Epidemiology and Risk Factors of COVID-19-Related Mortality

Introduction During the coronavirus disease 2019 (COVID-19) pandemic in India, several characteristics of hospitalized COVID-19 patients, based on demographics, mortality predictors, and presence of comorbidities, were found to be associated with poor outcomes. The objective of this study was to identify such epidemiological and clinical characteristics among the patients admitted at a tertiary-care center in India that may have predisposed them to COVID-19-related mortality. Methods This retrospective observational study conducted at the Department of Medicine, All India Institute of Medical Sciences, New Delhi, in May 2021 included 141 COVID-19 confirmed patients. The medical history, demographic characteristics, comorbidities, clinical findings, and laboratory data of each patient were obtained. The data were analyzed to identify significant clinical and laboratory parameters that led to the adverse final outcomes. Results Hypertension was the most common comorbidity and the presence of diabetes with hypertension led to poorer final outcomes. Lower oxygen saturation and requirement of oxygen supplementation at admission along with worse prognostic scores during admission led to poorer outcomes. Twenty-seven patients needed non-invasive ventilation (NIV) during the hospital course, and all ultimately landed up among the 56 patients who were managed on invasive mechanical ventilation (IMV). Multivariate logistic regression analysis performed identified COVID-19 severity at admission, co-existence of hypertension and diabetes mellitus, systolic blood pressure less than 90 mm Hg, and serum creatinine greater than 1.2 mg/dL to be associated with higher COVID-19 mortality. Conclusion COVID-19 patients having the co-existence of diabetes and hypertension constitute a high-risk group and may be targeted by prompt vaccination strategies. The presence of severe disease along with a need for oxygen therapy and other intensive care interventions ultimately led to unfavorable outcomes.


Introduction
India witnessed the second wave of coronavirus disease 2019  in March 2021, with 30 million confirmed cases and almost 0.4 million mortalities [1]. Several epidemiological studies from China, Europe, North America, Brazil, and India identified various characteristics of hospitalized COVID-19 patients based on demographics, mortality predictors, and association with comorbidities, which were associated with poor outcomes [2][3][4]. The majority of the studies have focussed on respiratory outcomes, mortality, risk factors, steroid use while a few have reported the multisystem outcomes and predictors of mortality mostly for critically ill COVID-19 patients. Mortality was extremely high in patients who were suffering from a severe form of the disease and with multiple comorbid [5][6]. Similarly, the use of various mortality predictors to prognosticate and approximate the further clinical course of illness in a COVID-19 patient is not wellestablished. In this small retrospective study, we compared various baseline demographic and clinical characteristics along with laboratory data and clinical outcomes in both survivor and non-survivor COVID-19-confirmed patients admitted to our center.

Study design
This was a retrospective observational study of COVID- 19

Clinical and biological data
In this study, the medical history, demographic characteristics, comorbidities, clinical findings, and laboratory data of each patient were recorded. The COVID-19 severity class along with Acute Physiology and Chronic Health Evaluation (APACHE) II, Sequential Organ Failure Assessment (SOFA), and Modified Early Warning Score (MEWS) scores at admission were considered. The baseline parameters and further hospital course were reviewed. Patient survival was the primary outcome. The final outcome of these patients was categorized as either survivor or non-survivor.

Statistical analysis
For categorical data, results were summarized as the actual number of patients (n) and percentages (cumulative incidence). All continuous variables were expressed as mean with standard deviation or median with inter-quartile range. Individual parameters among survivors and non-survivors were compared using the chi-square test or Fisher's exact test for categorical variables, independent t-test for continuous variables following normal distribution, and Wilcoxon rank-sum test for parameters following non-normal distribution. Univariable and then multivariable backward stepwise logistic regression analyses were performed to identify independent predictors of hospital mortality. A p-value of <0.05 was considered significant. All statistical analysis was conducted using StataCorp. 2019 (Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC).

Demographic and clinical features
A total of 141 patients were included in the study. Their baseline demographic characteristics and clinical features are shown in Table 1

Laboratory investigations
As listed in Table 2, non-survivors had a significantly higher mean total leukocyte count and median serum alkaline phosphatase, serum urea, serum creatinine, and serum lactate dehydrogenase. The significantly elevated markers of inflammation among non-survivors were C-reactive protein and serum ferritin.

Multivariate logistic regression analysis of mortality in COVID-19 patients
A multivariate logistic regression analysis was performed by considering the variables of significance as shown in Table 3. The variables identified to be independently associated with a statistically significant higher risk of mortality in COVID-19 disease were COVID-19 severity at admission, co-existence of hypertension and diabetes mellitus, systolic blood pressure less than 90 mmHg, and serum creatinine greater than 1.2 mg/dL.

Discussion
We conducted this small-scale, retrospective study and analyzed risk factors among these patients. Nonsurvivors had a higher mean age compared to survivors, similar to other studies from Wuhan and Pakistan [7][8]. More than half of the patients had some underlying chronic illness, commonly hypertension followed by diabetes and hypothyroidism, similar to studies on Chinese and Italian patients [7,9]. The meta-analysis by Sahni S et al., including more than 20000 patients from around the world, reported the same [10]. However, a South Indian study showed contrasting results, with diabetes more prevalent than hypertension [11]. One-fourth of our patients had both hypertension and diabetes significantly more than a Wuhan study where 6.7% of patients had both comorbidities [12].
The mean SpO2 recorded during admission was significantly lower among non-survivors, with 94% of them having SpO2 <90%, thus signifying that a significant majority of the non-survivors had severe COVID-19 disease at admission. Coincidentally, almost 90% of the non-survivors needed oxygen supplementation at the time of admission. This is in conjunction with similar studies where a majority of non-survivors had a lower SpO2 compared to the survivor group and had an early need of supplemental oxygen [11][12][13][14]. A study with ICU patients in Florida with a similar sample size showed 50% of non-survivors had SpO2 <90% during admission while 60% were in need of supplemental oxygen. In comparison, our non-survivors could be considered to have more severe COVID-19 disease. Similarly, the COVID-19 severity class along with the prognostic scores of APACHE II, SOFA, and MEWS score were all worse in the non-survivor group like various previous studies [13,[15][16].
Considering the various laboratory parameters, the non-survivor group had elevated mean total leukocyte count, serum urea, serum creatinine, serum alkaline phosphatase, and serum lactate dehydrogenase compared to the survivor group. The various other parameters were, however, almost similar in both groups. This is in conjunction with most other studies, thus signifying the role of these laboratory parameters in determining the severity of COVID-19 disease and the high rates of mortality [8,13,[17][18]. C-reactive protein and serum ferritin were elevated among non-survivors thus stressing the role of the inflammatory cytokine storm in the disease severity and poor prognosis [13].
Around 20% of our patients had severe respiratory distress and needed NIV management. Nevertheless, all these patients clinically deteriorated subsequently and were further managed on IMV. Thus, around 40% of our patients were managed on IMV and only two of them survived while the rest all succumbed to their illness. This was different compared to the study by Menzella F et al. where around 48% of patients were managed on NIV and the mortality rate recorded was 25% [19]. Around 10% of our patients were managed with prone positioning. This number was fewer than similar studies but our resource-poor setting had a role to play [20].
We performed a multivariate logistic regression analysis considering the variables that showed significance. A higher COVID severity at admission, co-existing hypertension and diabetes, lower systolic blood pressure at admission, and elevated serum creatinine were all identified to be significant independent risk factors for mortality in our regression model. Our findings could be considered in conjunction with various previous similar studies [12][13]. However, further studies with a greater sample size are suggested to better correlate these findings.
Our study has highlighted several important aspects but the interpretation should be done taking into account several limitations. The study was retrospective with a small sample size. It didn't consider the effect of various therapies namely antibiotics, anti-inflammatory drugs, and other supportive therapies on the results. Additionally, while examining multiple factors associated with survival using multivariate regression analysis, there may have been unidentified potential confounders.

Conclusions
Our study has highlighted several important aspects during the second COVID-19 wave in the Indian subcontinent. About 63% of our patients needed oxygen therapy at presentation, and 40% needed mechanical ventilation at some point of time during hospitalization. Both diabetes and hypertension were associated with a substantial increase in the risk of COVID-19 mortality. Our findings may be attributed to a sicker patient profile and further studies with a larger sample population are warranted to standardize them. The study may provide further guidance for future studies on outcome predictors in COVID-19 patients. The insight gained may guide future vaccination strategies in high-risk groups and hospital development policies with a vision to better fight further waves of this pandemic.

Additional Information
Disclosures