• Sunday, Jul 12, 2020
  • Last Update : 06:42 pm

DT Webinar: Lack of data makes it hard to see the exact situation of the pandemic in Bangladesh

  • Published at 10:06 pm May 29th, 2020
Webinar

Rumi Ahmed Khan MD FCCP, Director, Pulmonary Critical Care Fellowship Program, Director Cardiopulmonary Exercise and Lung Physiology Lab, Orlando Health, Associate Professor, UCF College of Medicine; Malay Kanti Mridha, Professor of Public Health, Director, Center for Non- Communicable Diseases and Nutrition. BRAC James P Grant School of Public Health; Rajiv Chowdhury, Principal Researcher/Associate Professor in Global Health, Department of Public Health and Primary Care, University of Cambridge,UK; Dipak Kumar Mitra, Professor and Chairman, Department of Public Health, School of Health and Life Sciences, North South University; Enayetur Raheem, PhD, Senior Data Scientist at Cigna, USA, Scientist and Head of Research, Biomedical Research Foundation (BRF), Bangladesh participated in the webinar which was moderated by Dhaka Tribune Editor Zafar Sobhan

Experts have said that lack of accessible data in Bangladesh has made it difficult to map out Bangladesh-specific models which could predict potential Covid-19 spread and mortality rate.  

They shared the opinion at a webinar titled Tribune Talks organized by Dhaka Tribune on Friday.

The experts said Bangladesh can seek inspiration from the models other hard-hit countries are following.

They said localized data would help map out models to make predictions more effectively and help the authorities to strategize to flatten the Covid-19 curve.

"Experts need to have access to data to make any prediction or design any model. We are not getting enough data and for this reason it is difficult for us to predict how many lives Covid-19 is going to claim in Bangladesh," said Malay Kanti Mridha, director of Centre of Excellence for Non-Communicable Diseases and Nutrition at Brac James P Grant School of Public Health.    

He said the government is releasing district-specific Covid-19 situation data and some works could be done based on the data. 

"I cannot predict what the mortality rate would be in the coming days in Bangladesh but a forecast can be done based on the number of deaths caused by the virus." 

He said any Bangladesh-specific prescriptive modelling could help reduce the reproduction number to less than one.

While moderating the webinar, Dhaka Tribune Editor Zafar Sobhan said the authorities are very sensitive about this issue and "they are telling us not to publish anything speculative." 

Echoing Sobhan, Enayetur Raheem, senior data scientist at the USA-based global health service company Cigna, said: "We cannot do modelling because the government is sensitive. 

"I speculate using the data...I am not getting the data I used to get from official and unofficial sources from Bangladesh and therefore, I could not make a model based on most recent data from Bangladesh."

Pointing to the importance of forecasts based on country-specific data, he said: "Everyone was happy seeing the Singapore model which predicts the curve will fall flat in May. Now we all can see where we stand."

'Lockdown alone cannot flatten Covid-19 curve'

The discussants also shed light on the fact that lockdown alone cannot do much to flatten the curve of coronavirus transmission if other preventive measures are not being followed. 

Rumi Ahmed Khan, associate professor of UCF College of Medicine in USA, pointed out that lockdown cannot work effectively to control Covid-19 transmission if the authorities do not work on other measures like ramping up testing capacity, enhancing the healthcare system, and making good policies.  

"Lockdown is one piece of a puzzle. It can work effectively if other pieces of the puzzle are in place. If we keep pushing the lockdown, we will only delay the surge. We would not really make anything tangible to make the curve flat," he said.

Sobhan questioned: "If you look at the trajectory in countries like the United States, Italy, and Spain, by this time we should have many more cases than we have here in Bangladesh...Does that mean things will be different in Bangladesh or it will happen in future?" 

"We are seeing exponential growth in case detection. On March 15, the case number was five, on April 15 case numbers crossed 1,200, and on May 15 the case numbers crossed 20,000. We do not know what the case number would be on June 15," Rumi said.

Referring to the UK's Imperial College forecast, Prof Dipak Kumar Mitra, chairman of the Department of Public Health at North South University, said they collected data from China in their model and then applied it to predict Covid-19 scenarios in western countries.

He said reporting on overall Covid-19 deaths, cases and recoveries could make a difference in Bangladesh. 

"USA predicted there would be 100,000 deaths and they already crossed the threshold," he said, adding that he doubts the reported numbers in Bangladesh are accurate.

61
59
blogger sharing button blogger
buffer sharing button buffer
diaspora sharing button diaspora
digg sharing button digg
douban sharing button douban
email sharing button email
evernote sharing button evernote
flipboard sharing button flipboard
pocket sharing button getpocket
github sharing button github
gmail sharing button gmail
googlebookmarks sharing button googlebookmarks
hackernews sharing button hackernews
instapaper sharing button instapaper
line sharing button line
linkedin sharing button linkedin
livejournal sharing button livejournal
mailru sharing button mailru
medium sharing button medium
meneame sharing button meneame
messenger sharing button messenger
odnoklassniki sharing button odnoklassniki
pinterest sharing button pinterest
print sharing button print
qzone sharing button qzone
reddit sharing button reddit
refind sharing button refind
renren sharing button renren
skype sharing button skype
snapchat sharing button snapchat
surfingbird sharing button surfingbird
telegram sharing button telegram
tumblr sharing button tumblr
twitter sharing button twitter
vk sharing button vk
wechat sharing button wechat
weibo sharing button weibo
whatsapp sharing button whatsapp
wordpress sharing button wordpress
xing sharing button xing
yahoomail sharing button yahoomail