The National Early Warning Score 2(NEWS2) to Predict Early Progression to Severe Community-Acquired Pneumonia

This study aimed to assess the predictive performance of the National Early Warning Score 2 (NEWS2) to identify the early progression to severe disease in patients with community-acquired pneumonia (CAP). A prospective-cohort study was conducted among patients with CAP admitted to a university hospital between October 2020 and December 2021. The endpoint of interest was the progression to severe CAP, defined as the requirement for a mechanical ventilator, a vasopressor, or death within 72 h after hospital admission. Among 260 patients, 53 (25.6%) had early progression to severe CAP. The median NEWS2 of the early progression group was higher than that of the non-progression group [8 (6–9) vs. 7 (5–8), p = 0.015, respectively]. The AUROC of NEWS2 to predict early progression to severe CAP was 0.61 (95% CI: 0.52–0.70), while IDSA/ATS minor criteria ≥ 3 had AUROC 0.56 (95% CI 0.48–0.65). The combination of NEWS2 ≥ 8, albumin level < 3 g/dL and BUN ≥ 30 mg/dL improved AUROC from 0.61 to 0.71 (p = 0.015). NEWS2 and IDSA/ATS minor criteria showed fair predictive-accuracy in predicting progression to severe CAP. The NEWS2 cut-off ≥ 8 in combination with low albumin and uremia improved predictive-accuracy, and could be easily used in general practice.


Introduction
Pneumonia is a major public-health problem in Thailand. In 2019, the Office of Epidemiology, Department of Disease Control, Ministry of Public Health Thailand found 254,876 pneumonia cases, representing an approximate rate of 384 cases per 100,000 population, with a mortality rate of 0.25 per 100,000 population. Two-thirds of pneumonia cases occurred in adults over 65 years and children under 5 years of age [1]. Communityacquired pneumonia (CAP) is caused by an infection that originates outside the hospital. The pathogens can be viruses, bacteria, and fungi, with a different prevalence in each age-group and immune status. The most common symptoms are fever, cough, shortness of breath, and chest pain. The symptoms of CAP can vary from mild symptoms to acute respiratory failure, depending on the type of pathogens, the severity of the disease, and the patient's original health background. The diagnosis is based on clinical signs and symptoms, along with chest radiographs indicating lung parenchymal abnormalities [2]. Determining which patients have severe CAP is important, because studies have shown that patients directly admitted to the intensive care unit (ICU) had a statistically significantly lower mortality-rate than those who received delayed intensive treatment [3]. 2

of 10
The CURB-65 score (confusion, uremia, respiratory rate, blood pressure, and age) is most commonly used to triage the site of care for patients with CAP; however, it seems to have favorable mortality-prediction but limited accuracy for identifying high-risk patients requiring intensive-care-unit (ICU) admission [4][5][6]. The official clinical-practice guidelines for diagnosis and treatment of adults with CAP 2019 approved by the American Thoracic Society (ATS) and Infectious Disease Society of America (IDSA) has recommended a level of care selection using the 2007 IDSA/ATS criteria. Patients who met one of two major criteria (the need for mechanical ventilation or vasopressors) or at least three of nine of the minor criteria were defined as having severe CAP and requiring intensive care [7]. Several validated studies of the 2007 IDSA/ATS criteria showed favorable predictive efficacy [8][9][10]. However, two systematic reviews and meta-analyses demonstrated the lower sensitivity of minor criteria in predicting the requirement for intensive care [5,6]. Moreover, because of the difficulty of using IDSA/ATS minor criteria in general practice with nine different parameters, several studies tried to simplify the score [11,12].
The National Early Warning Score (NEWS) was first published by the Royal College of Physicians (RCP) in 2012 [13] and updated to NEWS2 in 2017 [14], following improvements in the detection of, and response to, clinical deterioration in patients with acute illness. The NEWS2 assessment method focuses on simple data collection which is suitable for practical use. NEWS2 ≥ 7 indicates high clinical risk that requires an immediate response, usually with admission to the ICU. The recent study demonstrated the additional prognostic accuracy of the NEWS on the pneumonia severity index (PSI) and CURB-65 in predicting ICU admission [15].
To our knowledge, the comparative data between the NEWS2 and IDSA/ATS minor criteria has not been evaluated. Therefore, this study aimed to assess the predictive performance of NEWS2 in identifying CAP patients with early progression to severe disease, and compared it to the 2007 IDSA/ATS minor criteria.

Study Design
This was a prospective cohort study carried out at Chiang Mai University Hospital during October 2020 to December 2021. The study population was adult inpatients (age ≥ 18 years) diagnosed with CAP. The diagnosis of CAP required new or increasing lung-infiltration on a chest radiography, and at least one of the following symptoms or signs: fever, productive cough, dyspnea, chest pain, or new abnormal findings of the chest auscultation (crepitation or rhonchi). The exclusion criteria included patients with severe CAP who met at least one of two IDSA/ATS major criteria at admission, pregnant patients, and COVID-19 patients.

Data Collection
Data was collected using case record forms, with informed consent being taken at the first evaluation. The patients were followed from hospital admission until discharge. Evaluation of vital signs was carried out at the first visit to the emergency department (ED) by a nurse or suitably trained assistant nurse. Data pertinent to demographics, comorbidities, vital signs, symptoms and signs, laboratory, and imaging were obtained from the medical records. The decision to transfer to a general ward or ICU was left to the discretion of the treating physicians, who were not part of this study. Due to limited ICU resources, almost all patients who did not meet the major criteria were transferred to a general ward.
In all patients, NEWS2 was calculated as part of the routine at the Emergency Department. The NEWS2 includes six physiological assessments (respiratory rate, SpO 2 , systolic blood pressure, pulse rate, consciousness, and temperature) plus one weighting score (need for oxygen supplement), with each item having a detailed score. The total score indicates the severity of the acute illness [14].

Endpoint of Interest
In general, most patients with CAP should achieve clinical improvement within 48-72 h after appropriate treatment [2]. The endpoint of interest was early progression to severe pneumonia, defined as the progression of the disease to respiratory failure requiring mechanical ventilation, circulatory failure requiring vasopressors, or death within 72 h after admission. Hospital mortality and hospital length-of-stay (LOS) were also evaluated.

Statistical Analysis
All statistical analysis was performed using STATA/MP version 14.2 (Stata Corp., College Station, TX, USA). Categorical data of baseline characteristics between early progression and non-progression groups were expressed as numbers and percentages, while continuous data were expressed using the median and interquartile range or mean and standard deviation. A comparison of categorical variables between groups was analyzed using the chi-square test or Fisher's exact test. Continuous variables were analyzed, using Student's t-test or the Wilcoxon rank-sum test, as appropriate for data distribution. A p-value < 0.05 was considered statistically significant. The predictive performance of NEWS2 and IDSA/ATS minor criteria for prediction of early progression to severe pneumonia was assessed, using the area under the receiver-operating-characteristic (AUROC) curves, with a 95% confidence interval (CI).

Patient Population
A total of 53 (25.6%) of the 260 enrolled patients had early progression to severe pneumonia, in which 42 (79.2%) patients progressed to respiratory failure and required mechanical ventilation, 17 (32.1%) patients had hemodynamic instability requiring a vasopressor, and 6 (11.3%) patients died within 72 h. Overall demographic data such as gender, age, comorbidities, symptoms, signs, and the IDSA/ATS minor criteria presented on admission were not significantly different in both groups. According to the initial laboratory results, the early-progression group had significantly lower serum-albumin levels compared to the non-progression group. In addition, the number of patients with a high BUN (≥30 mg/dL) was greater in the early-progression group. The in-hospital mortality rate of patients with early progression to severe disease was higher than those without progression. The hospital length-of-stay also tended to be longer in the early-progression group (Table 1). Thirty-six patients died in hospital, with a median LOS of 11 days. No significant difference in initial median NEWS2 was observed in the non-survivors and the survivors.

Performance of NEWS2 in the Prediction of Early Progression to Severe Disease
The median NEWS2 of the early-progression group were higher than the non-progression group, with 8 (6-9) vs. 7 (5)(6)(7)(8), p = 0.015, respectively. The hospital mortality of the patients with early-progression to severe pneumonia was significantly higher than that of the non-progression group (30.2% vs. 9.7%, p < 0.001, respectively). The median time of hospital-LOS was also longer in the early-progression group, at 15 vs. 10 days, p = 0.053 respectively ( Table 1).

Discussion
This was the first prospective cohort study that compared the accuracy of NEWS2 to 2007 IDSA/ATS minor criteria and CURB65 in the prediction of early progression to severe CAP. When patients progress to severe CAP, they have significantly greater mortality and longer hospital length-of-stays than patients with non-severe CAP. Our findings show that IDSA/ATS 2007 minor criteria, as well as initial NEWS2 for early progression to severe CAP, had a low predictive performance. The cut-off of NEWS2 at ≥ 8 did not improve the result. The assessment of the commonly used scores including CURB-65 ≥ 3 also showed low predictive-performance, with AUROC < 0.7. However, significant improve-

Discussion
This was the first prospective cohort study that compared the accuracy of NEWS2 to 2007 IDSA/ATS minor criteria and CURB65 in the prediction of early progression to severe CAP. When patients progress to severe CAP, they have significantly greater mortality and longer hospital length-of-stays than patients with non-severe CAP. Our findings show that IDSA/ATS 2007 minor criteria, as well as initial NEWS2 for early progression to severe CAP, had a low predictive performance. The cut-off of NEWS2 at ≥ 8 did not improve the result. The assessment of the commonly used scores including CURB-65 ≥ 3 also showed low predictive-performance, with AUROC < 0.7. However, significant improvement in the discriminative value was found after adding clinical parameters including low serumalbumin level and high BUN level. The best cut-off value of the BUN level to predict poor outcomes in this study was ≥ 30 mg/dL, while the BUN level used in the CURB-65 and the 2007 IDSA/ATS was ≥ 20 mg/dL. Our study was also found albumin level < 3 g/dL and BUN ≥ 30 mg/dL added to NEWS2 ≥ 8 increased the predictive value for mortality, shown in Figure 2.
Several studies demonstrated a favorable predictive-performance of IDSA/ATS minor criteria for ICU admission, the AUROC 0.85-0.88 [8][9][10]. Two systematic reviews and meta-analyses in 2011 and 2012 demonstrated moderate sensitivity (55.7-57.0%) and high specificity (90.5-91.7%) of IDSA/ATS minor criteria for prediction of ICU admission, while CURB65 provided lower sensitivity (50.0-56.2%) and specificity (72.1-74.2%) [5,6]. The AUROC of CURB65 was 0.69 [5]. Sbiti-Rohr et al. reported a low discrimination for NEWS in predicting thirty-day and six-year mortality in patients with CAP presenting at the Emergency Department, with AUROC < 0.70. However, it showed a moderate prediction for ICU admission within 30 days (AUROC 0.73) and additional predictive-benefit when added to the PSI and CURB65 [15]. According to the data of the current study, the predictive accuracy of the IDSA/ATS and the CURB65 for ICU admission and mortality were lower than previously reported. The differences in mortality rate and primary outcomes could be the explanation.
At the present, NEWS2 has been widely used in clinical practice, aiming at early detection and making appropriate responses for clinical deterioration in patients with acute illness. The results from this study confirm that the accuracy of initial NEWS2 to predict early progression to severe disease was comparable to the known tools, including IDSA/ATS and CURB-65. The additional advantage of NEWS2 over the other tools is that NEWS2 does not require laboratory results, especially the arterial-blood-gas test to measure the PaO 2 /FiO 2 ratio, which is an invasive procedure and may not be available in all hospitals. NEWS2 can be quickly evaluated at the bedside by many levels of health-care workers, using the routine vital-signs measurement and the level-of-consciousness test. Thus, it can be inserted into the nurse-driven protocol to alert physicians and triage the level of care in patients with CAP, who present with a mild disease initially. The association between low albumin-levels (≤3.1 g/dL) and poor outcomes, including longer time of mechanical ventilation and vasopressors in sepsis patients, has been reported earlier [16].
Holder AL et al. demonstrated that serum albumin < 3.5 g/dL was independently associated with disease progression to severe sepsis or shock within 96 h [17]. Several studies demonstrated that high BUN-levels are associated with increased mortality in sepsis patients; however, the cut-off varies in each study [18][19][20]. Moreover, a high serum-BUN-to-albumin ratio was found to be a good predictor of mortality in patients with sepsis and septic shock [21][22][23]. This study found the addition of the high-BUN and low-albumin values to NEWS2 helped prediction of early progression to severe disease in patients with non-severe CAP. However, further external-validation studies are required to confirm our findings.
Our study did have some limitations. First, it was a single-center study that enrolled only CAP patients who presented at the ED and were admitted to hospital. We did not follow the patients who were discharged from the ED. There was a possibility that patients were subsequently admitted to another hospital, due to clinical deterioration. In addition, we did not collect subsequent NEWS2 after patient admission, which could be more useful in real-time clinical assessment than a single-score measurement. Further study aimed at evaluating the benefit of NEWS2 for the prediction of severe disease and/or mortality with CAP, in the setting of general and community hospitals is recommended. Lastly, the result of our study may not be applicable to COVID-19 pneumonia. Further study to evaluate the role of NEWS2 in the COVID-19 setting would be interesting.

Conclusions
The NEWS2 could help us to plan appropriate bed-management and effective use of limited resources. The study results support the routine use of screening for NEWS2 in non-COVID-19, CAP patients. The predictive accuracy for early progression to severe pneumonia of NEWS2 was comparable to 2007 IDSA/ATS minor criteria and CURB65. The cutoff of initial NEWS2 ≥ 8 is suggested. The combination of NEWS2 ≥ 8, albumin < 3 g/dL, and BUN ≥ 30 mg/dL was a good predictor of early progression to severe disease and mortality in patients with CAP.
Author Contributions: Conception and methodology, P.T., P.S., A.L. and J.I.; data collection and resources, P.S. and J.I.; data curation and analysis: P.S. and P.P.; writing-original draft preparation: P.T. and P.S.; writing-review and editing: P.T., A.L., P.P. and J.I.; supervision, A.L. All authors have read and agreed to the published version of the manuscript. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The datasets that support the findings of this study are available on request from the corresponding author.