On June 25, the Sasin Research Seminar series continued online with a fascinating talk by Assistant Professor Chonawee Supatgiat, Ph.D., from the Sasin Finance Faculty. The talk looked at the “Effects of control measures and their impacts on COVID-19 transmission dynamics”. The study arose in response to the pandemic. For well over a year, governments around the world have been grappling with methods to fight COVID-19. Various control measures have been used, including face masks in public places, social distancing, travel limitations, bans on mass gathering, contact tracing, mass testing, population education and engagement, and the closures of schools, businesses, and restaurants. However, there have been no studies that rank the interrelated effectiveness of these measures, and this knowledge gap can result in government responses that are uncoordinated and inefficient. The study aims to change that. The lecture began with Assistant Professor Chonawee looking at new global cases, followed by a more detailed analysis of the USA, Brazil, and India. In the USA, infections have fallen sharply thanks to aggressive vaccination and control measures. Brazil’s more relaxed approach has resulted in a steady rise, and India saw a massive spike in April brought down by a nationwide lockdown. Vaccination rates in these countries were also compared. While there were only a few cases last year in Thailand, the situation has now changed, with a rise in numbers starting in April. It is currently the 14^th in the world for daily new cases. To see the dynamic of the epidemic, Assistant Professor Chonawee talked about the standard epidemiology model SIR, which looks at those who are Susceptible, Infectious, and Recovered. However, due to the nature of the virus and the incubation period, the compartmental model became SEIR after Exposed was added. One major shortfall of SEIR models is the lack of the abilities to capture the dynamic temporal transmission and recovery probabilities of the disease. The probability of when symptoms first appear (incubation period), is not constant but instead depends on how long the patient has received the virus. The chance of passing on an infection (infectivity probability) is also not constant but depends on how long the patient has the first symptom. Moreover, the chance that an infected person can pass the virus on to others is also not constant but depends on how long the patient has received the virus. Moreover, there are some factors that could not be captured in the SEIR compartmental models when looking at their effectiveness at controlling COVID-19. For example, some people might not comply with the government’s rules and around a third of cases are asymptomatic where they cannot be detected and isolated. To resolve the shortfalls and to capture the important factors, Assistant Professor Chonawee developed a Discrete Time Markov Chain model that factors in non-compliance and asymptomatic cases. It also accounted for dynamic temporal transmission and recovery probabilities. The new model has a total of 89 compartments. The model’s parameters allow for the fact that measures and population activities change over time. These parameters include the reproduction number, cooperative percentages, undetected infected proportion, reduction in activities, testing delay, tracing delay, and contact tracing coverage. To show model implementation, Assistant Professor Chonawee used two case studies – the USA and Thailand – to see the effect of different control measures. The data was presented in a tornado diagram looking at the worst case, base case, and best case. The results show that the most effective measures in the USA were reducing the reproduction rate, population education and engagement, and reducing the testing delay. Thailand had similar results, except that a vaccination program was far more critical than reducing the testing delay. Those involved in the study had given warnings about the potential dangers of travel during the Songkran holiday, but they were ignored. Further warnings were issued that stricter measures were needed, or cases would rise and risk a health system breakdown, which also went unheeded. Assistant Professor Chonawee then showed what the study had predicted for different scenarios and revealed the rise in cases had matched their calculations. Future predictions were then presented, along with the assumptions made. The lecture ended by identifying items of serious concern. These included the recent drop in testing by hospitals, vaccine shortages, and an increasingly uncooperative population. Assistant Professor Chonawee then gave a strong recommendation for a swift and hard lockdown. It is hoped that if implemented, the study’s novel and unique model could be of benefit to other countries as they make pandemic control decisions. A lively Q&A followed the talk.