Take a look at the Recent articles

Keywords

 COVID-19, SARS-CoV-2, CURB-65, clinical decision-making tools, risk stratification

Introduction

The SARS-CoV 2 pandemic was a major societal event that acted as a stress test for much of the world’s healthcare systems. The pandemic arose in 2019 as a small cluster of index cases presenting with a pneumonia-like illness in Wuhan city, China, originating from the Huannan fish market. Within weeks, the virus spread at an alarming rate both locally and globally, owing to its high transmissibility [1]. On the 23rd of January 2020, the World Health Organization (WHO) issued a formal statement evaluating the evolving outbreak as being of high global risk [2]. Soon after, the outbreak spread rapidly to over 20 countries, prompting the WHO to label the outbreak as a global pandemic on March 11, 2020 [3]. The virus’ spread was difficult to contain, and initiated a sprint towards developing appropriate management and diagnostic pathways for infected patients. Nevertheless, many of the world’s healthcare systems were overwhelmed by the sheer magnitude of cases, including well-established ones such as China, the United States, and Italy [4]. As of today in 2025, the total infection toll as reported by the WHO sits at 778 million cases worldwide, with at least 7 million deaths [5]. Although the SARS-CoV-2 pandemic has officially passed, it remains to be a pivotal chapter in history from which much can be learned. This will ring particularly true as we examine the evolution of both the clinical and public health response to the pandemic, from initial scramble to a more stable steady-state, with a critical lens on whether the early use of pneumonia severity tools could have mitigated early strain on healthcare infrastructure

During the early phases of the outbreak, the primary focus of many countries revolved around appropriate detection and containment of the virus. By the time the WHO had released its initial safety report in January 2020, a number of countries had already taken action against minimizing spread including China, Thailand, Japan, South Korea, and the USA. These interventions included screening individuals at risk, contact tracing of flight passengers, and quarantining suspect patients, with some, namely Japan and the USA, having developed their own PCR assay for detection of the virus [2].

The use of pre-validated clinical decision making tools, such as CURB-65 or the PSI was only seriously considered in the later stretches of the pandemic, often producing good sensitivity and specificity. With the benefit of hindsight, it is important to analyze this evidence to guide future preparedness for respiratory illness outbreaks, avoiding early pitfalls which may have strained healthcare systems worldwide.

To the best of our knowledge, we have not found any publications that examined the literature produced during the timeline of the outbreak to evaluate the use of pre-validated clinical decision making tools (CDMs) such as CURB-65 as an early proxy for triage, risk stratification, and prognostication. This study aims to address this gap through a retrospective cohort analysis, integrating these findings in a narrative review to evaluate whether the use of pre-validated clinical decision-making tools would be an appropriate initial approach to guide disposition planning in very early stages of future outbreaks.

Methodology

Study design and participants

This was a single-center, retrospective cohort study, conducted at King Faisal Specialist Hospital and Research Centre, Riyadh (KFSH&RC). The study population included adult patients aged 18 and above who were diagnosed with COVID-19 pneumonia at our center between January 2020 and August 2021. A patient was defined as having COVID-19 pneumonia if: (1) reverse-transcriptase polymerase chain reaction (PCR) results for COVID-19 were positive; and (2) there was evidence of pulmonary infiltrates on chest radiography or chest CT scan. The CURB-65 scoring system was applied retroactively to all study participants in order to calculate sensitivity and specificity of the scoring system at predicting mortality.

Data collection

Retrospectively from patient electronic medical records (EMRs). All data files were stored on a secure platform and only shared with the research team. The following variables were extracted:

• Demographic data: Age, sex, comorbidities
• Clinical data: triage vitals, COVID severity class at triage, Symptoms at presentation (including new-onset confusion), laboratory results (including BUN), imaging findings
• Outcomes: Patient disposition and hospital course, 30-day mortality

During this time, patients at our center were being triaged into classes based on symptoms and severity of illness:

• Category A: asymptomatic
• Category B: mild symptoms such as fever, mild cough, myalgia, head ache, sore throat, gastrointestinal (GI) symptoms
• Category C: Pneumonia-Sick e.g. shortness of breath, severe head ache, chest pain
• Category D: Severe disease

 

A ‘CURB65 -based predicted disposition’ was calculated using the following CURB65 scores:

• Scores 0-1: Outpatient
• Scores 2: Inpatient, admitted to floors
• Scores 3-5: ICU admission

Statistical analysis

Descriptive statistics were used to summarize the baseline characteristics of the cohort. Continuous variables were expressed as mean ± standard deviation or median (interquartile range), and categorical variables were expressed as frequencies and percentages. Comparisons between 30-day mortality and characteristics were performed using Chi square test for qualitative variables and independent sample t-tests for continuous variables. A contingency table analysis was used to calculate the proportion of agreement between patient hospital course and predicted disposition based on CURB 65 scores. A p-value of <0.05 was considered statistically significant. The analysis was computed using SPSS.

Ethical considerations

This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board (IRB) of KFSH&RC.

Results

Data were collected from 531 patients with a mean age of 57.2 years (SD 15.9). As shown in (Table 1), 275 patients (51.8%) were male. Fever was the most commonly reported symptom (64.8%), followed by dyspnea (53.9%), dry cough (46.7%), and myalgia (19.8%). A small proportion of patients (4.9%) reported no symptoms.

Table 1. Baseline characteristics of participants

Variable	N (%)
Gender
Male	275 (51.8%)
Female	256 (48.2%)
Symptoms
Fever	344 (64.8%)
Confusion	14 (2.6%)
Dry cough	248 (46.7%)
Headache	82 (15.4%)
Myalgia	105 (19.8%)
Sore throat	75 (14.1%)
Diarrhea	93 (17.5%)
Nasal congestion	34 (6.4%)
Sneezing	9 (1.7%)
Dyspnea	286 (53.9%)
Productive cough	88 (16.6%)
Loss of taste	33 (6.2%)
Loss of smell	20 (3.8%)
No symptoms	26 (4.9%)
Comorbidities
Smoker	35 (6.6%)
Diabetes	241 (45.4%)
Hypertension	247 (46.5%)
Stroke	24 (4.5%)
Chronic kidney disease	60 (11.3%)
Cardiomyopathy	21 (4.0%)
Previous myocardial infarction	48 (9.0%)
Arrhythmia	37 (7.0%)
Liver disease	25 (4.7%)
COPD	14 (2.6%)
Asthma	50 (9.4%)
Immunocompromised	155 (29.2%)
No comorbidities	77 (14.5%)
COVID severity class on triage
Class A	31 (5.8%)
Class B	117 (22.0%)
Class C (Moderate)	288 (54.2%)
Class D (Severe)	95 (17.9%)

Regarding comorbidities, hypertension (46.5%) and diabetes (45.4%) were the most prevalent. Other notable conditions included chronic kidney disease (11.3%), asthma (9.4%), and previous myocardial infarction (9.0%). A total of 29.2% of patients were immunocompromised, either due to malignancy or immunosuppressive therapy. In contrast, 14.5% of patients reported no comorbidities.

On arrival to the emergency department, patients were triaged into four severity classes according to institutional policy. The majority were categorized as Class C (moderate symptoms, 54.2%), followed by Class B (mild symptoms, 22.0%), Class D (severe symptoms, 17.9%), and Class A (asymptomatic, 5.8%).

Actual disposition vs calculated disposition using CURB 65

The actual disposition of patients showed that 58.4% were admitted to the hospital floor, 40.7% required ICU-level care, and only 0.9% were discharged for outpatient follow-up. Among those admitted to the ICU, the mean length of stay was 7.2 days. The overall 30-day mortality rate for the cohort was 7.3% (Table 2).

Table 2. Patient outcomes

Characteristic	N=531 (%)
Hospital course
Admitted to floor	310 (58.4%)
ICU	216 (40.7%)
Outpatient	5 (0.9%)
Outcome 30 days later/ mortality
Alive	492 (92.7%)
Dead	39 (7.3%)

CURB-65 scores were calculated for each patient, and predicted dispositions were assigned based on standard scoring guidelines: scores of 0–1 indicated outpatient management, 2 suggested inpatient admission, and scores ≥3 warranted ICU-level care. Based on these predictions (Table 3), 63.1% of patients were expected to be managed as outpatients, 19.4% as floor admissions, and 17.5% as ICU-level admissions.

Table 3. CURB-65 scores and expected disposition

Characteristic	N=531 (%)
CURB 65 Scores
0	176 (33.1%)
1	159 (29.9%)
2	103 (19.4%)
3	70 (13.2%)
4	22 (4.1%)
5	1 (0.2%)
CURB 65 predicted disposition
Outpatient	335 (63.1%)
Admit for inpatient observation	103 (19.4%)
ICU	93 (17.5%)

Sensitivity of CURB 65

A cross-tabulation comparing the CURB-65 code–predicted level of care with actual patient disposition is shown in (Tables 4). Overall agreement between the CURB-65 code and observed outcomes was low, with only 22.98% (122/531) of cases showing exact matches. Agreement was highest for ICU-level care, with 72 patients correctly classified as requiring ICU admission.

Given the small number of patients discharged for outpatient care (n = 5), a secondary analysis focused on ICU vs. non-ICU admissions. In this binary classification, the CURB-65 code demonstrated a sensitivity of 55.81% (95% CI: 48.84–62.78), specificity of 69.70% (95% CI: 63.25–76.15), positive predictive value of 78.26% (95% CI: 72.47–84.05), and negative predictive value of 44.66% (95% CI: 37.68–51.64). Agreement in this subset improved to 60.51%.

Table 4. Relationship between actual disposition and expected outcomes

Actual disposition
CURB 65 expected outcomes	Outpatient	Admitted to floor	ICU	Total
Outpatient	4	244	87	335
Inpatient admission	0	46	57	103
ICU	1	20	72	93
Total	5 (0.94%)	310 (58.38%)	216 (40.68%)	531 (100%)

Different variables and 30 day mortality

Significant associations were found between mortality and COVID-19 Severity Class on triage (p < 0.001), CURB-65–predicted disposition (p < 0.001), and symptoms of myalgias (p = 0.017) and sore throat (p = 0.031) (Table 5).

Table 5. Significance of qualitative patient characteristics related to 30-Day mortality outcomes

	Outcome 30 days later/ mortality
Characteristic	Alive =N (%)	Dead=N (%)	Total=N	p-value
Gender
Male	249 (90.6%)	26 (9.5%)	275	0.053
Female	243 (94.92%)	13 (5.08%)	256
COVID Severity Class on Triage
Class A	28 (90.32%)	3 (9.68%)	31	<0.001
Class B	117 (100%)	0 (0.0%)	117
Class C (Moderate)	270 (93.75%)	18 (6.25%)	288
Moderate) D (Severe)	77 (81.05%)	18 (18.95%)	95
Symptoms Fever
No	171 (91.44%)	16 (8.56%)	187	0.43
Yes	321 (93.31%)	23 (6.69%)	344
Confusion
No	478 (92.46%)	39 (7.54%)	517	0.286
Yes	14 (100%)	0 (0.0%)	14
Dry cough
No	258 (91.17%)	25 (8.83%)	283	0.16
Yes	234 (94.35%)	14 (5.65%)	248
Headache
No	412 (91.76%)	37 (8.24%)	449	0.064
Yes	80 (97.56%)	2 (2.44%)	82
Myalgia
No	389 (91.31%)	37 (8.69%)	426	0.017
Yes	103 (98.10%)	2 (1.9%)	105
Sore throat
No	418 (91.67%)	38 (8.33%)	456	0.031
Yes	74 (98.67%)	1 (1.33%)	75
Diarrhea
No	403 (92.01%)	35 (7.99%)	438	0.215
Yes	89 (95.70%)	4 (4.30%)	93
Nasal congestion
No	460 (92.56%)	37 (7.44%)	497	0.735
Yes	32 (94.12%)	2 (5.88%)	34
Sneezing
No	484 (92.72%)	38 (7.28%)	522	0.662
Yes	8 (88.89%)	1 (11.11%)	9
Dyspnea
No	225 (91.84%)	20 (8.16%)	245	0.503
Yes	267 (93.36%)	19 (6.64%)	286
Productive cough
No	412 (93%)	31 (7%)	443	0.492
Yes	80 (90.91%)	8 (9.09%)	88
Loss of taste
No	459 (92.17%)	39 (7.83%)	498	0.095
Yes	33 (100%)	0 (0.0%)	33
Loss of smell
No	472 (92.37%)	39 (7.63%)	511	0.199
Yes	20 (100%)	0 (0.0%)	20
No symptoms
No	470 (93.07%)	35 (6.93%)	505	0.107
Yes	22 (84.62%)	4 (15.38%)	26
CURB 65 Code
Outpatient	331 (67.3%)	4 (10.3%)	335	<0.001
Admitted for Inpatient observation	98 (19.9%)	5 (12.8%)	103
ICU	63 (12.8%)	30 (76.9%)	93

Patients who died within 30 days had significantly higher age (66 vs. 56 years, p = 0.0002), BMI (29.6 vs. 23.4 kg/m², p < 0.001), and PaO₂ (8.22 vs. 6.83 kPa, p = 0.026), along with higher WBC count (8.6 vs. 6.6 ×10⁹/L, p = 0.012), neutrophil percentage (71.2% vs. 59.8%, p = 0.006), PTT (46.1 vs. 39.3 seconds, p < 0.001), D-dimer (2.64 vs. 1.29 mg/L, p < 0.001), BUN (10.4 vs. 5.9 mmol/L, p < 0.001), creatinine (139.7 vs. 89.3 µmol/L, p < 0.001), AST (97.6 vs. 60.1 U/L, p = 0.001), ferritin (1291.8 vs. 927.6 µg/L, p = 0.033), and procalcitonin (4.14 vs. 0.86 ng/mL, p = 0.016). Hemoglobin and lymphocyte levels were lower among deceased patients (111 vs. 125 g/L, p < 0.001; 12% vs. 21.4%, p < 0.001, respectively) (Table 6).

Table 6. Significance of quantitative characteristics related to 30-Day mortality outcomes

Outcome 30 days later/ mortality
Characteristic	Alive    Alive (N=492, Mean (SD)	Dead      Alive (N=39, Mean (SD)	Total Alive (N=531, Mean (SD)	p-value
Age	56.490 (15.667)	66.205 (15.845)	57.203 (15.869)	0.0002
BMI	25.402 (6.984)	29.692 (8.649)	25.718 (7.197)	<0.001
1PaCO2	5.250 (1.501)	5.544 (1.399)	5.275 (1.494)	0.246
2PaO2	6.828 (3.661)	8.226 (3.957)	6.947 (3.704)	0.026
HCO3	25.137 (4.630)	26.189 (5.996)	25.226 (4.762)	0.193
WBC	6.587 (4.763)	8.575 (4.611)	6.733 (4.776)	0.012
Hemoglobin	125.208 (23.588)	111.103 (26.860)	124.172 (24.099)	<0.001
Platelets	225.959 (93.492)	212.833 (138.497)	224.993 (97.396)	0.418
Neutrophils	59.886 (24.068)	71.210 (25.943)	60.698 (24.357)	0.006
Lymphocytes	21.409 (13.987)	12.003 (13.505)	20.740 (14.149)	<0.001
Eosinophils	1.245 (2.090)	0.785 (0.797)	1.212 (2.026)	0.224
Basophils	0.366 (0.931)	0.316 (0.275)	0.363 (0.903)	0.769
PT	15.651 (4.789)	17.441 (3.963)	15.786 (4.752)	0.024
PTT	39.338 (9.682)	46.138 (18.094)	39.851 (10.677)	<0.001
Fibrinogen	7.529 (49.677)	5.744 (5.238)	7.401 (47.885)	0.834
D-Dimer (fixed) 3	1.286 (2.192)	2.638 (3.773)	1.383 (2.361)	<0.001
ESR	40.177 (30.460)	47.731 (38.174)	40.670 (31.021)	0.23
Blood Urea Nitrogen	6.922 (6.295)	16.256 (10.008)	7.609 (7.061)	<0.001
Serum Creatinine	100.199 (113.019)	182.695 (154.585)	106.258 (118.367)	<0.001
Potassium	4.621 (6.844)	6.577 (12.426)	4.765 (7.398)	0.112
Sodium	137.393 (7.990)	139.821 (7.973)	137.571 (8.006)	0.068
HBA1c	15.364 (55.475)	6.475 (1.420)	15.012 (54.383)	0.75
LDH	340.807 (452.746)	491.649 (325.557)	351.969 (446.082)	0.048
ALT	38.484 (41.357)	50.245 (83.106)	39.335 (45.643)	0.126
AST	37.767 (29.394)	78.087 (123.474)	40.685 (44.578)	<0.001
Ferritin	656.671 (868.332)	1,882.370 (2,984.871)	744.391 (1,192.284)	<0.001
Procalcitonin	0.861 (6.985)	4.141 (16.318)	1.106 (8.081)	0.016

Discussion

The evolution of the COVID-19 response

As the initial days of the outbreak unfolded, a far graver problem began to arise which would mark the true hallmark of what most narrators would consider the brunt of the SARS-CoV 2 outbreak. Initial triage tools were developed under pressure to screen and stratify COVID-related admission. At that point in time, data regarding the clinical history and laboratory findings of SARS-CoV 2 was still evolving. Triage typically segregated patients into high risk and low risk patients on the basis of fever exposure history, clinical symptoms of respiratory illness, positive PCR testing, laboratory derangements consistent with viral infections, and positive imaging results [6,7]. These criteria were broad and sufficient for detecting patients at risk of infection, and facilitated isolation measures to help quell spread. However, they lacked the specificity of established pneumonia scoring systems at predicting mortality or severity, which did little to aid downstream decisions regarding resource allocation; namely informing the need for ICU admission. Many centers were not adequately prepared for the large influx of COVID positive patients that would greet their door, with many hospitals quickly reaching functional capacity [8-10].

As the early days of March 2020 came to an end, the question of “Is this a patient with COVID-19?” quickly evolved into “Does this COVID-19 patient require intensive care?”. A concern magnified by the sheer number of patients requiring respiratory support. Grasseli et al. reported that out of a sample of 1300 patients with confirmed SARS-CoV 2 infection, 99% required respiratory support [9]. ICU admission rates ranged from 5% to 45% [11-14], varying widely between countries, often necessitating pressors and/or advanced respiratory support as a result of hypoxemic respiratory failure [12,15] The need for efficient triage pathways that guided intensive care resource allocation was superseded by the unique challenge of developing such pathways presented by the urgent nature of a never before seen, rapidly transmissible virus [16].

COVID risk modifiers

Baseline characteristics and comorbidities

Data quickly emerged showing that older patients presented with disproportionately more severe disease as compared to younger patients [17,18], a finding which was found to be among the most predictive of mortality in some studies [18]. This was likely attributed to a more aggressive cytokine response in these populations, concomitant comorbidities, and susceptibility of infection [13,16,17]. A correlation between mortality and age was noted in our sample. Furthermore, comorbidities such as diabetes mellitus, hypertension, and atherosclerotic cardiovascular diseases have been consistently shown to be associated with more severe disease, likely due to increased host susceptibility of infection caused by upregulated ACE II receptors, an important cell surface molecule implicated in the pathogenesis of the disease [16]. Concurrent respiratory disease, specifically COPD, is an important predictor of severity among patients with COVID-19, in light of the physiologic dysregulation triggered by chronic hypoxia among other factors [18]. Obesity is a well-described risk factor which was implicated in more severe disease as well as with mortality [19]. In our sample, higher BMI was associated with increased mortality.

Laboratory derangements

Lymphocytopenia was among the earliest findings captured among patients with COVID-19, and was shown to predict disease severity [16]. SARS-CoV 2 is unique in the sense that it produces a uniquely potent cytokine signal in hosts, contributing to its virulence [16]. The resulting systemic inflammatory response is reflected in some of the laboratory derangements seen in more severe disease. Markers such as CRP, ESR, and D-Dimer All of these findings were also shown in our sample to be correlated with increased mortality, corroborating the present literature.

CURB 65 and other risk stratification tools during the pandemic

In light of the aforementioned circumstances, a number of authors sought to develop recommendations that informed clinical decision making, the earliest of which (to the best of the authors’ knowledge) included the UK NICE NG159 rapid guideline (a), as well as Zhou et al (b), advising the use of the Clinical Frailty Scale (CFS) and SOFA respectively in stratifying patients, both published in March of 2020. The following days and months, namely June 2020 onward, was when we began to see the emergence of more publications evaluating the utility of both pre-validated clinical decision making tools, as well as those developed specifically to predict COVID-19-related mortality.

In our analysis, we retroactively applied the CURB-65 criteria to a sample of 531 patients with confirmed COVID-19 pneumonia, which demonstrated a significant correlation with 30-day mortality but showed only moderate agreement with actual patient disposition. CURB-65 scores correlated significantly with mortality among our sample, which mimics similar studies in the literature. Interestingly, CURB-65 predicted that only 17.5% of patients among our sample would require ICU-level care as opposed to the 40.7% who received ICU-level care when stratified by COVID severity and clinical gestalt. This is underscored by the fact that agreement between actual patient dispositions as per COVID severity class only agreed with the CURB-65 disposition about 60% of the time. When we analyze the subgroup of ICU admitted patients, we begin to see that CURB-65 was more predictive of mortality than clinical gestalt + COVID severity. The authors are cautious about conflating mortality risk with the need for ICU admission, which is a common talking point in triage literature. We do note however that CURB-65 could have been a more useful adjunct to clinical gestalt and typical ICU admission criteria at triaging ICU-level care as compared to the de-novo triage tool used at our center. The primary question the authors would like to pose, as well as the underpinning of our discussion here is: Given that SARS-CoV 2 caused pneumonia, why weren’t these pneumonia-specific tools repurposed sooner in the outbreak? In the light of the evolving pandemic, large volumes of research on the topic were published by the day. Triage in such circumstances would be inherently messy as more evidence accumulates elucidating risk factors and laboratory derangements predictive of mortality and increased resource utilization. However, the use of pre-validated clinical decision making tools (CDMs) such as CURB-65 at the onset of the pandemic would have offered a useful foundation to managing resource distribution, as well as guide early research. These tools, although not a “perfect” fit (given that CURB 65 was originally validated for bacterial pneumonia!), may later be augmented and tested with additional modifiers as new evidence emerges regarding risk predictors. In hindsight, it is apparent that even tools created specifically for COVID-19 were poorly calibrated, and were of limited utility as compared to CURB-65 and qSOFA, as illustrated by Gupta et al’s analysis of 22 risk prediction models. And although the authors of this analysis continue to the culmination of this philosophy manifested in the 4C score, a modified version of CURB-65 developed and validated by the point of this discussion is not to argue whether CURB-65 would be a perfect tool for triaging ICU-resources in future outbreaks. But rather, to argue for the early use of pre-validated CDMs over clinical gestalt in the face of an unknown illness.

Limitations

Our study comes with a number of limitations we would like to address. Firstly, the design of our study is not sufficiently powered to truly detect AUC performance of CURB-65 in stratifying patients. Our analysis was exploratory to assess whether CURB-65 showed agreement with current practice in our center, as well as whether it correlated with mortality among a unique subset of patients with high disease complexity, including a large proportion of cancer, organ transplant, and immunocompromised patients. As such, greater emphasis should be placed on the narrative review portion of this article.

 In addition, it is important to note that CURB-65 was originally validated for bacterial pneumonia. Its mapping onto a disease causing viral pneumonia is only justifiable due to the overlapping clinical presentations of both pathologies. The later developed 4C score represents how more disease-specific criteria improve the performance of the CDM. Finally, the purpose of this article is not to compare CURB 65 to other CDM tools per se, but rather to highlight that, in the early stages of an outbreak, applying the most specific pre-validated CDM could guide resource allocation, and can then be refined as evidence evolves.

Conclusion

We present a combined retrospective cohort study and narrative review examining how different Clinical Decision Making (CDM) tools were applied during the pandemic to COVID-19 mortality and severity. We note the possible role CURB-65 could have had during the early days of the outbreak in predicting mortality and the need for ICU admission in conjunction with relevant clinical findings. We propose the early deployment of pre-validated CDM tools such as CURB-65 to guide resource allocation in the very early stages of future outbreaks, with subsequent refining as new data emerges. This approach may help improve system preparedness during the early stages of future pandemics.

References

1. Mohan BS, Nambiar V (2020) COVID-19: An insight into SARS-CoV-2 pandemic originated at Wuhan City in Hubei Province of China. J Infect Dis Epidemiol 6: 146.
2. Novel Coronavirus (2019-nCoV) SITUATION REPORT – 3 23 JANUARY 2020. https://iris.who.int/bitstream/handle/10665/330762/nCoVsitrep23Jan2020-eng.pdf
3. Khan M, Adil SF, Alkhathlan HZ, Tahir MN, Saif S, et al. (2020) COVID-19: A global challenge with old history, epidemiology and progress so far. Molecules 26: 39. [Crossref]
4. Jahangir MA, Abdul Muheem, Midhat Fatima Rizvi (2020) Coronavirus (COVID-19): History, current knowledge and pipeline medications.
5. WHO COVID-19 dashboard. https://data.who.int/dashboards/covid19
6. Wang Q, Wang X, Lin H (2020) The role of triage in the prevention and control of COVID-19. Infect Control Hosp Epidemiol 41: 772-776. [Crossref]
7. Ayebare RR, Flick R, Okware S, Bodo B, Lamorde M (2020) Adoption of COVID-19 triage strategies for low-income settings. Lancet Respir Med 8: e22. [Crossref]
8. Comelli I, Scioscioli F, Cervellin G (2020) Impact of the COVID-19 epidemic on census, organization and activity of a large urban Emergency Department. Acta Biomed 91: 45. [Crossref]
9. Grasselli G, Zangrillo A, Zanella A, Antonelli M, Cabrini L, et al. (2020) Baseline characteristics and outcomes of 1591 patients infected with SARS-CoV-2 admitted to ICUs of the Lombardy Region, Italy. JAMA 323: 1574-1581. [Crossref]
10. Emanuel EJ, Persad G, Upshur R, Thome B, Parker M, et al. (2020) Fair allocation of scarce medical resources in the time of Covid-19. N Engl J Med 382: 2049-2055. [Crossref]
11. Immovilli P, Morelli N, Antonucci E, Radaelli G, Barbera M, et al. (2020) COVID-19 mortality and ICU admission: The Italian experience. Crit Care 24: 228. [Crossref]
12. Guan WJ, Ni ZY, Hu Y, Liang WH, Ou CQ, et al. (2020) Clinical characteristics of coronavirus disease 2019 in China. N Engl J Med 382: 1708-1720. [Crossref]
13. Huang C, Wang Y, Li X, Ren L, Zhao J, et al. (2020) Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet 395: 497-506. [Crossref]
14. Richardson S, Hirsch JS, Narasimhan M, Crawford JM, McGinn T, et al. (2020) Presenting characteristics, comorbidities, and outcomes among 5700 patients hospitalized with COVID-19 in the New York City area. JAMA 323: 2052-2059. [Crossref]
15. Mara G, Nini G, Cotoraci C (2025) Chronic obstructive pulmonary disease and COVID-19: The impact of hematological biomarkers on disease severity and outcomes. J Clin Med 14: 2765. [Crossref]
16. Tyrrell CS, Mytton OT, Gentry SV, Thomas-Meyer M, Allen JL, Narula AA, et al. (2021) Managing intensive care admissions when there are not enough beds during the COVID-19 pandemic: A systematic review. Thorax 76: 302-312. [Crossref]
17. Bonanad C, García-Blas S, Tarazona-Santabalbina F, Sanchis J, Bertomeu-González V, et al. (2020) The effect of age on mortality in patients with COVID-19: A meta-analysis with 611,583 subjects. J Am Med Dir Assoc 21: 915-918. [Crossref]
18. AlGhatrif M, Cingolani O, Lakatta EG (2020) The dilemma of coronavirus disease 2019, aging, and cardiovascular disease: insights from cardiovascular aging science. JAMA Cardiol 5: 747-748. [Crossref]
19. Singh R, Rathore SS, Khan H, Karale S, Chawla Y, et al. (2022) Association of obesity with COVID-19 severity and mortality: An updated systemic review, meta-analysis, and meta-regression. Front Endocrinol 13: 780872. [Crossref]