Background: Ever since the Coronavirus disease (COVID-19) outbreak emerged in China, there has been several attempts to predict the epidemic across the world with varying degrees of accuracy and reliability. This paper aims to carry out a short-term projection of new cases; forecast the maximum number of active cases for India and selected high-incidence states; and evaluate the impact of three weeks lock down period using different models. Methods: We used Logistic growth curve model for short term prediction; SIR models to forecast the maximum number of active cases and peak time; and Time Interrupted Regression model to evaluate the impact of lockdown and other interventions. Results:The predicted cumulative number of cases for India was 58,912 (95% CI: 57,960, 59,853) by May 08, 2020 and the observed number of cases was 59,695. The model predicts a cumulative number of 1,02,974 (95% CI: 1,01,987, 1,03,904) cases by May 22, 2020. As per SIR model, the maximum number of active cases is projected to be 57,449 on May 18, 2020. The time interrupted regression model indicates a decrease of about 149 daily new cases after the lock down period, which is statistically not significant. Conclusion:The Logistic growth curve model predicts accurately the short-term scenario for India and high incidence states. The prediction through SIR model may be used for planning and prepare the health systems. The study also suggests that there is no evidence to conclude that there is a positive impact of lockdown in terms of reduction in new cases.

Background As the number of COVID-19 cases continues to rise, public health efforts must focus on preventing avoidable fatalities. Understanding the demographic and clinical characteristics of deceased COVID-19 patients; and estimation of time-interval between symptom onset, hospital admission and death could inform public health interventions focusing on preventing mortality due to COVID-19. Methods We obtained COVID-19 death summaries from the official dashboard of the Government of Tamil Nadu, between 10th May and July 10, 2020. Of the 1783 deaths, we included 1761 cases for analysis. Results The mean age of the deceased was 62.5 years (SD: 13.7). The crude death rate was 2.44 per 100,000 population; the age-specific death rate was 22.72 among above 75 years and 0.02 among less than 14 years, and it was higher among men (3.5 vs 1.4 per 100,000 population). Around 85% reported having any one or more comorbidities; Diabetes (62%), hypertension (49.2%) and CAD (17.5%) were the commonly reported comorbidities. The median time interval between symptom onset and hospital admission was 4 days (IQR: 2, 7); admission and death was 4 days (IQR: 2, 7) with a significant difference between the type of admitting hospital. One-fourth of (24.2%) deaths occurred within a day of hospital admission. Conclusion Elderly, male, people living in densely populated areas and people with underlying comorbidities die disproportionately due to COVID-19. While shorter time-interval between symptom onset and admission is essential, the relatively short time interval between admission and death is a concern and the possible reasons must be evaluated and addressed to reduce avoidable mortality.

Background: Since the onset of the COVID-19 in China, forecasting and projections of the epidemic based on epidemiological models have been in the centre stage. Researchers have used various models to predict the maximum extent of the number of cases and the time of peak. This yielded varying numbers. This paper aims to estimate the effective reproduction number (R) for COVID-19 over time using incident number of cases that are reported by the government. Methods: Exponential Growth method to estimate basic reproduction rate R 0 , and Time dependent method to calculate the effective reproduction number (dynamic) were used. "R0" package in R software was used to estimate these statistics. Results: The basic reproduction number (R 0 ) for India was estimated at 1.

Background Global research is running towards to find a vaccine to stop the threat of the COVID-19. The Bacillus Calmette–Guérin (BCG) vaccine that prevents severe forms of tuberculosis is getting more attention in this scenario. The objective of our study was to determine the association between BCG vaccine coverage and incidence of COVID-19 at a national-level across the Globe. Methods The data of 160 countries were included in the study. Meta-regression was done to estimate the difference in the incidence of COVID-19 cases between countries with BCG vaccination coverage. BCG coverage was categorized as ≤70%, >70% and no vaccination. The analyses were carried out by adjusting for factors such as population density, income group, latitude, and percentage of the total population under age groups 15–64 and above 65 years of each country. Results The countries that had ≤70% coverage of BCG vaccine reported 6.5 (95% CI: −8.4 to −4.5) less COVID-19 infections per 10,000 population as compared to countries that reported no coverage. Those that had >70% coverage reported 10.1 (95% CI: −11.4 to −8.7) less infections per 10,000 population compared to those with no BCG countries. Conclusion Our analysis suggests that BCG is associated with reduced COVID-19 infections if the BCG vaccine coverage is over 70%. The region-wise analyses also suggested similar findings, except the Middle East and North African region.

Introduction: At the end of the second week of June 2020, the SARS-CoV-2 responsible for COVID-19 infected above 7.5 million people and killed over 400,000 worldwide. Estimation of case fatality rate (CFR) and determining the associated factors are critical for developing targeted interventions. Methodology: The state-level adjusted case fatality rate (aCFR) was estimated by dividing the cumulative number of deaths on a given day by the cumulative number confirmed cases 8 days before, which is the average time-lag between diagnosis and death. We conducted fractional regression analysis to determine the predictors of aCFR. Results: As of 13 June 2020, India reported 225 COVID-19 cases per million population (95% CI:224-226); 6.48 deaths per million population (95% CI:6.34-6.61) and an aCFR of 3.88% (95% CI:3.81-3.97) with wide variation between states. High proportion of urban population and population above 60 years were significantly associated with increased aCFR (p=0.08, p=0.05), whereas, high literacy rate and high proportion of women were associated with reduced aCFR (p<0.001, p=0.03). The higher number of cases per million population (p=0.001), prevalence of diabetes and hypertension (p=0.012), cardiovascular diseases (p=0.05), and any cancer (p<0.001) were significantly associated with increased aCFR. The performance of state health systems and proportion of public health expenditure were not associated with aCFR. Conclusions: Socio-demographic factors and burden of non-communicable diseases (NCDs) were found to be the predictors of aCFR. Focused strategies that would ensure early identification, testing and effective targeting of non-literate, elderly, urban population and people with comorbidities are critical to control the pandemic and fatalities.