Projections of Corona Virus Infections in Bangladesh through SIR Modeling

By Topon Paul, PhD

Projections under 12 Scenarios

Note: These are the personal views of the researchers, which might be or might not be correct epidemiologically. These are simulation results but not actual results.

Source of Data: JHU HDX and IEDCR, Bangladesh

Observations and/or Additional Information

1. The most deciding factor in SIR modeling is Infection Rate (reproduction number), the average number of persons infected (new cases) by an already infected person (patient/case). Globally, the current estimated range of reproduction number of COVID-19 is 1.4-5.7. It appears that the current infection rate in Bangladesh is above 2.0; I have tried with infection rate of 2.0 but it does not produce meaningful results.
2. Not all the infected persons will be detected by PCR tests as many patients may not show any symptoms. The number of persons tested positive will be lower than the actual number of infected persons. For example, in New York City, estimation from the antibody testing is that about 1.7 million residents of NYC have had corona virus infections but only 138,000 cases had been confirmed by (PCR) tests at hospitals. Details: Intelligencer (2020/4/23)


Summary of Projections based on Preliminary Projections and Probable Scenario (#2) (2020/3/8-2020/9/30; these projections were made in April, 2020)
Total Infections	998,722~1,582,294
Probable Number of Patients with Symptoms	316,458
Peak of Infection	May 19-June 21, 2020
Total Deaths	4,746-23,734
Total expected Corona virus tests in Bangladesh	320,000-1,120,000 (5,000-10,000 testing capacity per day)
Total positive cases expected to be detected in Bangladesh	38,400-134,400 (12% of testing)

Note: 'End of Infections' is the date when the number of infections will drop below 10 persons. The infection will continue but I have used it to get a cut-off value.
Recovery means recovery from the disease or death when the person can no longer infect other persons.
The simulation results may change depending on the situations and the various counter-measures taken by the Govt. of Bangladesh.

Click on the image to get a larger one.

1. Optimistic Scenario	2. Probable Scenario	3. Worst Case Scenario

4. Extremely Optimistic Scenario with 0.001% Susceptible Population, Maximum 3.61 Infections through Contact and Recovery Rate of 11 days	5. Very Optimistic Scenario with 0.005% Susceptible Population, Maximum 3 Infections through Contact and Recovery Rate of 7 Days	6. Very Optimistic Scenario with Different Infection and Recovery Rates

7. Scenario with Maximum 2 Infections through Contact and Recovery Rate of 7 days	8. Optimistic Scenario with 0.01% Susceptible Population, Maximum 3.5 Infections Through Contact and Recovery Rate of 7.8 Days	9. Optimistic Scenario with 0.01% Susceptible Population, Maximum 4 Infections Through Contact and Recovery Rate of 10.8 Days

10. Optimistic Scenario with 0.01% Susceptible Population, and Other Parameters	11. Scenario with 1% Susceptible Population, Maximum 3 Infections Through Contact and Recovery Rate of 7 Days	Worst Case Scenario with 1% Susceptible Population, Maximum 4 Infections through Contact and Recovery Rate of 11 Days

Assumptions:

Population of Bangladesh: 165,000,000
Infection rate range (basic reproduction number): [1.4, 5.7]
Recovery rate range: [7, 21] days

Citation

Topon Paul, PhD (2020): Projections of Corona Virus Infections in Bangladesh through SIR Modeling.

Disclaimer

These models are built using the COVID-19 infection data of Bangladesh and by fitting the models with the available data until April 27, 2020. The simulation results may change depending on the situations and the various counter-measures taken by the Govt. of Bangladesh.
These are research results. There is no warranty that the projection results will reflect the real situations of Bangladesh. Therefore, use these projection results at your own risk.
I have tried my best to give you the accurate results based on the data but there is no warranty that these results are 100% accurate.
The researcher is NOT responsible for any damage incurred by using the projection results.

Page Last update: 2020/6/24 22:20 (JST).