR, Quintile 1 (24.4%) showed the highest mortality followed by Quintile 5, Quintile 2, Quintile 4, and Quintile 3 (16.1%, 12.2%, 11.8%, and 11.1%, respective- ly) (P b .01). ICU admission rate and in-hospital mortality were also sig- nificantly different. The Kaplan-Meier survival curves based on the initial NLR quintiles are shown in Fig. 1a.

Comparisons of 28-day mortality, according to the initial NLR and NLR on day 2 quintiles, are shown in Table 3. The 28-day mortality was relatively higher in both Quintile 1 (36.4%) and Quintile 5 (26.2%) of initial NLR and NLR on day 2. Three groups were classified by change in NLR, which was combined with initial NLR to compare mortality (Fig. 2): increased NLR (delta NLR b -4.3), minimal or no change in NLR (delta NLR -4.3 to 3.99), and decreased NLR (delta NLR N 3.99). The combinations of Quintile 5 with minimal or no changes in NLR, Quintile 1 with minimal or no change in NLR, and Quintile 5 with in- creased NLR showed higher 28-day mortality of 40.9%, 33.5% and 28.2%, respectively.

Based on these results, we defined two risk groups: the persistently low NLR group, Quintile 1 with minimal or no change in NLR, and the persistently high NLR group, Quintile 5 with minimal or no change and Quintile 5 with increased NLR. The Kaplan-Meier survival curves based on the risk groups are shown in Fig. 1b.

Cox Regression Analysis for 28-Day Mortality

After adjusting for potential confounding factors, Quintile 1 and Quintile 5 of initial NLR were significant risk factors for predicting 28-day mortality compared to Quintile 3 [adjusted HR: 1.79 (95% CI, 1.15-2.78, P = .01) in Quintile 1; 1.67 (95% CI, 1.04-2.66, P = .03) in

Table 1

Baseline characteristics of initial neutrophil-to-lymphocyte ratios across quintiles.

	Overall	Quintile 1	Quintile 2	Quintile 3	Quintile 4	Quintile 5	P-value
		(n = 279)	(n = 279)	(n = 279)	(n = 279)	(n = 279)
Age (years)	65 (55-73)	64 (54-70)	63 (53-73)	66 (55-74)	67 (56-74)	67 (54-74)	b0.01
Gender (male)	787 (56.4)	159 (57.0)	161 (57.7)	147 (52.7)	151 (54.1)	169 (60.6)	0.36
Comorbidities
Hypertension	466 (33.4)	75 (26.9)	93 (33.3)	106 (38.0)	86 (30.8)	106 (38.0)	0.02
Diabetes	314 (22.5)	43 (15.4)	56 (20.1)	75 (26.9)	65 (23.3)	75 (26.9)	b0.01
Cardiovascular disease	152 (10.9)	29 (10.4)	24 (8.6)	30 (10.8)	34 (12.2)	35 (12.5)	0.58
Chronic lung disease	80 (5.7)	15 (5.4)	20 (7.2)	14 (5.0)	15 (5.4)	16 (5.7)	0.83
Chronic renal disease	73 (5.2)	7 (2.5)	13 (4.7)	20 (7.2)	13 (4.7)	20 (7.2)	0.07
Liver cirrhosis	90 (6.5)	5 (1.8)	16 (5.7)	22 (7.9)	24 (8.6)	23 (8.2)	b0.01
Metastatic solid cancer	325 (23.3)	94 (33.7)	74 (26.5)	54 (19.4)	52 (18.6)	51 (18.3)	b0.01
Hematologic malignancy	128 (9.2)	71 (25.5)	27 (9.7)	14 (5.0)	9 (3.2)	7 (2.5)	b0.01
Suspected infection focus Intra-Abdominal infection	476 (34.1)	59 (21.2)	101 (36.2)	100 (35.8)	107 (38.4)	109 (39.1)	b0.01
Pneumonia	514 (36.9)	136 (48.8)	103 (36.9)	94 (33.7)	90 (32.3)	91 (32.6)	b0.01
Urinary tract infection	172 (12.3)	15 (5.4)	33 (11.8)	46 (16.5)	41 (14.7)	37 (13.3)	b0.01
Other	233 (16.7)	69 (24.7)	42 (15.1)	39 (14.0)	41 (14.7)	42 (15.1)	b0.01
Bacteremia	609 (43.7)	133 (47.7)	98 (35.1)	108 (38.7)	126 (45.2)	144 (51.6)	b0.01
Initial presentation of septic shock	755 (54.1)	169 (60.6)	134 (48.0)	148 (53.1)	160 (57.4)	144 (51.6)	0.03
Use of vasopressors	878 (62.9)	197 (70.6)	162 (58.1)	176 (63.1)	181 (64.9)	162 (58.1)	0.02
Mechanical ventilation	119 (8.5)	34 (12.2)	22 (7.9)	21 (7.5)	16 (5.7)	26 (9.3)	0.08
SOFA score	7 (4-9)	8 (5-10)	6 (4-9)	6 (4-9)	6 (4-9)	7 (4-9)	b0.01
APACHE II score	15 (11 -20)	20 (14-24)	14 (10-19)	13 (10-18)	13 (10-18)	15 (10-19)	b0.01
Laboratory tests White blood cell count, 10⁹/L	8.7 (2.7-15.1)	0.7 (0.3-1.8)	4.8 (2.0-9.2)	9.6 (5.7-14-3)	13.0 (8.5-18.5)	16.3 (11.0-21.8)	b0.01
Leukopenia b 4000 cells/mm³	428 (30.7)	241 (86.4)	119 (42.7)	40 (14.3)	15 (15.4)	13 (4.7)	b0.01
Leukocytosis N 12 000 cells/mm³	519 (37.2)	18 (6.5)	36 (12.9)	107 (38.4)	158 (56.6)	200 (71.7)	b0.01
ANC, 10⁹/L	6.9 (1.7-12.8)	0.1 (0.02-0.5)	3.4 (1.5-6.6)	8.2 (4.9-11.8)	11.4 (7.5-16.4)	15.2 (10.1-20.4)	b0.01
ALC, 10⁹/L	0.7 (0.3-1.2)	0.4 (0.2-1.0)	1.0 (0.4-1.8)	1.0 (0.6-1.4)	0.8 (0.5-1.1)	0.4 (0.3-0.6)	b0.01
Initial NLR	8.6 (2.6-17.8)	0.2 (0.1-0.7)	3.4 (2.6-4.7)	8.6 (7.1-9.9)	15.4 (13.3-17.8)	31.0 (24.6-46.8)	b0.01
Presence of band cells	298 (21.4)	51 (18.3)	53 (19.0)	44 (15.8)	69 (24.7)	81 (29.0)	b0.01
Hemoglobin, g/dL	11.3 (9.6-13.2)	9.5 (8.2-11.1)	11.4 (9.9-13.5)	11.9 (10.1-13.4)	11.8 (10.3-13.4)	11.5 (9.9-13.3)	b0.01
Platelet count, 10⁹/L	140 (73-212)	61 (18-140)	142 (89-212)	166 (91-225)	168 (102-234)	154 (84-234)	b0.01
C-reactive protein, mg/dL	11.0 (4.2-22.0)	13.2 (5.0-26.1)	7.4 (2.2-20.0)	8.6 (3.4-18.1)	13.0 (4.5-21.4)	12.8 (6.4-23.0)	b0.01
Initial serum lactate, mmol/L	4.3 (2.7-5.7)	4.5 (3.0-6.0)	4.5 (3.2-6.0)	4.2 (2.2-5.2)	4.2 (2.4-5.4)	4.4 (2.6-6.0)	b0.01

Data are shown as median with interquartile range or n (%).

SOFA, Sequential Organ Failure Assessment; ANC, absolute neutrophil count; ALC, absolute lymphocyte count; NLR, neutrophil-to-lymphocyte ratio.

Quintile 5] (Table 4). The persistently low NLR group (adjusted HR: 2.25, 95% CI, 1.63-3.11, P b .01) and persistently high NLR group (ad- justed HR: 2.65, 95% CI, 1.64-4.29, P b .01) were associated with signif- icant risk factors for predicting 28-day mortality compared to patients without risk.

Discussion

The results indicated that the initial NLR measured on ED admission was an independent risk factor for predicting 28-day mortality in pa- tients with severe sepsis or septic shock. In addition, both the persis- tently low NLR group and the group that had a persistently high NLR during the first two days after ED admission were also associated with increased risk for 28-day mortality.

NLR has been used in critical care medicine and may prove to be a valuable Prognostic tool based on the results of this study as well as others. Akilli et al. conducted a prospective study of critically-ill patients who presented in the ED and required ICU care [26]. In their study, high NLR measured at ED was independently associated with in-hospital

mortality and 6-month mortality. In addition, high NLR was also related to a risk of multi-organ failure and sepsis development. More recently, Riche et al. revealed an association between NLR and risk of death in pa- tients with septic shock. [13]. They also suggested that NLR could be used as an indicator of early (before day 5) and late (on or after day 5 after septic shock onset) death. In a study conducted by Salciccioli et al., NLR measured at the time of ICU admission was associated with 28-day mortality in unselected critically ill patients. In subgroup analy- sis, however, there was no association between NLR and mortality in sepsis patients [12].

Some of the pathophysiologic processes in severe sepsis and septic shock were potentially associated with the results in this study, includ- ing both widespread activation and dysfunction of the immune system [2,30]. Neutrophils rapidly respond to microbial infection, resulting in a dramatic increase in number of circulating neutrophils that migrate to the affected area. Neutrophils play pivotal roles in the innate immune response including directly killing pathogens by phagocytosis, releasing a variety of cytokines, and activating T cells. During sepsis progression, large quantities of pro-inflammatory cytokines are released; however,

Table 2

Comparison of initial neutrophil-to-lymphocyte ratios across quintiles.

	Overall	Quintile 1 (n = 279)	Quintile 2 (n = 279)	Quintile 3 (n = 279)	Quintile 4 (n = 279)	Quintile 5 (n = 279)	P-value
ICU admission from the ED	624 (44.7)	142 (50.9)	127 (45.5)	132 (47.3)	110 (39.4)	113 (40.5)	0.04
ICU LOS (days)	6 (3-15)	6 (2-14)	6 (3-14)	6 (3-16)	7 (2-14)	7 (3-17)	0.85
In-hospital mortality	228 (16.3)	69 (24.7)	43 (15.4)	37 (13.3)	34 (12.2)	45 (16.1)	b 0.01
28-Day mortality	211 (15.1)	68 (24.4)	34 (12.2)	31 (11.1)	33 (11.8)	45 (16.1)	b 0.01

Data are shown as median with interquartile range or n (%).

ICU, intensive care unit; ED, emergency department; LOS, length of stay.

Fig. 1. Kaplan-Meier survival curves. (a) Comparison of initial neutrophil-to-lymphocyte ratios across quintiles. (b) Comparison of neutrophil-to-lymphocyte ratio changes during the first 48 h.

severe and uncontrolled innate immune system activation can result in tissue injury followed by multiple organ failure and even death. On the other hand, immunosuppression is induced by the release of various anti-inflammatory cytokines, and apoptosis is induced in a large num- ber of lymphocytes. The NLR indicates the balance between neutrophil and lymphocyte levels in the body and is an indicator of systemic in- flammation [7]. In this context, the persistently high NLR group might represent patients that have an ongoing severe Inflammatory process due to infection. In this situation, the bone marrow production of neu- trophils persistently increases, resulting in a massive recruitment of

immature neutrophils into circulation. In addition, evidence has shown that neutrophil apoptosis is delayed in complicated sepsis cases [31]. Lymphocytopenia is another important component that could be involved in such a scenario. Persistent lymphocytopenia could be an indicator of sepsis-induced immunosuppression and has been found to be a predicting factor for mortality in septic patients [32]. The NLR is influenced by several conditions, including medications and comorbidities that affect the neutrophil and lymphocyte count. In addition, some disease- or treatment-related changes in peripheral leukocytes might occur in sepsis patients. In this context, careful

Table 3

Twenty-eight-day mortality by quintile of neutrophil-to-lymphocyte ratios during the first two days.

	NLR on day 2^a
	Quintile 1	Quintile 2	Quintile 3	Quintile 4	Quintile 5	P-value
	(n = 279)	(n = 279)	(n = 279)	(n = 279)	(n = 279)
Initial NLR Quintile 1 (n = 279)	36.4 (51/140)	11.1 (5/45)	6.5 (3/46)	3.7 (1/27)	14.3 (3/21)	b 0.01
Quintile 2 (n = 279)	22.2 (8/36)	16.4 (11/67)	10.0 (7/70)	4.9 (3/61)	4.4 (2/45)	0.35
Quintile 3 (n = 279)	9.5 (2/21)	15.1 (8/53)	9.5 (6/63)	12.9 (9/70)	11.1 (8/72)	0.10
Quintile 4 (n = 279)	11.8 (4.34)	13.6 (6/44)	22.0 (11/50)	14.1 (10/71)	20.0 (16/80)	0.09
Quintile 5 (n = 279)	6.3 (3/48)	5.7 (4/70)	8.0 (4/50)	20.0 (10/50)	26.2 (16/61)	0.02

Data are shown as % (n).

NLR, neutrophil-to-lymphocyte ratio.

^a The median of NLR on day 2 were 8.2 (IQR, 2.8-24.3) in overall group, 0.8 (IQR, 0.1-1.4) in Quintile 1, 3.6 (IQR, 2.8-4.4) in Quintile 2, 8.2 (IQR, 6.8-9.9) in Quintile 3, 18.9 (IQR, 15.0- 24.3) in Quintile 4, and 63.7 (IQR, 44.5-137.6) in Quintile 5.

Fig. 2. Comparison of 28-day mortality according to changes in the neutrophil-to-lymphocyte ratio during the first two days after ED admission (Group 1: increased NLR, delta NLR b -4.3; Group 2: minimal or no change in NLR, delta NLR -4.3 to 3.99; Group 3: decreased NLR, delta NLR N 3.99; *P b .01, **P N .5).

interpretation of the NLR is required. For instance, in contrast to the pre- vious studies, we included neutropenic patients regardless of precipitat- ing factors. The majority of patients included in the first quintile of initial NLR showed neutropenia, which is a risk factor for bacterial infec- tion, and subsequent severe sepsis and septic shock can result in high mortality [33]. This could be a major factor affecting the similar high mortality rates in the low and high NLR groups. Among these patients, the persistently low NLR group showed higher mortality compared to patients with increased NLR. We believe that this finding could reflect real clinical situations in predicting outcomes. Additional analysis that excluded patients with potential contributing factors including neutropenia or hematologic disease showed the same trend of high NLR associated with poor outcomes.

A perfect biomarker for sepsis has not yet been identified, although various tests have been investigated [6,18,27-29]. The NLR has a strength in that it is derived from extremely common laboratory values. Change in NLR as well as initial NLR could be used for identifying pa- tients at high risk of poor outcomes. These values may also provide helpful information about the initial therapeutic response and aid in evaluating host immune responses. However, further clinical studies are required to evaluate the benefit of NLR to sepsis care improvement and the additional role of the NLR compared with other common prog- nostic markers including lactate and procalcitonin.

There were several limitations to this study that should be taken into consideration when interpreting the results. First, the study was conduct- ed retrospectively in a single center, without intervention. Therefore, it was subject to selection bias and may not be generalizable to other

Table 4

Cox regression analysis of 28-day mortality.

Variable Univariable Multivariable

Unadjusted HR P Adjusted HR P (95% CI) (95% CI)

NLR at the time of initial evaluation

Quintile 1 2.37 (1.55-3.63) b0.01 1.79 (1.15-2.78) 0.01

Quintile 2 1.10 (0.67-1.79) 0.70 1.18 (0.72-1.94) 0.52

Quintile 3 Reference – Reference –

Quintile 4 1.06 (0.65-1.73) 0.80 1.18 (0.72-1.93) 0.52

Quintile 5 1.48 (0.94-2.34) 0.09 1.67 (1.04-2.66) 0.03

Risk groups by NLR change

No risk group^a Reference – Reference –

settings. In addition, the causal relationship between NLR and mortality could not be confirmed. Second, we divided our patients into quintiles based on the initial NLR on ED admission; however, the baseline charac- teristics of patients were heterogeneous and not well-balanced. Although we attempted to adjust for potential confounding factors, these discrep- ancies could affect the results, and differences in the patient composition of the different groups should be taken into consideration. Third, a new definition of sepsis and septic shock has been released recently, and the concept of severe sepsis has been replaced by sepsis. However, this defi- nition was not implemented in this study [34-36].

Conclusion

The initial NLR measured at ED admission was independently associated with 28-day mortality in patients with severe sepsis or septic shock in the ED. Persistently low NLR or high NLR values were also significant predictors for 28-day mortality.

Conflict of Interest

The authors have no potential conflicts of interest or funding sources to declare.

References

AJEM