ABSTRACT
The novel coronavirus (2019-nCoV) pandemic’s global regulatory authorities are issuing provisional market authorization to current public health for timely detection of COVID-19 infections based on the analytical performance data, limited clinical performance data, and manufacturers’ declarations. Importers, local manufacturers, government procurement agencies, and other stakeholders should be made aware of the regulatory process for market authorization of COVID-19 in vitro diagnostic (IVD) kits in India. The objective is to study in detail about COVID-19 IVD guidelines. Concentrating on the recent updates in COVID-19 IVD guidelines, we analyze the necessity for India to become efficient in manufacturing SARS-CoV-2 IVDs. In India, Central Drugs Standard Control Organization (CDSCO) coordination with The Indian Council of Medical Research (ICMR) is committed to ensuring the rapid availability of diagnostic tests for COVID-19. To address the urgent need for timely detection of COVID-19 infection, it is important to scale-up the testing capacity to maximum possible levels according to the CDSCO interim licensing process. This has helped to timely expand the variety of quality diagnostic tests available in India. The regulatory tools described in this article are made based on a comparison of data available on the procedure, followed by other major countries under the health emergency without compromising on the quality of the product within the existing provisions given in the Drugs and Cosmetic Act and Medical Devices Rules 2017. © 2022
ABSTRACT
Background: Several mathematical models have been developed to project the COVID-19 cases since the beginning of the pandemic, which are helps for administration to plan effectively and strengthening public health services and facilities. Many researchers incorporated long short term memory model (LSTM) and reported a better performance as compared to classical models. However, they have not clearly explained the workflow and steps involved in the LSTM models in a simple manner. Moreover, these models are not popularized among many biomedical researchers due to lack of expertise. Therefore, our aim of the study is to present such models as a tutorial, in a simple way with Python codes using real time data with good amount of instructions so that many researchers could use these methods to project pandemics like COVID-19.Data and Methods: A daily case in India from 1st December 2021 to 10th February 2022, and in UK from 1st May 2021 to 10th February 2022 was used to train the models. We used Convolutional Long Short Term Memory (CNN-LSTM) model and simple LSTM models to forecast COVID-19 cases. Both models were validated using data from 11th - 25th February 2022.Results: CNN-LSTM and simple LSTM models fit very well with R2 0.95 and 0.97 for India. Validation data, RMSE and MAE of CNN-LSTM model were 9972.81 and 8993.33. And these were 19285.57 and 18870.87 for simple LSTM model. The R2 value of CNN-LSTM and simple LSTM models for UK data were 0.77 and 0.84 respectively. And validation data, RMSE was 12111.95 for CNNLSTM and 8935.75 for simple LSTM. The MAE of CNN-LSTM and simple LSTM model were 10060.9 and 7824.821 respectively.Conclusion: Simple LSTM model works better than CNN-LSTM model while training. The performance of CNN is better in the validation for India. Therefore, our suggestion is, train various models instead of sticking into one model to project the future cases and report the range of values. Revise the models weekly once or once in a couple of week as the behaviour of an epidemic may change over a time.
Subject(s)
COVID-19ABSTRACT
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 R0, 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 (R0) for India was estimated at 1.379 (95% CI: 1.375, 1.384). This was 1.450 (1.441, 1.460) for Maharashtra, 1.444 (1.430, 1.460) for Gujarat, 1.297 (1.284, 1.310) for Delhi and 1.405 (1.389, 1.421) for Tamil Nadu. In India, the R at the first week from March 2-8, 2020 was 3.2. It remained around 2 units for three weeks, from March 9-29, 2020. After March 2020, it started declining and reached around 1.3 in the following week suggesting a stabilisation of the transmissibility rate. CONCLUSION: The study estimated a baseline R0 of 1.379 for India. It also showed that the R was getting stabilised from first week of April (with an average R of 1.29), despite the increase in March. This suggested that in due course there will be a reversal of epidemic. However, these analyses should be revised periodically.
ABSTRACT
Since December 2019, a novel coronavirus disease (COVID-19) creates a global threat. Medical students are more susceptible to be infected by the virus. This study aimed to assess COVID-19 related knowledge, attitude towards COVID-19, and preventive behaviours against COVID-19 among medical students within the first month of the onset of the outbreak in Malaysia. We collect data from medical students using an online Google survey form. Out of 696 students, 467 responded to the questionnaire. The analysis revealed that the mean percentage of knowledge was (85.04), attitude (84.12), and preventive practice (77.75) respectively. Hierarchical multiple linear regression analysis revealed that living with family (p<0.01) and knowledge of COVID-19 (p<0.001) appeared to be important predictors of attitude toward COVID-19. However, gender (p<0.001), living status (p<0.001), frequency of travel during movement control order (p<0.01), attitude towards COVID-19 (p<0.001) have appeared significant predictors for preventive practice against COVID-19. But knowledge of COVID-19 had no impact on preventive practice against COVID-19 (p>0.05). We found a high level of COVID-19 related knowledge, attitude, and preventive practice against COVID-19 among medical students. A sustained knowledge, attitude, and preventive behavioural strategy could play an ingredient in upholding the student’s learning and practice against any disease like COVID-19. © 2020
ABSTRACT
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.