|Year : 2020 | Volume
| Issue : 2 | Page : 79-84
Factors associated with case fatality in COVID-19
Aastha Poddar1, Abhishek Gogate1, Anusha S Kumbar1, Kurt Cordeiro Sydney Francis1, Jayaprakash Appajigol1, Shivalingappa B Javali2
1 Department of General Medicine, Jawaharlal Nehru Medical College, KLE Academy of Higher Education and Research, Belagavi, Karnataka, India
2 Department of Community Medicine, USM KLE International Medical Programme, Belagavi, Karnataka, India
|Date of Submission||26-May-2020|
|Date of Acceptance||12-Jun-2020|
|Date of Web Publication||11-Sep-2020|
Dr. Jayaprakash Appajigol
Department of General Medicine, KLE Academy of Higher Education and Research, Jawaharlal Nehru Medical College, Nehru Nagar, Belagavi, Karnataka
Source of Support: None, Conflict of Interest: None
Introduction: The ability of the novel coronavirus SARS-CoV-2 to selectively target high-income countries, has puzzled many epidemiologists. Several possible protective factors such as Bacillus Calmette–Guérin (BCG) vaccine coverage, higher environmental temperature, higher proportion of the younger population, prevalence of hypertension (HTN), diabetes, and smoking have been suggested. This study aims at understanding the influence of each of these factors on case fatality rate due to COVID-19. Methods: We selected two or more countries in each geographical region, which were severely affected by SARS-CoV-2 infections. Case fatality rate for each country was calculated obtaining data on the number of cases and deaths in each country. The details of coverage of BCG vaccination, prevalence of HTN and diabetes, age structure of the population, and average maximum and minimum temperatures during February and March months of 2020, and prevalence of smoking for each country were taken from the World Health Organization and other standard databases till May 9, 2020. Statistical analysis was done using Karl Pearson's correlation and multiple linear regression model. Results: Case fatality rate was negatively correlated with BCG coverage of the country, percentage of population below 14 years age and average maximum temperatures. Conclusions: Environmental temperature, wider BCG coverage, and higher proportion of the younger population in middle- and low-income countries might have played a protective role against COVID-19.
Keywords: Bacillus Calmette–Guerin, COVID-19, temperature, young age
|How to cite this article:|
Poddar A, Gogate A, Kumbar AS, Francis KC, Appajigol J, Javali SB. Factors associated with case fatality in COVID-19. J Sci Soc 2020;47:79-84
| Introduction|| |
The infection caused by novel coronavirus (SARS-CoV-2) has brought the world to a standstill. To curtail the spread of the virus, governments are trying valiantly to isolate the patients, asking citizens to stay at home, stopping all public gatherings, and imposing lockdowns. On the other hand, the scientists from various countries are trying their best to discover an effective vaccine and a medication to control this disease. In spite of all these collective efforts, the virus is spreading rapidly. As expected, during the initial phase of the spread of COVID-19, the number of cases was maximum in China, the country from where COVID-19 was first reported. With time, this coronavirus (SARS-CoV-2), one of the most contagious viruses, started spreading across the globe. The United States reported its first case of confirmed COVID-19 on January 20, 2020. From then onward, the virus is spreading relentlessly across the country. During the month of January, a major concern for the World Health Organization (WHO) was the spread of SARS-CoV-2 to low- and middle-income countries with fragile health systems. India detected its first confirmed case on January 30, 2020, when a student returning from Wuhan tested positive in the southern state of Kerala., By the end of January, around 21 countries reported nearly 10,000 confirmed cases. On the contrary, the number of new cases in China started decreasing during late February. At the same time, new cases started appearing in Europe, mainly affecting Italy and France. Within a short period of time, Italy witnessed more than 2 lakh cases and more than 28,000 deaths, and the count is rising ever since. By the end of March, the epicenter of the disease had shifted to the United States of America. The United States has become the world's worst-affected country with more than a million cases and more than 50,000 deaths, and the figures are exponentially increasing. Today, the United States and Europe are failing to contain the rising infection rates and mortalities in spite of having one of the best health-care systems in the world. On the other hand, the developing countries with fragile health-care systems are not witnessing the similar exponential spread of the virus, and they also have reported lower death rates. This disproportionately smaller number of cases and deaths reported from low-income countries remains unclear. Few attempts have been made to solve this puzzle. Bacillus Calmette–Guerin (BCG) vaccine which is known for protection against tuberculosis also has protective effects against other viral diseases. BCG has been used for the treatment of urinary bladder malignancy and has been found to have immunomodulatory effect. Wider BCG vaccine coverage has been studied, and it has been found that countries with universal BCG immunization program have fewer cases of COVID-19. Researchers also pointed out that higher environmental temperature and higher ultraviolet (UV) index are protective against COVID-19. Developed countries such as Italy and the United States have a higher proportion of the older population. An Italian study has demonstrated that a higher proportion of the older population may be the reason for more cases and higher case fatality rates.
We hypothesize that the combined effect of BCG vaccination policy, weather conditions, and age structure of the populations in different countries might have impacted the transmission patterns and SARS-Cov-2, associated infection, and mortality. The aim of our study is to understand the effects of each of these factors on the infection rates and fatalities in different countries.
| Methods|| |
In this random cross-sectional study, we selected two or more countries in each geographical region which were severely affected by SARS-CoV-2 infections. Thus, we chose the United States, Canada, and Mexico from the North American region; the UK, Italy, France, Germany, Belgium, Spain, and Russia from Europe; Turkey, Singapore, South Korea, Qatar, Iran, Japan, Sri Lanka, India, China, Bangladesh, and Pakistan from Asia; Peru and Brazil from South America; Nigeria, Senegal, South Africa, Morocco, and Egypt from Africa; and New Zealand and Australia from Australia. The data were collected about the age group structure of the population of each country from the World Bank website., BCG vaccine coverage of the population, prevalence of hypertension (HTN), and prevalence of Type 2 diabetes mellitus (DM) were collected for each country from the WHO.,, We failed to get data on the number of tests done for Sri Lanka, Brazil, China, and Egypt. Hence, we could not calculate the number of detected cases per 100 tests. There is a lack of adequate information in regard to the temperature of each city in a given country at different times of the year and the contribution of each city to the number of COVID-19 cases. Therefore, in our study, for the purpose of analysis, we have assumed that the capital of each city represents the temperature of the entire country. We have further taken the entire number of positive COVID-19 cases in each country for analysis. On January 30, 2020, the WHO declared COVID-19 as a public health emergency of international concern. Hence, we presumed active worldwide spread might have occurred during the month of February and March 2020. The average maximum temperature and average minimum temperatures for the months of February and March of 2020 were taken for analysis.,, The total number of COVID-19-infected cases and total number of deaths due to COVID-19 in each country depends on the number of tests conducted in the country. Countries who have done lesser number of tests overall are going to report fewer cases and deaths, hence the number of cases and the number of deaths do not denote the real burden of the disease. Therefore, we have considered the number of COVID-19-positive cases for every 100 tests done and the number of deaths for every 100 confirmed cases of COVID-19 as better predictors of disease burden in the country. These two parameters were taken as dependent variables. All the data were collected till May 9, 2020. The effects of age structure, BCG vaccine coverage of the country, average maximum and average minimum temperature during the months of February and March, prevalence of Type 2 diabetes, prevalence of systemic HTN, and prevalence of smoking of the population were studied as independent variables.
The collected data were entered into Microsoft Excel version 16.37 and were analyzed using the software SPSS 20.00 version (IBM corporation, Bangalore, Karnataka, India). The statistical analysis was performed using the following statistical procedures. The Karl Pearson's correlation for relationships between two variables and a multiple linear regression model was applied for the combined effect of independent variables.
| Results|| |
The United States has conducted the maximum number (8,105,513) of COVID-19 tests in the world, which contributes to 27.83% of the total tests of the countries which we have studied. Senegal, on the other hand, has conducted the least number of test which is 19,369 (0.07%) among the countries studied. The United States has detected the maximum number of cases of COVID-19 in the world. It has confirmed 1283,929 cases of COVID-19. Among the countries we have studied, Sri Lanka has confirmed the least number of cases that is 835 cases. Among the countries we have studied, Mexico has the highest cases per 100 tests conducted (32.03%). It has conducted a total of 98,399 tests and detected 31,522 cases of COVID-19. New Zealand has the minimum cases per 100 tests conducted (0.65%) and has detected 1142 cases.
The United States has suffered the most from this pandemic. It has recorded a total of 77,180 deaths due to COVID-19, followed by the United Kingdom, Italy, and France, which have recorded 31,241, 30,201, and 26,230 deaths, respectively. Sri Lanka has reported the least number of deaths (only nine). France has a case fatality rate of 18.95%, which is the highest, and Qatar has the least mortality rate of 0.06% among the countries we studied.
Qatar, Japan, Bangladesh, China, Sri Lanka, and Morocco have 99% BCG coverage of their population, whereas developed countries such as the USA, Canada, the UK, Italy, Spain, France, Belgium, Germany, New Zealand, and Australia have no documented BCG coverage data. Nigeria has 44% of its population <14 years' age group, and Japan has 27% of its population aged 65 years or more. Nigeria has recorded the highest average maximum temperature during February and March 2020, and Canada has recorded the lowest average minimum temperature during the same months. The highest prevalence of HTN, Type 2 DM, and smoking has been observed in Pakistan, Qatar, and Russia, respectively, among the countries we studied. The detailed data are shown in [Table 1]. As discussed earlier, the number of COVID-19-positive cases for every 100 tests done and the number of deaths for every 100 confirmed cases of COVID-19 are better predictors of disease burden in the country. Therefore, these two parameters are taken as dependent variables. The effect of various other variables on these dependent variables is analyzed below.
Correlation between cases per 100 tests with other variables by Karl Pearson's correlation coefficients did not show any significant correlation. Stepwise linear regression of cases per 100 tests with other variables showed that the combined effect of number of deaths, number of test done, and BCG coverage is found to be positive and significant. It means that these variables can be taken as the best predictors to predict cases per 100 tests. Therefore, the prediction equation is given by:
Cases per 100 tests = 6.2480 + 0.004 (deaths) + 0.001 (tests done) + 0.0845 (BCG coverage).
The model is found to be significant with R = 0.6618, F = (3,18) = 4.6782, P < 0.05 and R2 = 0.4381. Combined contribution of above-said three significant predictors is 43.81%.
Correlation between the number of deaths for every 100 tests with other variables by Karl Pearson's correlation coefficients showed that deaths/100 cases had positively correlated with the number of tests done, number of positive cases, and number of deaths in each country. This has a negative correlation with BCG coverage of the country, percentage of population below the age group of 14 years, and average maximum temperatures during February and March 2020.
The combined effect of the average minimum temperature in March and number of confirmed cases is positive, but deaths are found to be negative and significant. It means that these are three best predictors to predict death per 100 cases. The regression model is given by:
No. of deaths per 100 cases = −8360.5373 + 7.1137 (min temp March) – 0.2317 (deaths) + 0.0006 (no. of confirmed cases).
This model is found to be significant with R = 0.9157 and F = 31.1660, P < 0.05.
It means that this model can be used to predict the number of deaths per 100 cases in future, with total contribution around 83.85%. The predicted and observed values of number of deaths by other variables are depicted in [Figure 1].
|Figure 1: The predicted and observed values of the number of deaths by other variables|
Click here to view
| Discussion|| |
The infection patterns and fatality rates of SARS-CoV-2, grievously affecting only a few selected countries, have raised many questions. After its origin in China, the way it has spread and severely affected the faraway places such as Italy, the United Kingdom, and the United States and relatively sparing the neighboring countries such as India and other Southeast nations has puzzled many epidemiologists. Most severely affected countries are high-income-developed countries such as Italy, the United Kingdom, and the United States. This observation has attracted global researchers to identify protective factors among relatively spared middle- and low-income countries. Many hypotheses have been proposed to explain this complex relation.
Sachin S. Gunthe et al. studied the potential effect of location-specific temperature and UV index on COVID-19. They found that 90% of the total confirmed cases in the world are confined to a narrow temperature range between 3°C and 12°C. They also noticed that higher UV index areas have a lesser number of cases. They went on to suggest that artificial UV radiation could be effective methods of sterilizing built-up environments for preventing viral spread. A study in Jakarta, Indonesia, concluded that the weather is an important factor in determining the incidence rate of COVID-19 in Jakarta.
A Japanese study by Ujiie et al. showed a possible association between low temperatures and the risk of COVID-19 infection. They demonstrated a strong correlation between lower temperatures and larger number of cumulative cases. In this study, they considered the number of cases for every million population as denominator. In larger countries such as India, the number of confirmatory tests done for the population was relatively low at the time of our data collection. Hence, considering the number of cases for every fixed number of population did not represent the true number of cases and the burden of the disease. Hence, we considered the number of cases for every 100 tests done. In our study, it was observed that the average maximum temperature during February and March had a significant negative correlation with the case fatality rate. It means that the lower environmental temperature correlated with higher fatality rate.
Studies have demonstrated that the BCG vaccine's immunomodulatory activity can safeguard against other respiratory infections. Hegarty et al. showed that the countries with a national program for universal BCG vaccination had a lower incidence and death rate from SARS-CoV-2. A study by Arts et al. showed that BCG vaccination conferred genome-wide epigenetic reprogramming in monocytes and correlated it with protection against other experimental viral infections. In our study, BCG vaccine coverage was negatively associated with case fatality rate. Better BCG coverage was associated with protective effect against COVID-19 death. On the other hand, BCG coverage did not influence the number of confirmed cases detected.
Worldwide data have shown almost negligible death rates among children. Children aged <18 years only account for 2% of severely affected patients. Possible explanation for this difference in severity of COVID-19 between adults and children might be due to the differences in receptors in the renin–angiotensin system and altered inflammatory responses to pathogens. Pro-inflammatory cytokines associated with neutrophil function increase with increasing age, which may be one more reason for the severity of acute respiratory distress syndrome in adults. Our study shows that case fatality rate had negatively correlated with the percentage of population in the age group of <14 years.
| Conclusions|| |
We conclude that higher environmental temperature, extensive BCG coverage, and higher proportion of younger age population in the low- and middle-income countries have a protective effect on infection with SARS-Cov-2 infection. Further research is required to understand the physiological basis for these protective effects against the COVID-19.
Strengths and limitations of the study
There are many limitations of our study; we did not collect the data for COVID-19-positive cases, and the data for the factors presumed to be influencing them from the exact same location. National data were collected pertaining to the total number of COVID-19 cases, number of deaths, BCG coverage, prevalence of DM, prevalence of HTN, tobacco consumption, and age distribution structure, whereas we considered the environmental temperature of the capital city of each country and presumed it to be representative of the overall climatic conditions of the country. This was done due to the lack of availability of accurate data at a local and provincial level. It was also not possible to track the data for all the variables from each city individually. There is a huge variation between the total number of tests conducted in each country, which may directly influence the burden of COVID-19 in each country. However, we tried to eliminate this discrepancy using cases detected per 100 tests done and deaths per 100 cases. The number of COVID-19 cases is rising relentlessly every day, and we are not completely sure as to where our findings will fit in the coming days.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Bhatnagar T, Murhekar MV, Soneja M, Gupta N, Giri S, Wig N, et al
. Lopinavir/ritonavir combination therapy amongst symptomatic coronavirus disease 2019 patients in India: Protocol for restricted public health emergency use. Indian J Med Res 2020;151:184-9.
] [Full text]
Holshue ML, DeBolt C, Lindquist S, Lofy KH, Wiesman J, Bruce H, et al
. First case of 2019 novel coronavirus in the United States. N
Engl J Med 2020;382:929-36.
Hegarty P, Kamat A, Zafirakis H, Dinardo A. BCG Vaccination May Be Protective against Covid-19; 2020.
Onder G, Rezza G, Brusaferro S. Case-fatality rate and characteristics of patients dying in relation to COVID-19 in Italy. JAMA 2020;323:1775-6.
Tosepu R, Gunawan J, Effendy DS, Ahmad OA, Lestari H, Bahar H, et al
. Correlation between weather and Covid-19 pandemic in Jakarta, Indonesia. Sci Total Environ 2020;725:138436.
Ujiie M, Tsuzuki S, Ohmagari N. Effect of temperature on the infectivity of COVID-19. Int J Infect Dis 2020;95:301-3.
Arts RJ, Moorlag SJ, Novakovic B, Li Y, Wang SY, Oosting M, et al
. BCG vaccination protects against experimental viral infection in humans through the induction of cytokines associated with trained immunity. Cell Host Microbe 2018;23:89. e5.