a b s t r a c t

Study objective: The number of critically ill patients admitted to the emergency department increases daily. To de- crease mortality, interventions and treatments should be conducted in a timely manner. It has been found that the neutrophil-lymphocyte ratio (NLR) is related to mortality in some disease groups, such as acute coronary syn- drome and pulmonary emboli. The effect of the NLR on mortality is unknown in critically ill patients who are ad- mitted to the emergency department. Our aim in this study is to evaluate the effect of the NLR on mortality in critically ill patients.

Methods: This study was planned as a prospective, observational cohort study. Patients who were admitted to the emergency department because they were critically ill and required the intensive care unit were included in the study. Demographic characteristics, Acute Physiology and Chronic Health Evaluation II (APACHE II), Sepsis- related Organ Failure Assessment, Glasgow Coma Score, and NLR values were recorded upon emergency depart- ment admission. The patients were followed up for sepsis, ventilator-associated pneumonia, multiorgan failure, in-hospital mortality, and 6-month mortality.

Results: The median (interquartile range) age of the 373 patients was 74 (190) years, and 54.4% were men. Neutrophil-lymphocyte ratio values were divided into quartiles, as follows: less than 3.48, 3.48 to 6.73, 6.74- 13.6, and more than 13.6. There was no difference among these 4 groups regarding demographic characteristics, APACHE II score, Sepsis-related Organ Failure Assessment score, Glasgow Coma Score, and length of hospital stay (P N .05). In the multivariable Cox regression model, in-hospital mortality and 6-month mortality NLR were hazard ratio (HR), 1.63 (1.110-2.415; P = .01) and HR, 1.58 (1.136-2.213; P = .007), respectively, and APACHE II scores were detected as independent indicators.

Conclusion: The NLR is a simple, cheap, rapidly available, and independent indicator of short- and long-term mor- talities. We suggest that the NLR can provide direction to emergency department physicians for interventions, particularly within a few hours after admission, in the critically ill patient group.

(C) 2014

Introduction

Although advancements in science and technology have increased the average Life expectancy, it has also created a patient group with high comorbidity within the advanced age group. Therefore, the num- ber of critically ill patients in the emergency department is increasing daily, and these patients’ medical conditions are more complicated, and their emergency department services required gets longer. In a study that compared the length of stay of patients in the emergency de- partment during the years 1988 and 1997, the length of stay increased by 152% [1]. early diagnosis and treatment of these patients in a timely

* Corresponding author. Konya Egitim ve Arastirma Hastanesi Acil Servis Bolumu 42080-Konya, Turkey. Tel.:+90 332 2236409; fax: +90 332 2237941.

E-mail address: [email protected] (N.B. Akilli).

manner are of vital importance. The best example of this is early goal- directed therapy for sepsis [2]. According to previous studies, impaired physiologic parameters can be reversed with interventions in the emer- gency department, and these several hours are as equally important as the first 3 days in the intensive care unit (ICU). This state will be called the “golden hour” and “silver day [3-5]“.

white blood cell count is a parameter that is included in many scoring systems that are frequently used in daily practice for the diagnosis and follow-up of diseases. According to recent studies, impor- tant changes occur in WBC subtypes under stress. Galus and Stern [6] re- ported that extreme lymphocytopenia may occur during infectious emergencies, such as toxic Shock syndrome. Jilma et al [7] detected an increase in the number of neutrophils and a decrease in monocytes after inflammation. Along with these changes, Zahorec [8] identified a new parameter known as the neutrophil-lymphocyte stress factor.

http://dx.doi.org/10.1016/j.ajem.2014.09.001

0735-6757/(C) 2014

According to this parameter, when the ratio of peripheral neutrophils to lymphocytes (NLR) is compared with the WBC count, a better prognos- tic indicator is achieved. Furthermore, studies based on acute coronary syndrome, congestive heart failure, and pulmonary emboli have dem- onstrated that the NLR is related to mortality [9-12]. The relationship of the NLR to mortality is unknown in critically ill patients.

In this study, we aimed to evaluate the effect of the NLR on in- hospital mortality and 6-month mortality in critically ill patients who were admitted to the emergency department.

Materials and methods

Study design and setting

This prospective, observational cohort study was conducted in a re- gional academic hospital emergency department that provides tertiary care in Konya. This emergency department provides care to approxi- mately 300 000 patients annually, and 2.6% of the patients are admitted to the ICU. The study was conducted between January 1, 2013, and Au- gust 10, 2013. Approval from the local ethics committee was obtained for the study and is compatible with the Helsinki Declaration.

Selection of participants

Critically ill adult patients were included in the study and were de- fined as patients accepted by the emergency department who had 2 of the 4 systemic inflammatory response syndrome criteria: heart rate more than 90 beats per minute, respiratory rate more than 20 breaths per minute, temperature more than 38?C or less than 36?C, WBC count more than 12 000 cells/mm³ and less than 4000 cells/mm³ or more than 10% band cells as well as systolic blood pressure below 90 mm Hg or 2 of the systemic inflammatory response syndrome criteria and a lactate level above 4 mmol/L [13]. Exclusion criteria were age younger than 18 years, pregnancy, known hematologic dis- ease, previous chemotherapy treatment (within the last month), blood transfusion (within the last 2 weeks), chronic hepatic disease, trauma, or poisoning.

Outcome measures

Demographic data, comorbidities, and diagnoses of all patients were recorded. Vital findings upon emergency department admission, labo- ratory findings, and neutrophil and lymphocyte values from whole blood counts were taken. Neutrophil-lymphocyte ratio, Acute Physiolo- gy and Chronic Health Evaluation II (APACHE II) [14], Sepsis-related Organ Failure Assessment (SOFA) [15], and Glasgow Coma Score scores were calculated. After the patients were accepted to the ICU from the emergency department, they were evaluated for the development of sepsis, ventilator-associated pneumonia, and multiorgan failure (MOF). duration of hospital stay was recorded. In- hospital and 6-month mortalities of the patients were determined from medical records. In the cases where data were not available from the records, data were obtained via phone call. The primary end point blood analysis“>was in-hospital mortality and 6-month mortality; secondary end points were sepsis, ventilator-associated pneumonia, and MOF, respectively.

Blood analysis

Complete blood counts and differentials were studied in the periph- eral blood samples. Blood samples were taken in calcium-EDTA tubes. Complete blood counts were performed with an autoanalyzer (SIEMENS ADVIA 120 hematology analyzer; SIEMENS, Eschborn, Germany). The NLR was calculated after the whole blood count NLR.

Data analysis

Statistical analysis was performed using SPSS version 15.0 for Win- dows (SPSS, Chicago, IL). Both visual (histogram and probability graphs) and analytical (Kolmogorov-Smirnov and Shapiro-Wilk tests) methods were used to determine if the data were normally distributed. Descrip- tive variables are expressed as the mean +- SD for data that are normally distributed and as the median and interquartile range (IQR) for vari- ables that are not normally distributed. The ?² or Fisher exact test was used to compare categorical values. Clinical and laboratory characteris- tics, according to the NLR quartile, were evaluated via one-way analysis of variance for normally distributed variables, whereas the Kruskal- Wallis test was used for variables without a normal distribution. When necessary, the Mann-Whitney U test with the Bonferroni correc- tion was used to compare variables. Univariate and multivariate Cox re- gression models were used to evaluate the independent relationships of in-hospital and 6-month mortalities with NLR values. Age, sex, conges- tive heart failure, coronary artery disease, diabetes mellitus, malignan- cy, corticoSteroid treatment, creatinine, hematocrit, WBC, APACHE II score, and SOFA scores were included in this model. This model was considered to be 11.9 cutoff value of NLR. The effect of the NLR on the survival of critically ill patients was investigated using the log-rank test. Kaplan-Meier survival estimates were calculated. The utility of the NLR in long-term mortality in critically ill patients was evaluated via receiver operating characteristic curves, and the cutoff value was de- termined using Youden’s index. A P b .05 was considered statistically significant. Study power was calculated as 92% using a free Web site (http://www.statisticalsolutions.net/pss_calc.php).

Results

A total of 373 patients were included in the study. Of the 373 pa- tients, 203 were men (54.4%) and 170 were women (45.6%). The medi- an (IQR) age was 74 (19) years. The mean APACHE II score was 20.7 +- 10.1, the SOFA score was 4.9 +- 2.2, and the median (IQR) NLR was 6.72 (10.3). Demographics and clinical characteristics of the patients are pro- vided in Table 1.

In the comparison, according to NLR quartiles, there were no differ- ences detected for age, sex, APACHE II, SOFA score, GCS, comorbid dis- eases, mechanical ventilation, and duration of hospital stay (P .05). Because there was not any difference for ventilator-associated pneumo- nia time on the ventilator in terms of NLR quartiles, the development of MOF was at significantly higher levels in the fourth quartile than the other quartiles (P = .04) (Table 2).

Table 1

Demographics and laboratory findings of the study population

No. of patients	373
Age, y, median (IQR)	74 (19)
Female sex, n (%)	170 (45.6)
Diabetes mellitus, n (%)	153 (41.0)
Coronary artery disease, n (%)	78 (20.9)
Congestive heart failure, n (%)	83 (22.3)
Chronic renal failure, n (%)	40 (10.7)
Chronic obstructive pulmonary disease, n (%)	130 (34.9)
Malignancy, n (%)	21 (5.6)
Sepsis, n (%)	35 (9.4)
Corticosteroid treatment, n (%)	34 (9.1)
APACHE II score, mean +- SD	20.7 +- 10.1
SOFA score, mean +- SD	4.9 +- 2.2
GCS, median (IQR)	12 (8)
Creatinine, milligrams per deciliter, median (IQR)	1.17 (1.09)
Hematocrit, percent, mean +- SD	39.5 +- 8.8
WBC count, thousands per microliter, median (IQR)	12.8 (7.02)
Neutrophil, thousands per microliter, median (IQR)	9.0 (6.1)
Lymphocyte, thousands per microliter, median (IQR)	1.3 (1.4)
NLR, median (IQR)	6.72 (10.3)
Length of stay in-hospital, d, median (IQR)	6.0 (9.1)

Table 2

Patients characteristics according to quartile of NLR

	b 3.48	3.48-6.73	6.74-13.6	N 13.6	P
Age, y, median (IQR)	70 (25)	73.5 (29)	74 (19)	77 (15)		.06
Female sex, n (%)	48 (51.1)	44 (47.3)	38 (40.4)	40 (43.5)		.49
Diabetes mellitus, n (%)	39 (41.5)	43 (46.2)	28 (29.8)	43 (46.7)		.66
Coronary artery	25 (26.6)	15 (16.1)	14 (14.9)	24 (26.1)		.08
disease, n (%)
Congestive heart 26 (27.7)		18 (19.4)	16 (17)	23 (25)	.27
failure, n (%)
Chronic renal	10 (10.6)	5 (5.4)	14 (14.9)	11 (12)	.20
failure, n (%)
Malignancy, n (%)	2 (2.1)	6 (6.5)	6 (6.4)	7 (7.6)	.30
Corticosteroid	8 (8.5)	9 (9.6)	7 (7.4)	10 (10.8)	.59
treatment, n (%)
APACHE II score,	21.3 +- 10.1	19.8 +- 9.6	20.7 +- 9.9	20.9 +- 8.9	.75
mean +- SD
SOFA score, mean +- SD	4.4 +- 2.2	5.21 +- 2.4	4.9 +- 2.2	4.8 +- 2.0	.27
GCS, median (IQR)	13 (12)	12 (8)	10 (11)	13 (5)	.38
Development of	22 (23.7)	16 (17.2)	17 (18.1)	29 (31.5)	.07
sepsis, n (%)
Creatinine, milligrams	1.09 (1.01)	1.06 (0.98)	1.17 (1.11)	1.52 (1.32)	.02
per deciliter,
median (IQR)
Hematocrit, percent,	39.5 +- 9.3	39.6 +- 9.4	39.5 +- 8.5	39.8 +- 7.8	.99
mean +- SD
WBC count, thousands	10.2 (8.02)	10.4 (6.15)	11.6 (6.41)	13.6 (7.78)	b.001
per microliter,
median (IQR)
Application of MV,	50 (53.2)	38 (40.9)	44 (46.8)	53 (57.6)	.11
n (%)
Duration of MV, d,	1 (6)	0.5 (4)	1 (2)	1 (5)	.62
median (IQR)

Fig. 1. Receiver operating characteristic curve of NLR to predict mortality.

Ventilator-associated

pneumonia, n (%)
MOF, n (%)	62 (65)	50 (53.7)	51 (54.2)	65 (70.7)	.04
Length of stay	6 (9)	6 (10)	5 (8)	7 (9)	.52

in-hospital,

d, median (IQR)

21 (22.6) 14 (15.1) 14 (14.9) 19 (20.7) .41

significantly different (P = .04). When the in-hospital mortalities were compared according to the NLR quartiles, the fourth quartile (52 [31.7%]) had significantly higher mortality rates than the first, sec- ond, and third quartiles (38 [23.2%], 39 [23.8%], and 35 [21.3%], respec- tively; P b .041). Survival rates for in-hospital mortality, calculated with the Kaplan-Meier curve, were almost significantly lower in the fourth quartile (log-rank, 7.82; P = .05), and survival rates were significantly lower for 6-month mortality (log-rank, 10.45; P = .01) (Fig. 2). In

Abbreviation: MV, mechanical ventilator.

There was not any correlation detected between the NLR and age, APACHE II, SOFA, duration of hospital stay, or duration of mechanical ventilation (P N .05).

Performance characteristics (area under the curve, sensitivity, spec- ificity, predictive value, and likelihood ratios) of the NLR for predicting mortality were calculated (Table 3). The cutoff value for mortality ob- tained using the ROC curve was 11.9 (sensitivity, 37.0%; specificity, 81.8%; area under the curve, 0.61; 95% confidence interval [CI], 0.55-

0.65; P = .01) (Fig. 1).

Sepsis development, mechanical ventilation, and ventilator- associated pneumonia were analyzed according to the cutoff value. When the NLR is less than 11.9, sepsis developed in 19.5% (50) of the pa- tients, whereas when the NLR is greater than or equal to 11.9, sepsis de- velopment significantly increased to 29.6% (34) (P = .03). However, there was not a significant difference detected for mechanical ventila- tion and ventilator-associated pneumonia development (P N .05).

During follow-up, 164 mortalities occurred in the hospital. When the dead and aLive patients were compared, NLR values were found to be

Table 3

Performance parameters of NLR as a predictor of mortality

Performance parameters	95% confidence interval
Sensitivity	37.04 (30.6-43.9)
Specificity	81.8 (74.8-87.6)
Positive predictive value	74.1 (64.7-82.1)
Negative predictive value	48.1 (41.9-54.3)
Positive likelihood ratio	2.04 (1.4-3.0)
Negative likelihood ratio	0.77 (0.7-0.9)
Receiver operating characteristic curve area	0.61 (0.55-0.65)

the multivariable Cox regression model, for in-hospital mortality and 6-month mortality, NLR was (Hazard ratio (HR), 1.637 [1.110-2.415]; P = .01 and HR, 1.585 [1.136-2.213]; P b .007, respectively). In this model, accepted cutoff value of NLR was 11.9. Acute Physiology and Chronic Health Evaluation II scores were detected as independent indi- cators (Table 4). In the Kaplan-Meier curve of patients with NLR more than 11.9, in-hospital mortality and 6-month mortality had significantly lower survival rates. (log-rank, 5.61; P = .002 and log-rank, 9.12; P =

.018, respectively) (Figs. 3 and 4).

Fig. 2. Comparison of Kaplan-Meier survival curves of NLR quartiles.

Table 4

The independent predictors of mortality by multivariate logistic regression analysis

Predictor	Hazard ratio	95% CI	P
In-hospital mortality
NLR	1.637	1.110-2.415	.01
APACHE II	1.084	1.061-1.108	b.001
180-Day mortality NLR	1.585	1.136-2.213	.007
APACHE II	1.082	1.064-1.100	b.001

Discussion

This study was a prospective cohort study evaluating the prognostic value of the NLR in the critically ill patient group. According to the data we obtained from this study, high NLR levels in critically ill patients are an independent indicator of in-hospital and 6-month mortality. Neutrophil-lymphocyte ratio levels more than 11.9 increase in- hospital mortality by 64% and increase 6-month mortality by 59%. Fur- thermore, the risks of MOF and sepsis development are higher with higher NLR values. To our knowledge, this study is the first prospective study that presents the prognostic value of the NLR in critically ill pa- tients in the emergency department.

The diagnosis and resuscitation of critical patients are the reason for emergency service. Each day, the number of patients admitted to emer- gency departments with complex geriatric syndromes or decompensa- tion of chronic diseases rises [16]. In particular, 19% of emergency department admissions consisted of elderly patients in 2000, and this rate is estimated to reach 34% by 2050 [17]. Most of our study group was elderly. Advancements in science and technology have prolonged life expectancy and increased the number of comorbid patients. In the study by Vinton et al [18], patients were grouped according to the num- ber of emergency department admissions they had in 1 year, and they found that patients who were admitted to the emergency department most frequently (>=10 visits/year) were the patients with chronic dis- ease and low socioeconomic status. Critical care services provided to this complex, high-volume patient group in the emergency department, and ICUs compose most health expenditures. In spite of scoring systems were improved, which are quality of intervention to patients can be evaluated and be able to predict prognosis, cheap and fast Biological markers are very less. Because interventions performed on critical pa- tients within the first several hours of admission play a very important

Fig. 3. Comparison of Kaplan-Meier survival curves of NLR cutoff for 6-month mortality.

Fig. 4. Comparison of Kaplan-Meier survival curves of NLR cutoff for in-hospital mortality.

role in reducing mortality and the first several hours in the emergency department are equal to 72 hours in the ICU, they are known as the “golden hour and silver day” proposal [3-5]. In this case, the importance of markers for prognosis increases.

White blood cell number is a frequently used parameter that is in- cluded in many scoring systems and is frequently used for the diagnosis and follow-up of disease in daily practice. White blood cells play a main role in the systemic inflammatory response [8]. Jilma et al [7] researched the changes in WBC types after inflammation and detected a 300% in- crease in circulating neutrophils, 96% decrease in monocytes, and 85% decrease in lymphocytes 4 to 6 hours after inflammation. Dionigi et al

[19] and Ayala et al [20] summarized the possible reasons of lymphocytopenia as induction of catecholamine, prolactin, and cortisol levels that increase with stress, margination of lymphocytes to the retic- uloendothelial system, and apoptosis. Menges et al [21] explain the pos- sible reasons for neutrophilia as demargination of neutrophils from the endothelium, delay of neutrophil apoptosis, and effect of growth factors on stem cells. Zahorec [8] reported a study conducted with 90 intensive care patients who developed neutrophilia and lymphocytopenia; how- ever, the NLR was more valuable, and the authors state that this ratio is a reliable parameter for reflecting the intensity of stress, evaluating sys- temic inflammatory response, and monitoring.

Recent studies regarding the NLR mostly suggest that the NLR is a prognostic factor for cardiovascular diseases. Uthamalingam et al [11] found that in acute decompensated heart failure, NLR at admission was related to 30-day mortality, and they recommended the use of NLR for risk stratification. Thus, it has been demonstrated in studies that NLR is a prognostic factor for acute coronary syndromes [9,10]. Kayrak et al

[12] found that NLR is an independent indicator of short-term mortality in pulmonary emboli at the NLR greater than or equal to 9.2 cutoff value (HR, 3.52 [1.36-9.10] 95% CI; P = .01). According to our literature search, we did not find a similar study for the critically ill patient group.

Our results demonstrate that in the critically ill patient group, NLR (HR, 1.637 [1.110-2.415]; P b .05 and HR, 1.585 [1.136-2.213]; P b .05)

and APACHE II (HR, 1.084 [1.061-1.108]; P b .001 and HR, 1.082

[1.064-1.100]; P b .001) can be independent indicators of short- and long-term mortality. The NLR is a simple, cheap, and rapidly available parameter. In our opinion, the NLR can be used to determine the prog- nosis of critical patients in the emergency department. Because it is rap- idly obtainable, the NLR can provide knowledge to perform the necessary interventions within the very important golden hours. For

us, this will be used as a biological marker to determine the prognosis on critical patients in emergency department.

Zahorec [8] demonstrated that the clinical improvement of major surgery, sepsis, and injury parallels an increase in the number of lym- phocytes and a decrease in the number of neutrophils, and persistent neutrophilia and lymphopenia may be related to development of MOF. Our results also demonstrated that the development of MOF in the fourth quartile was significantly higher (P = .03). For NLR less than 11.9, sepsis developed in 19.5% (50) of the patients, whereas for NLR greater than or equal to 11.9, sepsis development significantly increased to 29.6% (34) (P = .03). However, there was not a significant difference detected for mechanical ventilation and ventilator-associated pneumo- nia development (P N .05). The NLR may not only shed light on short- and long-term mortality but can also enlighten the physician to the de- velopment of MOF and sepsis.

Zahorec [8] performed the aforementioned study with oncology ICU patients who had sepsis or septic shock. Only 10% of our patient group was critical due to infectious causes. Therefore, the NLR cannot only be used in infectious cases as a prognostic factor but can also be used in noninfectious critical patient groups.

Limitations

Our study had several limitations. The first limitation was that the study was a single-center study. Our hospital is a regional hospital, where medical intensive care patients are frequently accepted. For ex- ample, chronic obstructive pulmonary disease patients are very com- mon in our region and constitute a major portion of the patients admitted to our emergency department, particularly in the winter months. For this reason, multicenter studies, including trauma centers, are needed for further studies.

The second limitation is that repetitive measurements of NLR were not performed. The effects of repetitive measurements on determining the effectiveness of treatment and the effect of the change on mortality are unknown.

Conclusion

This study investigated the prognostic value of the NLR in the critical patient group. The NLR, which is a simple measure, does not require ad- ditional costs, is rapidly accessible, and is an independent indicator of short- and long-term mortality. We suggest that the NLR can direct emergency department physicians toward interventions within several hours of admission in the critical patient group.

References

1. Meggs W, Czaplijski T, Benson N. Trends in emergency department utilizaition, 1988-1997. Acad Emerg Med 1999;6:1030-5.
2. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, et al. Early goal direct- ed therapy in the treatment of severe sepsis and septic shock. NEJM 2001;345: 1368-77.
3. Lundberg JS, Perl TM, Wiblin T, Costigan MD, Dawson J, Nettleman MD, et al. Septic shock: an analysis of outcomes for patients with onset on hospital wards versus in- tensive care units. Crit Care Med 1998;26:1020-4.
4. Garrard C, Young D. SubOptimal care of patients before admission to intensive care is caused by a failure to appreciate or apply the ABCs of life support (editorial). BMJ 1998;316:1841-2.
5. Blow O, Magliore L, Claridge JA, Butler K, Young JS. The golden hour and the silver day: detection and correction of occult hypoperfusion within 24 hours improves outcome from major trauma. J Trauma 1999;47:964-9.
6. Galus MA, Stern J. Extreme lymphocytopenia associated with toxic shock syndrome. J Intern Med 1998;244:351-4.
7. Jilma B, Blann A, Pernerstorfer T, Stohlawetz P, Eichler HG, Vondrovec B, et al. Regu-

   lation of adhesion molecules during human endotoxemia. Am J Respir Crit Care Med 1999;159:857-63.

   Zahorec. Ratio neutrophil to lymphocyte counts-rapid and simple parameter of sys- temic inflammation and stress in critically ill. Bratisl Lek Listy 2001;102:5-14.

8. Tamhane UU, Aneja S, Montgomery D, Rogers EK, Eagle KA, Gurm HS. Association between admission neutrophil to lymphocyte ratio and outcomes in patients with acute coronary syndrome. Am J Cardiol 2008;102:653-7.
9. Azab B, Zaher M, Weiserbs K, Estelle T. Usefulness of neutrophil to lymphocyte ratio in predicting short and long term mortality after non-ST elevation myocardial in- farction. Am J Cardiol 2010;106:470-6.
10. Uthamalingam S, Patvardhan EA, Subramanian S, Ahmed W, Martin W, Daley M, et al. Utility of the neutrophil to lymphocyte ratio in predicting long-term outcomes in acute decompensated heart failure. Am J Cardiol 2011;107:433-8.
11. Kayrak M, Erdogan HI, Solak Y, Akilli H, Gul EE, Yildirim O, et al. Prognostic value of neutrophil to lymphocyte ratio in patients with acute pulmonary embolism: a retro- spective study. Heart Lung Circ 2014;23:56-62.
12. Nguyen HB, Rivers EP, Havstad S, Knoblich B, Ressler JA, Muzzin AM, et al. Critical care in the emergency department: a physiologic assessment and outcome evalua- tion. Acad Emerg Med 2000;7:1354-61.
13. Knaus Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med 1985;13:818-29.
14. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, Bruining H, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/fail- ure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 1996;22:707-10.
15. Arendts G, Lowthian J. Demography is destiny: an agenda for Geriatric EMergency medicine in Australasia. Emerg Med Australas 2013;25:271-8.
16. Martin-Khan M, Burkett E, Schnitker L, Jones RN, Gray LC. Methodology for develop- ing quality indicators for the care of older people in the emergency department. BMC Emerg Med 2013;13:23-33.
17. Vinton DT, Capp R, Rooks SP, Abbott JT, Ginde AA. frequent users of US emergency de- partments: characteristics and opportunities for intervention. Emerg Med J 2014:1-7.
18. Dionigi R, Dominioni L, Benevento A, Giudice G, Cuffari S, Bordone N, et al. Effects of surgical trauma of laparoscopic vs. open cholecystectomy. Hepatogastroenterology 1994;41:471-6.
19. Ayala A, Herdon CD, Lehman DL, Ayala CA, Chaudry IH. Differential induction of ap- optosis in lymphoid tissues during sepsis: variation in onset, frequency, and the na- ture of the mediators. Blood 1996;15:4261-75.
20. Menges T, Engel J, Welters I, Wagner RM, Little S, Ruwoldt R, et al. Changes in blood lymphocyte populations after multiple trauma: association with posttraumatic com- plications. Crit Care Med 1999;27:733-40.