Indian J Med Res 152, October 2020, pp 368-377 Quick Response Code:

DOI: 10.4103/ijmr.IJMR_595_18

Comparison of pneumonia severity scores for patients diagnosed with

pneumonia in emergency department

Aynur Ecevit Kaya1, Seda Ozkan2, Eren Usul3 & Engin Deniz Arslan4

1
Department of Emergency Medicine, Bursa City Hospital, Bursa, 2Department of Emergency Medicine, Istanbul
University-Cerrahpasa, Cerrahpasa Faculty of Medicine, Istanbul, 3Department of Emergency Medicine, Sincan
State Hospital, Ankara & 4Department of Emergency Medicine, Antalya Training & Research Hospital,
Antalya, Turkey

Received March 24, 2018

Background & objectives: Sepsis due to pneumonia or pneumonia itself is one of the main causes of deaths
in patients despite the advanced treatment methods. The optimal prognostic tool in pneumonia is still
not clear. This study was aimed to compare the pneumonia severity scores and the possibility of using the
new scores in patients who were diagnosed with pneumonia in the emergency department.
Methods: Demographic data, laboratory and imaging results, confusion, elevated blood urea nitrogen,
respiratory rate and blood pressure plus age ≥65 yr (CURB-65), pneumonia severity index (PSI), national
early warning score (NEWS), NEWS-lactate (NEWS-L) scores, hospitalization, referral, discharge
and 30-day mortality of patients who were diagnosed with pneumonia in emergency department were
recorded.
Results: A total of 250 patients were included in the study. The most successful score in predicted mortality
was found to be NEWS-L. This was followed by NEWS, CURB-65 and PSI, respectively. Most successful
scores in anticipation of admission to the intensive care unit were NEWS-L followed by NEWS. This was
followed by CURB-65 and PSI scores, respectively. The most successful score in anticipation of hospital
admission was NEWS-L, followed by NEWS, CURB-65 and PSI, respectively. There was a significant
difference between all pneumonia severity scores of the patients who died and survived within 30 days.
There was a significant difference between the scores of patients in intensive care unit (ICU) and service,
compared to non-ICU patients.
Interpretation & conclusions: NEWS-L score was found to be the most successful score in predicting
mortality, ICU admission and hospitalization requirement. Both NEWS-L and NEWS scores can be used
in determining the mortality, need for hospitalization and intensive care of the patients with pneumonia
in the emergency department.

Key words Emergency department - hospitalization - mortality - pneumonia - pneumonia severity scores - prognosis

Pneumonia is responsible for a significant loss of work and school days and deaths1-3. Various
proportion of hospital admissions, treatment costs, scores have been sought to determine the relationship

© 2020 Indian Journal of Medical Research, published by Wolters Kluwer - Medknow for Director-General, Indian Council of Medical Research
368
 KAYA et al: PNEUMONIA SEVERITY SCORES IN EMERGENCY DEPARTMENT 369

between the clincial findings and the laboratory tests laboratory findings were included. Those with hospital-
of the patients with the severity of pneumonia and thus acquired pneumonia, aspiration pneumonia, pulmonary
predict the need for hospitalization, admission into tuberculosis, pulmonary oedema, and pulmonary
intensive care and/or mortality4,5. Confusion, elevated thromboembolism were excluded. During the study
blood urea nitrogen, respiratory rate and blood period, 314 patients diagnosed with pneumonia were
pressure plus age ≥65 yr (CURB-65) and pneumonia recruited. Thirteen patients with hospital-acquired
severity index (PSI) scores are the most commonly pneumonia and 29 patients with aspiration pneumonia
used scores6,7. PSI score assesses hospitalization and were excluded from the study. A total of 272 patients
mortality by evaluating the patient’s demographic who met the criteria were included. Twenty two
information, clinical findings, laboratory findings and patients who met the study criteria were lost to
co-morbid illnesses and categorizes the patient in a follow up. Vital signs, biochemistry, complete
risk group of 1-55. CURB-65 evaluates the patients’ haemogram, arterial blood gas values, chest X-ray
consciousness, blood urea level, systolic blood pressure, findings, accompanying diseases, consciousness
respiratory rate and age6. Studies3-8 comparing PSI and status, age, gender and treatment of the patients were
CURB-65 have shown that short-term mortality is recorded. CURB-65, PSI, NEWS and NEWS-L scores
more consistently determined by PSI. However, PSI is of the patients were calculated. All variables requiring
not as easy to calculate as CURB-65 and its variables the calculation of the score systems were obtained
are different8. from the routine examinations of the patients. Patients
National early warning score (NEWS) is a relatively who were diagnosed with pneumonia in the emergency
new score that includes systolic blood pressure, department were discharged home, hospitalized into
respiratory rate, heart rate, fever, consciousness and the general ward or admitted into ICU depending on
oxygen saturation level9. NEWS-lactate (NEWS-L) is their clinical conditions. The follow up and treatment
another scoring system, in which the data of the NEWS of patients included in this study were not affected by
score and the lactate level are evaluated together9. the risk score calculated. The intensive care admission
The purpose of establishing these two scores was to criteria (threatened airway, respiratory rate ≥40 or ≤8
determine the clinical prognosis in a practical way. breaths/min, oxygen saturation <90 per cent on ≥50
Better results were obtained with these two scoring per cent, respiratory and cardiac arrests, pulse rate
systems, which included vital signs10. <40 or >140 beats/min, systolic blood pressure <90
The number of studies comparing NEWS and mmHg, fall in the level of consciousness, rising arterial
NEWS-L with other scores is limited. This study carbon dioxide with respiratory acidosis) were used
was aimed to compare pneumonia severity scores in in determining whether intensive care was required.
patients who were diagnosed with pneumonia in the The hospital records and national healthy notification
emergency department and to investigate the utility of system were used for investigating the outcome at
new scores. 30 days. Telephone call was made for follow up if
no record was found from the above two systems.
Material & Methods Calculation of risk scores is shown in Tables I and II.
This study was carried out at the Emergency Statistical analysis: Statistical Package for the Social
department of Ankara Dışkapı Yıldırım Beyazıt Sciences (SPSS) programme (IBM SPSS Statistics
Training and Research Hospital, Ankara, Turkey. The for Windows, version 17.0, SPSS Inc, Chicago,
study was approved by the local ethical committee IL) was used to analyze the obtained information.
and written informed consent was obtained from each Kolmogorov-Smirnov test was used in groups with
participant. All consecutive patients over 18 years of >30 patients in the analysis of normality of the data,
age who were diagnosed with community-acquired and the Shapiro-Wilk test was used in those <30. The
pneumonia between October 1, 2015 and May 1, 2016 Mann-Whitney U-test was used to compare data that did
were included in this study. The minimum sample size
not fit the normal distribution, and the Student’s t test
required in this study was 234.
was used to compare data with a normal distribution.
Patients with new or worsening of pre-existing
infiltration in the chest X-ray and at least two of the Med Calc statistic software programme (Med Calc
symptoms associated with pneumonia (cough, sputum, Software, version 15, Mariakerke, Belgium) was used
dyspnoea and pleuritic chest pain) with significant for the analysis of receiver operating characteristic
370 INDIAN J MED RES, OCTOBER 2020

Table I. NEWS versus NEWS-L used in the present study

Physiologic component 3 2 1 0 1 2 3
SBP (mmHg) ≤90 91-100 101-110 111-219 ≥220
PR (beats/min) ≤40 41-50 51-90 91-110 111-130 ≥131
RR (breaths/min) ≤8 9-11 12-20 ≥25
Body temperature (°C) ≤35 35.1-36 36.1-38 38.1-39 ≥39, 1
SpO2 (%) ≤91 92-93 94-95 96
Any supplemental oxygen Yes No
Level of consciousness Alert Voice, pain, unresponsive
Laboratory component
Lactate level When calculating the NEWS-L score, serum lactate level is added on the score of the NEWS score
NEWS risk class: Score
I: 0-4 Low risk
II: 5-6 Moderate risk
III: ≥7 High risk
NEWS-L risk class: Score
I: 0-3 Low risk
II: 3.1-5.2 Low risk
III: 5.3-8.0 Moderate risk
IV: ≥8.1 High risk
SBP, systolic blood pressure; PR, pulse rate; RR, respiratory rate, SpO2, peripheral oxygen saturation; NEWS, national early warning
score; NEWS-L, national early warning score-lactate. Source: Ref. 4

(ROC) curves to compare the performance of the 23.6 per cent (n=59) of the patients were ex-smokers
pneumonia risk scores in predicting mortality, need of and 6.8 per cent (n=17) were active smokers.
admission into hospital and intensive care unit (ICU).
Pleural effusion on chest radiograph was detected
The area under the ROC curves (AUC) was calculated.
in 12.8 per cent (n=32) patients, lobar infiltration
Variables with risk scores were defined if a two-sided P
was detected in 74.8 per cent (n=187) patients and
value was 0.05 or less; 95 per cent confidence intervals
bilateral infiltration was detected in 20.4 per cent
(CIs) were calculated. The method of DeLong et al11
(n=51) patients (Table III). Bronchodilator therapy
was used to calculate the standard error of the area
was started with antibiotics in 246 of the patients in
under the curve. An exact binomial Ci was calculated
the Emergency department. One hundred and one
for the area under the curve. Youden index was used
patients were discharged from emergency services
to determine the optimal cut-off value, sensitivity and
with medical therapy. One hundred and forty nine
specificity on a ROC curve (Youden index J is defined as
patients were hospitalized, 69 were admitted to the
J=max (sensitivityc+specificityc−1). The cut-point that
wards, while 80 patients were admitted to the ICU.
achieved this maximum was referred to as the optimal
Mortality was seen in 27 patients during the 30 days
cut-point (c). Correlation tests of normally distributed
of follow up period.
data were performed by Spearman correlation test.
The mean values of the risk scores in the three
Results
outcome groups are shown in Table IV. NEWS-L was
The study included 250 patients. Mean age the most successful in predicting 30-day mortality with
was 72.3±14.25 yr and 41.6 (n=104) per cent of the a value of 0.96 AUC (95% CI: 0.928-0.981, cut-off
patients were females chronic obstructive pulmonary value: 13.7). This was followed by NEWS, CURB-65
disease (COPD) was the most common associated and PSI, respectively (Fig. 1 and Table V). Pneumonia
(39.2%) comorbid diseases (Table III). Of these, severity scores were compared between those who died
 KAYA et al: PNEUMONIA SEVERITY SCORES IN EMERGENCY DEPARTMENT 371

Table II. CURB-65 and pneumonia severity ındex

CURB-65 Points
Confusion 1
Blood urea nitrogen >20 mg/dl 1
Respiratory rate >29 breaths per minute 1
SBP <90 mmHg or diastolic blood pressure ≤60 mmHg 1
Age ≥65 yr 1
Score
0-1: Low risk Outpatient
2: Moderately severe Hospitalization required
3-5: Severe pneumonia Hospitalization required; consider ICU admission
PSI
Characteristic Points
Age-men yr
Age-Women yr-10
Nursing home resident 10
Comorbidity
Neoplastic disease 10
Liver disease 20
Congestive heart disease 10
Cerebrovascular disease 10
Renal disease 10
Examination
Altered mental state 20
Respiratory rate ≥30 breaths per minute 20
SBP <90 mmHg 20
Temperature <35°C or ≥40°C 15
Pulse ≥125 10
Laboratory findings
Arterial pH <7.35 30
Urea ≥30 mg/dl 20
Na <130 mg/dl 20
Glucose ≥250 mg/dl 10
Hct <30% 10
PaO2 <60 mmHg or SaO2 (air) <90% 10
Pleural effusion 10
PSI risk class: Score
I: <50 Outpatient
II: 51-70 Outpatient
III: 71-90 Outpatient or short inpatient
IV: 91-130 Hospitalization required
V: >130 Hospitalization required
PSI, pneumonia severity index; Hct, haematocrit; PaO2, partial oxygen pressure. Source: Ref. 3
 Table III. Demographic characteristics, vital signs and laboratory findings of the cases
372
Parameters All patients Survived Dead P Hospitalized Non- P ICU Non-ICU P
hospitalized admission admission
Age† 76;20 76; 20 77; 18 0.46 78;18 73;19 0.01 78; 16 76; 22 0.02
#
SBP (mmHg) 125.1±27.4 127.5±26.2 104.4±30.6 0.001 122.5±28.8 128.7±25.6 0.03 117.4±32.0 128.5±32.2 0.01
DBP (mmHg)# 73.3±15.4 74.9±14.7 60.1±16.1 0.001 72.8±16.1 74.0±14.7 0.52 70.5±17.5 75.4±13.9 0.16
#
PR (beats/min) 97.5±22.7 95.42±20.76 114.1±27.8 0.001 97.9±23.2 96.8±21.1 0.81 102.9±26.6 94.9±19.6 0.01
RR (breaths/min)† 22;4 21; 4 27; 12 0.001 22;5 20;2 0.001 24; 6 22;2 0.001
Body temperature (°C)# 37.2±0.8 37.3±0.8 37.2±0.8 0.43 37.3±0.8 37.3±0.8 0.71 37.3±0.8 37.2±0.8 0.50
#
SpO2 (%) 87.1±9.1 88.4±8.0 78.1±12.3 0.001 85.2±10.2 90.3±6.2 0.001 83.2±11.6 83.3±11.6 0.08
AVPU# 1;0.699 1; 1 2; 1 0.001 1;1 1;0 0.001 2; 1 1;0 0.001
Blood urea nitrogen (mg/dl)# 65.5±46.2 61.4±43.4 96.2±56.7 0.001 74.5±47.0 51.4±41.6 0.001 83.9±52.1 63.5±27.7 0.01
Creatinine (mg/dl)# 1.3±0.8 1.3±0.7 1.8±1.1 0.01 1.4±0.8 1.2±0.7 0.09 1.5±0.9 1.2±0.5 0.29
#
Glucose (mg/dl) 151.1±69.2 149.9±68.9 164.6±76.6 0.53 155.8±72.8 145.2±64.8 0.17 154.8+71.1 156.9±71.1 0.98
ALT (IU/l)# 39.6±125.4 31.7±48.2 102.4±355.4 0.04 50.7±162.7 22.9±23.5 0.02 61.5±211.8 49.5±119.9 0.02
#
AST (IU/l) 47.9±140.0 38.7±73 122±369.5 0.001 62.9±181.3 25.8±16.6 0.001 74.2±220.5 37.9±66.2 0.24
#
Sodium (mmol/l) 137±6.5 136.6±6.3 140.2±7.6 0.02 137.4±7.6 136.4±4.6 0.17 138.5±9.0 136.1±5.2 0.06
pH# 7.36±0.10 7.37±0.09 7.29±0.12 0.001 7.37±0.11 7.39±0.07 0.001 7.32±0.13 7.37±0.08 0.001
pCO2 (kPa)# 43.2±17.3 43.3±17.2 43.1±18.9 0.5 45.9±20.7 39.2±9.4 0.01 48.3±25.8 48.3±12.0 0.59
pO2 (kPa)# 57.8±23.6 58.8±23.6 49.8±23.1 0.01 55.3±24.2 61.5±22.5 0.001 58.5±21.5 54.0±19.3 0.65
#
HCO3 (mmol/l) 24.6±6.1 25.0±6.0 20.4±6.1 0.001 24.7±6.6 24.3±5.3 0.97 23.7±7.1 25.8±6.0 0.05
Lactate (mmol/l)# 2.7±2.0 2.4±1.5 5.2±3.5 0.001 3.0±2.4 2.2±1.2 0.02 3.5±2.8 2.4±1.8 0.001
Base deficit (mmol/l)# -0.2±5.51 3.7±3.6 5.8±5.3 0.001 4.2±4.0 3.5±3.7 0.05 5.3±4.7 3.0±2.5 0.001
INDIAN J MED RES, OCTOBER 2020

SAT O2 (%)# 81.7±12.1 82.6±11.3 73.4±15.6 0.001 79.5±12.8 84.6±10.5 0.001 78.8±14.3 80.3±10.9 0.94
#
WBC (µl) 12611.4±6630.1 12873.9±6641.4 11301.0±6272.3 0.43 12862.5±6687.4 12461.6±6515.2 0.48 12752±6891.7 12992.5±6487.5 0.96
#
Haematocrit (%) 38.7±6.8 38.8±6.5 38.0±8.6 0.42 38.0±6.8 39.8±6.7 0.15 37.6±7.5 38.5±5.9 0.38
Haemoglobin (g/dl)# 12.6±2.2 12.6±2.0 12.3±2.9 0.56 12.4±2.1 12.9±2.1 0.061 12.3±2.3 12.4±1.9 0.6
Comorbidity, n (%)
COPD pulmonary disease, n 98 (39.2) 91 (92.9) 7 (7.1) 0.13 55 (56.1) 43 (43.9) 0.37 22 (40) 33 (60) 0.01
(%)
Hypertension, n (%) 97 (38.8) 86 (88.7) 11 (11.3) 0.83 62 (63.9) 35 (36.1) 0.27 33 (53.2) 29 (46.8) 0.92
Diabetes mellitus, n (%) 61 (24.4) 53 (86.9) 8 (13.1) 0.50 45 (73.8) 16 (26.2) 0.01 18 (40) 27 (60) 0.03
Cardiovascular disease, n (%) 45 (18) 37 (82.2) 8 (17.8) 0.09 28 (62.2) 17 (37.8) 0.69 15 (53.6) 13 (46.4) 0.98
Contd...
 KAYA et al: PNEUMONIA SEVERITY SCORES IN EMERGENCY DEPARTMENT 373

and survived within 30 days of follow up. Significant

Median, IQR; #mean±SD. DBP, diastolic blood pressure; COPD, chronic obstructive pulmonary disease; ALT, alanine transaminase; AST, aspartate transaminase;
0.28 difference was found in all scores (Table IV). There was
0.01
0.75
0.01
0.46
0.01
0.14

0.52
0.25
0.19
P

a significant difference between the other parameters

except age, fever, glucose, pCO2, white blood cell
admission
Non-ICU

12 (57.1)

21 (67.7)

14 (51.9)
54 (49.1)
12 (36.4)
(WBC), haemoglobin (Hb) and haematocrit (Hct)
2 (12.5)
6 (55.5)

9 (69.2)
1 (100)

HCO3, bicarbonate; AVPU, alert-verbal-pain-unresponsive; SAT O2, oxygen saturation; pCO2, partial pressure of carbon dioxide; pO2, partial pressure of oxygen
values with regard to 30-day mortality (Table III).
0

ROC curve analysis was performed for admission

to the hospital for pneumonia severity scores. The
highest AUC score was for NEWS-L (AUC: 0.72,
admission

14 (87.5)

10 (32.3)

13 (48.1)
56 (50.9)
21 (63.6)
9 (42.9)

5 (45.5)

4 (30.8)
7 (100)
ICU

CI: 0.659-0.774, cut-off value: >7.7), while the

0

highest sensitivity was shown by NEWS (77.1%) and

the highest specificity by NEWS-L score (63.4%)
(Fig. 2 and Table VI). ROC curve analysis was
0.32
0.83

0.42
1.00
0.20
0.20

0.00
0.66
0.40
0.9
P

performed to predict the need for intensive care for

all scores. The highest AUC was shown by NEWS
hospitalized
10 (32.3)
10 (38.5)

28 (47.5)

77 (41.2)
18 (35.3)

(0.86) and NEWS-L (0.86), followed by CURB-65

8 (42.1)
8 (53.3)

4 (23.5)

5 (15.6)
1 (50)
Non-

(0.85). NEWS-L had the highest sensitivity (90%)

and CURB-65 had the highest specificity (94.7%)
(Fig. 3 and Table VII).
Hospitalized

The pneumonia severity scores of the patients

110 (58.8)
21 (67.7)
16 (61.5)

31 (52.5)
13 (76.5)

27 (84.4)

33 (64.7)
11 (57.9)
7 (46.7)
1 (50)

admitted to the ICU and ward (non-ICU) were

compared, and significant difference was detected
among the scores except PSI (Table IV). The pneumonia
severity scores of hospitalized patients and patients
0.76
0.43
1.00
0.21
1.00
0.05
0.70

1.00
0.30
0.80
P

discharged home were also and a significant difference

was observed between all the risk scores (Table IV).
4 (12.9)
4 (15.4)
2 (10.5)

18 (9.6)
2 (11.8)

6 (11.8)
2 (3.4)

3 (9.4)
3 (20)
Dead

169 (90.4)
27 (87.1)
22 (84.6)
17 (89.5)

57 (96.6)
15 (88.2)

29 (90.6)

45 (88.2)
Survived

12 (80)
2 (100)
All patients

187 (74.8)
31 (12.4)
26 (10.4)

59 (23.6)

32 (12.8)

51 (20.4)
19 (7.6)

17 (6.8)
2 (0.8)
15 (6)
Cerebrovascular disease, n (%)
Congestive heart failure, n (%)

Chronic renal disease, n (%)

Chronic liver disease, n (%)
Neoplastic disease, n (%)

Bilateral opacities, n (%)

Pleural effusion, n (%)
Lobar opacities, n (%)
Radiographic findings
Current smoker (%)
Ex-smoker, n (%)
Parameters

Fig 1. Comparison of pneumonia severity scores receiver operating

characteristic curves for 30-day mortality. CURB-65, confusion,
elevated blood urea nitrogen, respiratory rate and blood pressure
plus age ≥65 yr; NEWS, national early warning score; NEWS-L,
†

national early warning score-lactate; PSI, pneumonia severity index.

 374

Table IV. Distribution of scores according to patient results

Scores All Dead Survived P ICU Non-ICU P Hospitalized Non- P
(n=27) (n=223) admission admission (n=149) hospitalized
(n=80) (n=69) (n=101)
CURB-65 (mean±SD) 2.1±0. 8 3.11±0.75 1.96±0.75 0.001 2.81±0, 79 1.76±0.54 0.001 2.32±0.86 1.73±0.63 0.001
Class 1 54 (21.6) 1 (3.7) 53 (23.8) 5 (6.3) 20 (29) 25 (16.8) 29 (28.7)
Class 2 130 (52) 2 (7.4) 128 (57.4) 18 (22.5) 47 (68.1) 65 (43.6) 65 (64.4)
Class 3 66 (26.4) 24 (88.9) 42 (18.8) 57 (71.3) 2 (2.9) 59 (39.6) 7 (6.9)
PSI (mean±SD) 123.5±26.2 139.18±23.18 121.59±25.86 0.001 132.11±23.78 125.37±22.17 0.11 128.99±23.22 115.37±28.23 0.001
Class 1 14 (5.6) 1 (3.7) 13 (5.8) 2 (2.5) 2 (2.9) 4 (2.7) 10 (9.9)
Class 2 1 (0.4) - 1 (0.4) - 1 (1.4) 1 (0.8) -
Class 3 16 (6.4) - 16 (7.6) 5 (6.3) 5 (7.2) 10 (6.7) 6 (5.9)
Class 4 149 (59.6) 12 (44.4) 137 (61.4) 39 (48.8) 43 (62.3) 82 (55) 67 (66.3)
Class 5 70 (28) 14 (51.9) 56 (25.1) 34 (42.5) 18 (26.1) 52 (34.9) 18 (17.9)
NEWS (mean±SD) 5.9±3.1 10.29±2.50 5.38±2, 73 0.001 8.62±2.63 4.85±2.13 0.001 6.87±3.05 4.49±2.61 0.001
Class 1 79 (31.6) - 79 (35.4) 2 (2.5) 26 (37.7) 28 (18.8) 51 (50, 5)
Class 2 61 (24.4) 1 (3.7) 60 (26.9) 11 (13.8) 26 (37.7) 37 (24.8) 24 (23.8)
INDIAN J MED RES, OCTOBER 2020

Class 3 110 (44) 26 (96.3) 84 (37.7) 67 (83.8) 17 (24.6) 84 (56.4) 26 (25.7)

NEWS-L (mean±SD) 8.6±4.0 15.48±3.34 7.75±3.20 0.001 12.06±3.80 7.31±2.91 0.001 9.86±4.15 6.71±2.92 0.001
Class 1 16 (6.4) - 16 (7.2) - 3 (4.3) 3 (2) 13 (12.9)
Class 2 38 (15.2) - 38 (17) - 12 (17.2) 12 (8) 26 (25.7)
Class 3 61 (24.4) - 61 (27.4) 8 (10) 27 (39.1) 35 (23.5) 26 (25.7)
Class 4 135 (54) 27 (100) 108 (48.4) 72 (90) 27 (39.1) 99 (66.4) 36 (35.6)
Score values shown as n (%)
 KAYA et al: PNEUMONIA SEVERITY SCORES IN EMERGENCY DEPARTMENT 375

Fig 2. Comparison of pneumonia severity scores receiver operating Fig 3. Comparison of pneumonia severity scores receiver operating
characteristic curves for hospitalization. characteristic curves for intensive care unit needs.

Table V. Receiver operating characteristic analysis for 30-day mortality forecast of pneumonia severity scores
Scores AUC SE 95% Cl Sensitivity (%) Specificity (%) Cut-off Youden index J P
CURB-65 0.86 0.037 0.809-0.899 88.9 80.7 >2 0.6961 <0.001
NEWS 0.91 0.031 0.869-0.943 96.3 75.8 >7 0.7208 <0.001
NEWS-L 0.96 0.014 0.928-0.981 85.1 96.4 >13.7 0.8160 <0.001
PSI 0.71 0.054 0.645-0.762 66.7 73.5 >136 0.4021 <0.001
AUC, area under the curve; SE, standard error; CI, confidence interval

Table VI. Receiver operating characteristic curve analysis for hospitalization of pneumonia severity scores
Scores AUC SE 95% Cl Sensitivity (%) Specificity (%) Cut-off Youden index J P
CURB-65 0.69 0.029 0.627-0.745 40.3 60 >2 0.3334 <0.001
NEWS 0.71 0.032 0.655-0.771 77.1 53.5 >4 0.3065 <0.001
NEWS-L 0.72 0.032 0.659-0.774 69.1 63.4 >7.7 0.3249 <0.001
PSI 0.64 0.035 0.580-0.702 69.8 54.5 >117 0.2425 0.001

Table VII. Receiver operating characteristic analysis of scores for ıntensive care needs
Scores AUC SE 95% Cl Sensitivity (%) Specificity (%) Cut-off Youden index J P
CURB-65 0.85 0.027 0.798-0.891 72.5 94.7 >2 0.6721 <0.001
NEWS 0.86 0.023 0.812-0.901 82.5 73.5 >6 0.5603 <0.001
NEWS-L 0.86 0.023 0.809-0.699 90 62.9 >8 0.5294 <0.001
PSI 0.64 0.037 0.577-0.699 71.25 55.29 >122 0.2654 0.001
376 INDIAN J MED RES, OCTOBER 2020

Discussion were consistent with these studies. All these studies

indicate that lactate level plays an important role in
In the ROC analysis for 30-day mortality
the prognosis of patients with pneumonia. As the
prediction, the NEWS-L score performed the best with
NEWS-L score includes lactate acid, it performs better
a value of 0.96 AUC. It was followed by NEWS and
than other score in predicting the 30-day mortality,
CURB-65 scores, respectively. In a study conducted
need for hospitalization and intensive care. There was
by Jo et al4, the NEWS-L score was found to be
a significant difference between all the scores when
the first rank and the NEWS score was the second
the pneumonia severity scores of the patients who
rank, consistent with our results. CURB-65 and PSI were hospitalized were compared with those who
performed moderately with the latter having a better received medical therapy at home. These findings were
AUC than the former4. However, in our study, the PSI consistent with the study by Chen et al12 for CURB-65.
score was the weakest predictor of mortality. In the
PSI score, patients with neoplastic disease history and Our study had some limitations. COPD patients
liver disease were in high risk group5. It was presumed may need additional oxygen as part of the severity
that the blood glucose, Hct and sodium levels used in of the disease/acute exacerbation. In our study, the
the evaluation of PSI score were normal at the time of presence of 39 per cent COPD patients might have
our study, which categorized the patients as low risk. played a role in the success of NEWS and NEWS-L
Chen et al12 have found that the use of the CURB-65 scores more than other scores (PSI and CURB-65).
score alone predicts the mortality poorly. In a study by Most validation studies for NEWS score were single-
Gwak et al13, the mortality rate of patients with PSI centre studies with a short follow up of patients. Thus,
scores grades 1 and 2 was 2.4 per cent. Jo et al4 found external validity and long-term predictive ability of
that vital signs constituted more significant effects in NEWS remains unknown. As in other studies, only 30-
the NEWS and NEWS-L scores than in the PSI and day results were evaluated in our study. Furthermore,
CURB-65 scores, which led to better performance in NEWS was not designed to a single time point tool
predicting 30-day mortality. In our study, significant but rather a ‘track-and-trigger’ system in individual
differences were found in all scores when pneumonia patients. Accuracy of NEWS may thus be different
severity scores of patients who died in 30 days as if multiple measurements at different time points are
with that who survived. Jo et al4 found significant considered.
differences between the dead and survivors for NEWS, In conclusion, NEWS-L score was found to be
NEWS-L, CURB-65 and PSI scores. Gwak et al13 also the most successful score in predicting mortality, and
compared the PSI score in the deceased and surviving requirement for ICU admission and hospitalization
groups and found a significant difference. Our results in our study. NEWS-L and NEWS scores thus can
were consistent with these studies. be helpful in determining the mortality, need for
In the ROC analysis for hospitalization of hospitalization and intensive care of the patients with
pneumonia severity scores, the NEWS-L score ranked pneumonia in the emergency department.
first with an AUC of 0.72, followed by NEWS, CURB- Financial support & sponsorship: None.
65 and PSI scores, respectively. Chen et al12 calculated
an AUC of 0.61 for CURB-65 in ROC analysis for Conflicts of Interest: None.
hospitalization. This rate was comparable with our
study. These results indicated that the NEWS and References
NEWS-L scores predicted hospitalization better than 1. Özlü T, Bülbül Y, Alataş F, Arseven O, Sakar CA, Cilli A, et al.
the other scores. NEWS and NEWS-L scores ranked Turkish Thoracic Society Consensus Report on diagnosis and
first in ROC analysis of all pneumonia severity scores treatment of community acquired pneumonia. Turk Toraks
Dergisi 2009; 10 : 1-16.
in terms of ICU necessity. Chen et al12 found that ROC
2. Acute respiratory infections in under-fives: 15 million deaths
analysis value for intensive care requirement of the a year. Lancet 1985; 2 : 699-701.
lactate-CURB-65 combination was better than CURB-
3. Metlay JP, Waterer GW, Long AC, Anzueto A, Brozek J,
65 score alone. Crothers K, et al. Diagnosis and treatment of adults with
community-acquired pneumonia. An official clinical practice
There has been a significant difference between guideline of the American Thoracic Society and Infectious
lactate levels of deceased and alive patients in studies Diseases Society of America. Am J Respir Crit Care Med
conducted in patients with pneumonia12-14. Our results 2019; 200 : e45-67.
 KAYA et al: PNEUMONIA SEVERITY SCORES IN EMERGENCY DEPARTMENT 377

4. Jo S, Jeong T, Lee JB, Jin Y, Yoon J, Park B. Validation detecting adult inpatient deterioration. Resuscitation 2010;
of modified early warning score using serum lactate 81 : 932-7.
level in community-acquired pneumonia patients. 10. Jo S, Lee JB, Jin YH, Jeong TO, Yoon JC, Jun YK, et al.
The national early warning score-lactate score. Am J Modified early warning score with rapid lactate level in
Emerg Med 2016; 34 : 536-41. critically ill medical patients: The ViEWS-L score. Emerg
5. Kwok CS, Loke YK, Woo K, Myint PK. Risk prediction Med J 2013; 30 : 123-9.
models for mortality in community acquired pneumonia: A 11. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing
systematic review. Biomed Res Int 2013; 2013 : 504136. the areas under two or more correlated receiver operating
6. Lim WS, van der Eerden MM, Laing R, Boersma WG, characteristic curves: A nonparametric approach. Biometrics
Karalus N, Town GI, et al. Defining community acquired 1988; 44 : 837-45.
pneumonia severity on presentation to hospital: An 12. Chen YX, Li CS. Lactate on emergency department arrival as
international derivation and validation study. Thorax 2003; a predictor of mortality and site of care in pneumonia patients:
58 : 377-82. A cohort study. Thorax 2015; 70 : 404-10.
7. Angus DC, Wax RS. Epidemiology of sepsis: An update. Crit 13. Gwak MH, Jo S, Jeong T, Lee JB, Jin YH, Yoon J, et al. Initial
Care Med 2001; 29 : S109-16. serum lactate level is associated with inpatient mortality in
8. Fine MJ, Auble TE, Yealy DM, Hanusa BH, Weissfeld LA, patients with community-acquired pneumonia. Am J Emerg
Singer DE, et al. A prediction rule to identify low-risk patients Med 2015; 33 : 685-90.
with community-acquired pneumonia. N Engl J Med 1997; 14. Liu W, Peng L, Hua S. Clinical significance of dynamic
336 : 243-50. monitoring of blood lactic acid, oxygenation index and C
9. Prytherch DR, Smith GB, Schmidt PE, Featherstone PI. reactive protein levels in patients with severe pnuemonia. Exp
ViEWS - Towards a national early warning score for Ther Med 2015; 10 : 1824-28.

For correspondence: Dr Seda Ozkan, Department of Emergency Medicine, Istanbul University-Cerrahpasa, Cerrahpasa Faculty of
Medicine, Istanbul 34098, Turkey
e-mail: sedacil@gmail.com