Risk factors of COVID-19 clinical worsening: A retrospective cohort study in COVID-19 referral hospital in west Java, Indonesia

Introduction

Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2) is a new type of coronavirus which causes respiratory infections and is currently becoming the center of attention in global health.¹ The clinical course of Coronavirus disease 2019 (COVID-19) ranges from asymptomatic to critically ill conditions with severe pneumonia and acute respiratory distress syndrome (ARDS), leading to death.² Rapid worsening is one sign of intense viral replication and uncontrolled inflammatory response.³ Systemic inflammatory response, immune system disorder, and dysfunction of the renin-angiotensin-aldosterone system (RAAS) can trigger a cytokine storm causing ARDS and leading to increased mortality.^4–7

Managing COVID-19 in limited resources requires a quick decision as to whether initiating antiviral or supportive treatment may lead to different outcomes and prognoses.⁸ Certain parameters, along with the patient's clinical appearance, may predict the worsening of COVID-19.⁹ Therefore, patients who tend to worsen need to be prioritized and require early aggressive treatment.¹⁰ Prompt and appropriate medical treatment can improve the prognosis of COVID-19 patients.¹¹ Several factors such as age, gender, comorbidities, and laboratory parameters may predict the occurrence of worsening.^6,12 On the other hand, some factors initially linked to the poor outcome of the disease but proven contrarily, such as asthma.¹³

A comprehensive understanding of the characteristics of COVID-19 patients and the risk factors for progression is fundamental. The findings of this study may be helpful not only in determining the prognosis but also in establishing appropriate treatment policies to prevent poor outcomes. Therefore, our study was conducted to assess whether male, elderly, several comorbidities and laboratory parameters could affect the clinical worsening in COVID-19 patients at the tertiary referral hospital in one of the most densely populated provinces in Indonesia, Hasan Sadikin General Hospital in Bandung, West Java, which can be used to determine the priority and predict the clinical outcome of the patients.

Methods

Results

A number of 537 subjects were included in our study, consisting of 336 patients with mild COVID-19 and 201 patients with moderate COVID-19. Analysis was carried out during the 14-day follow-up of treatment, with 72 patients had worsened, and those who did not experience worsening were 465 subjects. Further analysis revealed that 20 subjects experienced worsening from mild COVID-19 to severe or critical, and 52 subjects experienced worsening from moderate to severe COVID-19 progressed to severe and critical. Data is shown in Figure 1.

Figure 1. Research flow of subjects.

COVID-19: Coronavirus disease; REDCap: Research Electronic Data Capture.

ROC analysis

Numerical data that was statistically significant for the worsening of COVID-19, the cut-off value was determined based on the calculation of the ROC curve, as shown in Table 4 below.

Table 4. Cut-off values of variables affecting the worsening of COVID-19.

Variable	Cut off point	AUC (95% CI)	p-value	Sensitivity (%)	Specificity (%)	OR (95% CI)
Respiratory rate	>23	0.658 (0.617-0.699)	<0.001	64.71	68.87	4.06 (2.38-6.92)
Pulse rate	>99	0.574 (0.531-0.617)	0.065	41.18	77.19	2.37 (1.39-4.02)
Temperature	>37	0.562 (0.519-0.605)	0.103	36.76	77.35	1.99 (1.16-3.40)
ALC	≤1479.6	0.646 (0.603-0.687)	<0.001	74.24	51.79	3.10 (1.73-5.54)
NLR	>3.842	0.716 (0.675-0.755)	<0.001	77.27	59.87	5.03 (2.75-9.23)
CRP	>3.76	0.798 (0.756-0.836)	<0.001	85.71	70.34	11.85 (5.43-25.85)
Leukocyte	>8800	0.642 (0.599-0.683)	<0.001	58.21	68.14	2.98 (1.76-5.03)

Figure 2 shows the factors analyzed in the ROC curve. In the analysis through the ROC curve obtained from seven variables, only pulse rate and temperature whose cut-off values were not statistically significant (p < 0.05). Multiple logistic regression was carried out to analyze what factors were related to the worsening of COVID-19 patients using multivariate analysis. The variables included in this study had a p-value of < 0.25 from the bivariate analysis results or were considered important by the researcher. Variables in multivariate analysis were age > 60 years, male, hypertension, diabetes mellitus, chronic heart disease, chronic kidney disease, body mass index, dyspnea, fever, cough, respiratory rate, pulse, body temperature, leukocyte value, ALC value, NLR, CRP value, pneumonia in the chest X-ray.

Figure 2. Factor affecting COVID-19 progressivity into severe and critical case in ROC curve analysis.

AUC: Area under the ROC curve; ALC: Acute lymphocyte count; NLR: Neutrophil-to-lymphocyte ratio; CRP: C-Reactive protein.

Discussion

Our study classified the patients with and without worsening states, then retrospectively analyzed the factors affecting the outcome. Previous studies showed several factors are associated with the deterioration of the disease and the poor outcome, including older age, male, comorbidities such as obesity, hypertension, heart disease, chronic obstructive pulmonary disease, and chronic kidney disease, and also laboratory parameters such as C-Reactive Protein and NLR.^{3,8,10,20–24} In our study, age over 60 years old, previous history of heart disease, diabetes mellitus, respiratory rate > 23x/minute, and NLR value > 3.8 were found to be statistically significant as the factors affecting the worsening of COVID-19 in multivariate analysis. Several conditions may affect the results.

Subjects aged over 60 affected the risk of worsening COVID-19 by 4.20 times higher (95% CI 2.128 – 8.320) to severe and critical degrees compared to younger ages on treatment for 14 days at the hospital. A previous meta-analysis combining various registries around the world, including China, Italy, Spain, United Kingdom, and the United States involving 611.583 patients, explained that the risk of COVID-19 worsening increases exponentially with age, especially in the elderly.²⁵ The elderly have a higher risk of worsening because of the susceptibility to the infection and more severe clinical manifestations due to the physiological process of aging and various comorbidities that can reduce the functional capacity of the body's defense mechanism to combat the invasion of various microorganisms that enter and cause the infection.^25,26

Patients with comorbidities were more susceptible to COVID-19 infection and poor outcomes. Our study found that hypertension was the most common comorbid (38.9%), followed by diabetes mellitus (31.9%), chronic heart disease (18.1%), and chronic kidney disease (15.3%) in deteriorated patients. These comorbidities were significant in bivariate analysis, but only chronic heart disease and diabetes mellitus were statistically significant in multivariate analysis. A meta-analysis by Chang et al. found that hypertension, diabetes mellitus, and cardiovascular disease were the most common comorbidities associated with the severity of disease and mortality from COVID-19.²⁷ The immune system in diabetic patients became weak and compromised, thus worsening the COVID-19 condition.²⁸ The effect of COVID-19 on the cardiovascular system is also remarkable. Rapid release of high level of cytokines due to the COVID-19 infection causes the dysfunction of vascular endothelial cells and abnormal coagulation state, increasing the likelihood of thromboembolic events to occur.²⁹ Therefore, patients with coronary artery disease (CAD) and COVID-19 present a greater risk of mortality. In groups of survivors against non-survivors, the incidence of CAD was 8.5% versus 21.6%, respectively.³⁰ Conversely, the later analysis shows that hypertension and chronic kidney disease were not statistically significant. As the risk linked with hypertension is accentuated by its confounding influence on diabetes mellitus, Sun et al. discovered that hypertension was not an independent factor (95% CI 0.33 – 1.61).³¹ Another study also stated that there was no correlation between renal disease and aggravation of COVID-19 (95% CI 0.76 – 7.50), as renal involvement is not caused solely by one factor.^32,33

Patients with respiratory rate > 23x/minute at first admission had a higher risk of pulmonary deterioration. An increase in the respiratory rate indicates an early sign of disturbed airway physiology before it progressively becomes the symptom of severe COVID-19.³⁴ Early detection of respiratory rate > 23x/minute on physical examination is a beneficial sign, as stated in the minor criteria for severe pneumonia according to ATS/IDSA, where the higher respiratory rate is an alarm sign to notify the physician to evaluate and treat effectively and immediately.³⁵

Leukocyte, lymphocyte, neutrophil, platelet, and neutrophil-lymphocyte ratio may be used to predict the worsening.^1,36 Patients with severe COVID-19 symptoms should have laboratory parameters checked for hyperinflammatory markers to improve mortality rates.³⁷ One of the causes of inflammation is infection. A severe inflammatory response contributes to a weak adaptive immune response. This causes an imbalance in the immune response.^36,38 Circulating biomarkers can represent inflammatory and immune status, thus can be used as potential predictors in the prognosis of COVID-19 patients.²⁴ Yang et al. found that patients with worsening of the disease usually have progressively lower lymphocyte, higher neutrophil, and increased NLR.³⁶ Higher NLR was found to be independently correlated with severe COVID-19.³⁹ The amount of the increase in neutrophil counts seen during the immunopathological phase, even in the absence of bacterial co-infection, nonetheless reflects the intensity of the inflammatory response.⁴⁰ Moreover, several chronic conditions may affect NLR, such as hypertension, diabetes mellitus, and cardiovascular disease.⁴¹ Previous studies also reported that severe lymphopenia and higher WBC were linked to the poor prognosis of COVID-19.^42,43 These findings align with our study result, as the severe and critical patients had higher leukocyte, more severe lymphopenia, neutrophilia, and increased NLR.

Older age is associated with elevated proinflammatory cytokines, thus contributed to prolong viral RNA shredding in COVID-19 patients.^44,45 High neutrophil was observed during a cytokine storm caused by COVID-19.⁴⁶ A previous study found that neutrophils will increase and enter the lungs during a cytokine storm that triggers ARDS, causing organ damage and death in COVID-19. This was found after post-mortem autopsies were carried out on COVID-19 patients.⁴⁷

According to the result of our study, the five significant factors could be used to forecast COVID-19 deterioration and determine which patients need to start receiving treatment immediately. Recent research has shown that the use of monoclonal antibodies (mAbs) like ritonavir-boosted nirmatrelvir (paxlovid), remdesivir, bebtelovimab, and molnupiravir reduce the risk of disease deterioration.⁴⁸ To avoid an increase in mortality rate, our study findings can be utilized to identify which patients need to start receiving these mAbs earlier.

However, some limitations should be noted. As the data were collected retrospectively, there were missing or incomplete data, including anosmia, ageusia, malignancy, cerebrovascular disease, and chronic liver disease. Therefore, role of these variables as risk factors for clinical worsening could not be investigated.

Conclusion

Five independent variables are interrelated and can affect the deterioration of COVID-19, including older age, chronic heart disease, diabetes mellitus, higher respiratory rate, and high NLR values. These risk factors help prioritize the patients and predict treatment outcomes since these factors are routinely measured in managing COVID-19 patients.