Introduction
The factor that started in Wuhan city of China and caused the pandemic was defined as severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) and the pneumonia caused by it was named Coronavirus disease-2019 (COVID-19) (1). There are no specific antivirals for patients and it seems impossible to predict how the pandemic will progress in the future (2,3). Community-acquired pneumonia (CAP) is often caused by Streptococcus pneumoniae, Staphylococcus aureus, Mycoplasma pneumoniae, Chlamydophila pneumoniae, rhinovirus, influenza, respiratory syncytial virus, continues to be a persistent global public health problem (4,5). It can be more difficult to determine the empirical treatment to be applied in periods of pandemic, especially when the incidence of CAP is increasing (5). The mortality of COVID-19 pneumonia is up to 14.4 times higher than CAP, and timely delivery of medical treatments for SARS-CoV-2 is required to reduce mortality (6,7). Distinguishing covid-pneumonia patients from CAP subjects will limit unnecessary antibiotic use. A biomarker that differentiates COVID-19 pneumonia from CAP could be important for an early and appropriate empirical treatment algorithm. In our study we evaluated features of two different pneumonia groups in order to discover a readly available and favourable marker that differs etiologies.
Materials and Methods
Patients
Two groups of patients were included in study. First group (CoV) comprising patients >18 years old, diagnosed with COVID-19 pneumonia according to a positive nasal swap sample for SARS-CoV-2 virus in polymerase chain reaction and a thorax computed tomography imaging suitable for COVID-19 pneumonia, and hospitalized from date 15 April 2020 to 30 April 2020. Hospitalization was decided according to guideline provided by Turkish Republic Health Ministry which suggest hospitalization of patients >50 years, having additional comorbidity, with a pulse saturation of ≤94, and one laboratory criteria of poor prognosis. Second group (CAP) including patients >18 years old, diagnosed with CAP according to clinical examination and imaging data, and hospitalized from date 1 December 2019 to 30 January 2020, the time gap in which Influenza used to make a peak in our country and any COVID-19 pneumonia has not been detected yet. Hospitalization was assessed according to CURB-65 criteria which contains a score of five items, one point for each of: Confusion, urea >7 mmol/L, respiratory rate >/=30/min, low systolic (<90 mm Hg) or diastolic (/=65 years (8). Those who scored two points or more were admitted to the hospital. Outpatients were excluded.
All data regarding socio-demographics, admissional laboratory [hemoglobin, platelet, white blood cell (WBC), lymphocyte, neutrophil, monocyte, D-dimer, ferritin, lactatedehydegenesis, C-reactive protein (CRP), alanine amino transferase, aspartate amino transferase, gamma glutamyl transferase, alkaline phosphatase, sodium, potassium, creatine, blood urea nitrogen, troponine], comorbid diseases, admissional physiological values (fever, oxygen saturation, blood pressure, pulse), supplemental oxygen (O2) requirement, transfering to intensive care unit on follow-up, lenght of stay, and outcome was gathered retrospectively from hospital’s electronical records. Our study has been approved by local ethical commission with a registiration number of 2020/150. Patient consent was not obtained for our retrospective study.
Statistical Analysis
All values were represented as mean ± standard deviation, 95% confidence intervals (95% CI), percentages, medians with interquartile ranges as appropriate. Suitable analyzes including Mann-Whitney U/Student’s t, paired t and chi-square/exact tests were appropriately used. A sample analysis to detect a group difference regarding monocyte count was calculated with a power of 0.8 and an alpha of 0.05 that showed 36 patients a group was required. A ROC analysis using area under curve (AUC), has been performed by parameters considered to provide clinical meaning, while evaluating a favourable biomarker able to distinguish patients with COVID-19 pneumonia from others, then sensitivity and specificity analysis were appropriately calculated. A p-value of <0.05 was accepted as significant. All analyzes were calculated with SPSS 23 IBM® statistics program provided by Hacettepe University Faculty of Medicine, Ankara, Turkey.
Results
A total of 90 patients were evaluated. Mean age was 52±15 years [Female (F): 21 (57%)] in CoV group while it was 73±15 in [F: 24 (45%)] CAP group (p<0.001 for age, p=0.248 for gender). Comorbidities other than congestive heart failure, chronic lung and neurologic diseases did not differ between groups (Table 1). Cough as an onset symptom was much more common in group CoV than in group CAP (Table 1). When compared to group CoV, laboratory parameters related with WBCs and CRP level were higher in group CAP. Only median pulse saturation and mean heart rate (beat/min), unlike other admissonal physiologic parameters, differed between groups [96 (94-98) vs. 91 (86-96), p<0.001); 92±13 vs 103±21, p=0.006]. No mortality difference was detected between groups.
ROC analysis, with combined parameters in order to achieve a more extended AUC, revealed that an admissional neutrophil x CRP ratio to have an AUC of 0.91 (p<0.001) while a formula gathered age x CRP x monocyte (count) to have a that of 0.92 (p<0.001) (Figure 1) with a sensitivity and specificity of 0.83 (0.70-0.91) and 0.84 (0.67-0.93), respectively. The latter formula had a positive predictive value of 0.88 (0.75-0.95) and a negative predictive value of 0.75 (0.61-0.89) (Table 2). Only age x CRP x monocyte count with a cut-off value of 167.662 and having chronic obstructive lung disease were found to be an independent factor for predicting non-COVID-19 pneumonia while adjusted for age (>65 years), gender, and comorbidities [OR (95% CI): 18.5 (4.6-74.4), p<0.001; OR: 24.1 (2.4-246), p=0.007, respectively] (Data not shown).
Discussion
We have been getting involved in a period in which CAP is increasing while there is not a certain knowledge about how COVID-19 will evolve. In this study, we found that age x CRP x monocyte formula could be a reliable marker to distinguish COVID-19 pneumonia from CAP.
People in CoV group were detected to be younger than CAP group and found as statistically significant in our study which was similar to literature (9). When COVID-19 pandemic had been realized for the first time in our country, a curfew order for people older than 65 years was administered by the goverment. The restriction has to be taken into account, also. Especially old age condition and comorbid diseases, which were also known to be risk factors for CAP, were detected higher in CAP group (10-12). In addition to the fact that old age poses a risk for CAP, the younger age group is more likely to be in crowded environments due to working life and social life and is excluded from restrictions. For these reasons, the age difference made us think of it as a predictor, not a constraint. It is also stated in the literature that old age can be considered as diagnostic for COVID (13). Being elderly is a parameter in the PES score which is used to predict the etiology of pneumonia in the literature, and there are age-related studies in the prediction of etiology (14,15). One of the dynamics in the transmission of SARS-CoV-2 is age, and serological studies in England, Brazil, Japan and Germany estimate that the seroprevalence is highest in adults under 35 years of age (16,17).
Among the presentation symptoms, in our study, cough, fatigue, fever and myalgia were significantly common in COVID pneumonia, while dyspnea was more common in CAP patients. There are significant differences in the comparison of presentation symptoms of COVID and non-COVID pneumonias in the literature, and symptoms other than dry cough do not appear as appropriate data for the prediction of the causative agent of respiratory system infection (18,19). Among the vital signs of the patients at admission, tachycardia and relative hypoxia are more common in the CAP group and may be associated with an excess of comorbid diseases in the same group.
Hitherto, laboratory parameters especially lymphocyte and monocyte count have been investigated already in detail and those revealed findings similar to our study had revealed (19-22). In these studies, the leukocyte and neutrophil counts were relatively lower in COVID pneumonia than in non-COVID pneumonia including other viral pneumonia (18-21). Our study showed monocyte count to be an essential parameter to distinguish patients with COVID-19 and others. Migration and redistrubition of monocytes from blood to both lung and non-respiratory tissue, thus consumption might be the explanation of significantly reduced monocyte count in COVID-19 patients, while compared to patients those are not (18,22-25). It has been shown in most studies that the number of monocytes is decreased in COVID patients, and in addition, the correlation between severe alterations of monocyte subtypes and severe clinical findings shows the importance of monocyte in the pathogenesis of COVID and it should be clarified with ongoing studies (26,27).
CRP is an acute phase reactant that is released dependant on inflammation of the body and is also used in differential diagnosis between viral infections and bacterial infections (25,28). In our study, the CRP value in the CAP group was found to be statistically significantly higher than the CoV group. In the literature, there are studies in which the CRP value in non-COVID diseases including other viral diseases was found to be higher than COVID pneumonia (29).
As a consequence, we are hopeful of the vaccination programs implemented for the pandemic that caused serious losses all over the world for 1 year. The fact that RNA viruses can mutate frequently is the biggest obstacle to the success of vaccinations and the possibility of COVID-19 getting out of our lives. For this reason, studies for the development of early diagnosis methods and the discovery of direct-acting antivirals should continue in the fight against SARS-CoV-2.
Study Limitations
Our study has some limitations. First one is the retrospective nature of the study that might have caused data loss. Second is the possible selection bias due to the prohibition for elderly to be outside by goverment provision in Turkey that might be a protective factor for them which results in lower mean age in CoV group. Third, limited number of patients, eventhough our study is not underpowered, restrained subgroup analysis.
Conclusion
We aimed to establish a suitable marker to distinguish patients with COVID-19 pneumonia from those with CAP and found age CRP monocyte formula to be reliable. Validation of our result in larger studies will provide an easily accessed and readily available parameter for health workers.
Ethics
Ethics Committee Approval: Our study has been approved by Düzce University Local Ethical Commission with a registiration number of 2020/150.
Informed Consent: Patient consent was not obtained for our retrospective study.
Peer-review: Externally peer-reviewed.
Authorship Contributions
Concept: A.D.H., C.Ç., Design: A.D.H., S.D., Data Collection or Processing: Ö.A., F.M.S., Analysis or Interpretation: F.M.S., M.Y.P., Literature Search: S.D., Ö.A., Writing: A.D.H., M.Y.P.
Conflict of Interest: No conflict of interest was declared by the authors.
Financial Disclosure: The authors declared that this study received no financial support.
