1 Introduction

There was a reported outbreak in Wuhan, China, in December 2019, which promptly became a pandemic with unclear circumstances. At the beginning of the year 2020, scientists successfully isolated a novel virus that belongs to the Beta-corona virus genus of the Coronaviridae family. It was declared a pandemic by the World Health Organization (WHO) in February 2020 (Abbas et al., 2021; Li et al., 2020; Lotfy et al., 2021; Phelan et al., 2020).

In case of COVID-19 pneumonia,a high fever, dry cough, and difficult breathing are the predominant symptoms. The great majority of patients had a mild to moderate illness and were able to recover entirely with conservative therapy. However, 15–30 % of patients may develop severe pneumonia, leading to ARDS, multiple organ failure, or even death (AlOtaibi et al., 2021; Huang et al., 2020; Zayed et al., 2022).

Severely ill patients are challenging to treat due to lack of targeted therapies; so that it is obligatory for a healthcare worker to look for the clinical characteristics of severity and subsequent predictors of mortality to implement the appropriate and early intervention in the hopes of reducing death rates. Recently, it has been shown that age, the presence of cardiovascular co-morbid profile, and diabetes mellitus are factors that may be used to predict mortality. In addition, serum ferritin, D-dimer, and cardiac enzymes have all been found by other researchers as potential biomarkers for predicting severe and fatal illnesses (Weng et al., 2020).

Recent work has resulted in the developing of an integrated ANDC score, which serves for the early classification of COVID-19 patients and treatment guidance (Management protocol of COVID 19 patients by Ministry of Health and population, Egypt Version 1.5., 2021). Consequently, The aim of the current work is to investigate whether the ANDC sore and CRP-derived inflammatory markers might be used to predict COVID-19-infected adult patients with high probability of mortality.

2 Patient and methods

This retrospective study was conducted at Zagazig University Hospitals Isolation unit and Clinical Pathology Department, Egypt from March 2021 to August 2021. That inquiry is congruent with guidelines established by the World Medical Association in its Helsinki Declaration. This research included 98 adult patients who were confirmed by laboratory and radiologically as COVID-19. Patients were above the age of 18. They were diagnosed according to the Egyptian Ministry of Health's Scientific Committee (Goudouris, 2021). Throat swabs were taken from individuals suspected of having SARS-CoV-2 infection to confirm the. In addition, each patient underwent a chest computed tomography (C.T.) scan and a battery of laboratory tests, including measurements of leukocytes, neutrophils, lymphocytes, C-reactive protein, fibrin degradations (D-dimer), creatinine level, albumin, lactic dehydrogenase (LDH) and ferritin. In addition, the CRP to lymphocyte ratio, the CRP to platelet ratio and the CRP to albumin ratio (CLR,CPR,CAR, respectively). The ANDC score was calculated using the following formula: $$Totalpo\mathrm{i}\mathrm{n}\mathrm{t}s=1.14\ast age-20ys+1.63\ast NLR+5\ast D-\mathrm{d}\mathrm{i}\mathrm{m}er+0.14\times CRP(mg/L)$$

3 Methods

4 Results

A total of 98 confirmed COVID-19 patients were enrolled in the current study. Unfortunately, 51 patients had passed away by the time it was through, and 47 were still alive. Therefore, there was no statistically significant difference between both groups regarding sex or length of hospital stay. At the same time, there was a statistically significant difference between the two groups regarding age, TLC, ANC, NLR, D-Dimer, and albumin (p = 0.013, 0.028, 0.006, <0.001, and 0.029), Table 1.

Moreover, a highly statistically significant difference regarding CRP levels, CAR, CPR, CLR, and ANDC was noted (0.032, <0.001, <0.001, <0.001, and < 0.001, respectively), Table 2, Fig. 1A, Fig. 1B.

Close

Fig. 1A

Box-plot diagram represents the range of ANDCscore, CRP, CLR, and CAR in the studied groups; the upper & lower line in each box represents the 75th& 25th percentile, respectively, while the line through each box indicates the median. Whiskers represent the range between the minimum and maximum values.

Export to PPT

Close

Fig. 1B

Box-plot diagram represents the range of CPR in the studied groups.

Export to PPT

Our study showed that serum CRP level > 123 ng/ml, CAR > 36.77, CPR level > 462, and CLR > 84 had sensitivity; (64.71 %, 66.6 %, 72.5 %, and 76.4 %, respectively) and specificity; (85.1 %, 78.7 %, 72.3 %, and 72.3 % respectively) in mortality prediction. Meanwhile, the ANDC score was the most sensitive indicator (88.2 %) for mortality outcome, Fig. 2& Table 3.

Close

Fig. 2

ROC curve of serum CRP, ANDC, CAR, CLR, and CPR levels markers for mortality in COVID-19 patients.

Export to PPT

There was a significant difference in LOS between high and low levels of CAR, CPR, and CLR groups (p = 0.001, <0.001, 0.001; respectively), as well as a high level of ANDC score compared with a low-level group (p = 0.001). However, no significant difference in LOS was observed between high and low CRP (p = 0.224), Table 4, Fig. 3.

Close

Fig. 3

Kaplan– Meier survival curves illustrating hospital survival time differences in all patients and within each category as regards CRP, ANDC, CAR, CLR, and CPR levels.

Export to PPT

The effects of Age, Sex, CRP, CRP-derived inflammatory markers, ANDC level, Initial TLC, Ferritin, LDH, and D-Dimer on the likelihood of participants' mortality after ICU admission were investigated and ascertained by performing logistic regression. The univariate logistic regression analyses revealed that mortality was dependently associated with aging, CAR; CPR; CLR Levels, ANDC level, Ferritin, and LDH with H.R. [1.03 (1.01–1.06); 2.60 (1.44–4.71); 2.93 (1.58–5.46); 2.71 (1.42–5.19); 3.93 (1.67–9.26); 1.002 (1.001–1.003) and1.001 (0.999–1.002) respectively] and P-value was [0.008; 0.002;0.001,0.003;0.002; 0.001 and 0.004 respectively]. However, on multivariable Cox regression analysis, aging, CPR, and ANDC level were independently associated with mortality with H.R. [1.025 (1.002–1.050); 2.338 (1.189–4.599) and 2.896 (1.191–7.044)] and P-value was [0.034, 0.014 and 0.019 respectively], Table 5.

5 Discussion

Some tests can be performed in labs or imaging devices that may indicate the typical signs of COVID-19 and its consequences or risk factors for problems (Azkur et al., 2020 Jul). Complete blood count lymphopenia, eosinopenia, and neutrophil/lymphocyte ratio of less than 3.13 are connected to increased severity and a poorer prognosis (Chen et al., 2020; Vabret et al., 2020; Wynants et al., 2020). Higher CRP, ferritin, PCT, LDH and D-dimer are all associated with a more severe illness and a less favorable prognosis than lower levels of these markers in most studies.

Developing a reliable prediction tool to forecast how the illness would manifest itself clinically may greatly assist in risk stratification, clinical decision-making, and rational resource optimization. They are essential to prevent potentially life-threatening side effects and, eventually, lessen the severity of the disease's impact. Unfortunately, the scores and nomograms that have been made public up to this point are much more challenging to understand due to the inclusion of a significant increase in the number of criteria (some up to 23) (Bal et al., 2021).

In the present study, our objective was to evaluate the predictive usefulness of the aforementioned score and parameters in adult covid-19 patients necessitating hospital admission.

We used these four factors to develop a scoring system called the ANDC score for predicting mortality. On the other hand, it is essential to keep in mind that it forecasts death rates rather than the need for NIV, IVM, or ICU admission. As a result, it may be most effective at its extremes, such as when it gives doctors the confidence to release patients with low mortality ratings or prompts early talks about treatment escalation with patients who need oxygen.

CRP is a protein that may be used to locate or monitor ailments that produce inflammation. Viral infections are the most prevalent disorders that decrease the number of lymphocytes in the blood, and CRP can be used to detect or monitor these conditions. These findings support our earlier conclusion that CLR and NLR are both significant predictors of mortality. Although both NLR and LCR could identify seriously unwell patients and those critically ill, Bal and colleagues discovered that LCR was more effective than NLR (Tonduangu, et al., 2021). Compared to NLR, LCR showed a superior ability to discriminate between thoughtfully and critically sick individuals (Liu et al., 2020). The viral load of the SARS-CoV-2 virus is likely responsible for explaining our findings. This viral load has been linked to CRP and lymphopenia and has been demonstrated to correlate well with the severity of the disease (Kalabin et al., 2021).

The current analysis found that CAR was considerably more remarkable in the group of patients who passed away compared to those who survived, consistent with previous findings from past investigations (Wiedermann, 2021). Albumin is found in high concentrations in human blood; hypoalbuminemia, which is low albumin levels, is often caused by inflammation and is linked to worse outcomes across various illnesses (Poudel et al., 2021). This helps explain why the dying patients had a significantly elevated CAR level. Hypoalbuminemia in Covid-19 patients results from the complex interaction of systemic inflammation with successively increased capillary permeability and redistribution of albumin to interstitial fluids. This conclusion was supported by previously published data that revealed an association of severity of illness and greater D-dimer values; a prognostic mortality clue (Yao et al., 2020; Meini et al., 2021). According to the findings of this study, a higher level of D-dimer was significantly associated with a greater risk of passing away. When there is a systemic infection, both the extrinsic coagulation route and the contact coagulation pathway are active (Wool and Miller, 2021). The coagulation cascade activation, which may have been brought on by viremia, superinfection, cytokine storm, or organ failure, resulted in increased D-dimer levels in patients who later passed away. Disseminated intravascular coagulopathy may be a factor in COVID-19 (Fan et al., 2006), which might explain why D-dimer levels were more significant in individuals who passed away from the disease.

Our findings revealed that CRP, CAR, and CLR all had a high AUC for predicting mortality (0.772, 95 percent CI: 0.677–0.867 for CRP; 0.778, 95 percent CI: 0.683–0.856 for CAR; 0.772, 95 percent CI: 0.677–0.866 for CLR) and that using CAR and CLR boosted sensitivity at the expense of specificity. On the other hand, The NLR alone may predict mortality with a reasonably high AUC (0.764, 95 percent confidence interval (CI): 0.659–0.850), but it only has a sensitivity of 56.52 percent. This was determined via observational research and meta-analyses. The combination of CRP and the NLR combined led to an area under the curve (AUC) value of 0.804 (95 percent confidence interval [CI]: 0.702–0.883), as well as a considerable improvement in sensitivity from 56.52 percent to 73.92 percent, at the expense of a loss in specificity (Kalabin et al., 2021). Both the LCR and the NLR were able to identify critically ill patients from severe patients, with the CLR having a higher ROC AUC than the NLR (Tonduangu et al., 2021). This information lends credence to our hypotheses and reveals that published research supports them. On the other hand, Tonduangu and colleagues discovered that CLR was the only significant predictor of mortality out of the investigated variables (CRP level, lymphocyte level, and CLR level) (Liu et al., 2020).

With a cutoff score of > 72.6, we stratified patients according to the score into low score group and high score group. The ANDC score was 66.9 (30.7–153.3) in live Patients and 97.0 (37.8–160.9) in the deceased one, with a positive predictive value of the scoring system (70.3 %), and the negative predictive value was 82.4 % which showed good Discrimination using ROC curves (AUC:0.778;95 % CI; p < 0.001) as an AUC ROC value over 0.75 represents good clinical Discrimination (Liu et al., 2021).

One retrospective study used the ANDC score on 301 patients with COVID-19 to assess its prognostic usefulness in predicting hospital mortality. They found that the ANDC score provided a quantitative tool for identifying individuals with a high mortality risk on admission (AUC 0.912) and directing clinical care (Huang et al., 2020).

One significant disadvantage is that its use may need to be more practical in low- and middle-income nations (LMICs).

Unfortunately, in LMICs, where physiological scores may be more practical, restricted access to virological testing and laboratory facilities may limit their utility.

In conclusion: The utility of the ANDC score and the CRP-derived inflammatory indicators readily increases the prediction of identifying patients at high mortality risk.

Ethical consideration:

The current study was conducted in accordance with the strict guidelines and regulation such as Declaration of Helsinki.

The study was conducted according to the protocol approved by the Review Board of the Faculty of Medicine of Zagazig University IRB#9567–1-6–2022.

Informed consent was obtained from all subjects and their legal guardian(s).