Loading...
The Financial Express

Fighting Covid-19: Data analytics a tested way  

Himadri Nandi   | Published: May 16, 2020 21:38:17 | Updated: May 22, 2020 21:59:07


Fighting Covid-19: Data analytics a tested way   

As the pandemic is unfolding, the world grapples with health and economic crisis. When everything is at a standstill, technology has come to save the day. Data analytics in particular is playing a pivotal role. It is addressing both targets of the pandemic: healthcare and economy. Pandemic analysis we can call it, is processing raw data into compatible formats and designing forecasting models to derive cognitive decisions. Data scientists are joining forces with epidemiologists to stem the outbreak. International bodies, policymakers and researchers are relying on the prediction models to follow contagion progression and take essential decisions.

Few countries that have been successful in "flattening the curve" took data-driven initiatives at a very early stage. Being a neighbouring Island to China, Taiwan was feared to be severely affected. Yet, taking timely measures through capitalising on data, Taiwan averted the situation. They have integrated National Health Insurance Databases with National Immigration and Customs databases to track down people with recent travel history. Contact tracing and symptom tracking were conducted thoroughly. As a result, Taiwan controlled the transmission of the virus to a minimum and they have not seen any local transmission in the last 26 days. The country did not impose any lockdown and even schools remained open.

South Korea can be an example, too. It was hit hard by MERS-CoV in 2015 and used the experience to strengthen public healthcare with big data analysis. Smart contact tracing, app-based patient symptom and mobility monitoring, real time healthcare capacity monitoring and data sharing between hospitals are some examples where data-driven outbreak analysis helped South Korea to flatten the curve earlier than the rest of the world.

Coming to the case of Bangladesh, the situation is not quite the same. The growing numbers of infected people underscore a sobering reality for Bangladesh. Undoubtedly, the real picture is way more gruesome as the testing facility is yet to expand both in terms of volume and accessibility. DGHS and IEDCR websites are providing daily data of test numbers, number of those infected, fatalities, age and geo-distribution of the patients. The statistics are important but they are not sufficient to answer the bigger and burning questions. Bangladesh also should take immediate initiative to bring in data analytics into the scene.   

A central data repository must be built where data from various stakeholders will be stored and processed. COVID-dedicated hospitals must submit information of patients, confirmed deaths, treatment details, ICU and ventilator usages etc. Non-COVID healthcare centres should share the suspected cases and fatality details. Logistical data like stock of lifesaving drugs, oxygen, tools and protective gears are critical. Data from patients who are recovering from Covid-19 are significant, too. Their conditions should be closely monitored via a dedicated support system.

Contact tracing proved to be the most successful tool to stop the local transmission. Despite the initiatives to track traveller flow and enforce home quarantine, the result was not fruitful. But, it's not late yet as the virus is still at large. In the coming days, thousands of migrant workers might lose job and come back home. It's high time to think of a solution that integrates mobile operators and government stakeholders like the Department t of Immigration, Civil Aviation, local administration and law enforcement agencies.

Next step is to process all the epidemiological data, travel information, mobility logs, etc. A team of epidemiologists and data analysts will work jointly to prepare various predictive algorithmic models. Leveraging machine learning, Artificial Intelligence algorithms and decision-making models will be prepared. Based on requirement and attributes, these models will generate outcomes with minimum uncertainty.

One such possible model can be used for healthcare management based on the related variables. Doctors may know the effectiveness of various treatments and can curtail loss of precious life. For example, doctors can know which treatment plan is working better for COVID patients with kidney complication and administer that accordingly. Hospitals can also prepare for sudden surge and allocate staffs and resources.

Another important model can be prepared for contagion progression tracing. Aggregated mobility information from telecom operators backed by immigration data will help track traveller migration and contain the outsourced infection number. It will allow quantifying human mobility in areas and predict how the virus may spread. Government can identify high risk clusters, susceptible people and emphasise testing accordingly.

Integration of data analytics itself will bring about a paradigm shift in healthcare, but the complete task is time- consuming. With the coronavirus looming over us, the  government can phase out the project and fast-track the part that is dedicated for Covid19. Bangladesh has the technical expertise required. Forthe last couple of years, private IT-ITeS organisations are coming up with effective solutions to bridge technology and healthcare gap. With their and the ICT ministry's help, the system can be launched sooner than expected.

Nevertheless, there are various challenges to be addressed. Data reporting and data availability are the critical ones. The existing legacy of data reporting modality has to change and a phased shifting from paper-based data collection will be required. Acquisition of mobility data will raise privacy and confidentiality concerns. The government has to ensure budget and incorporate technology in the shortest possible time. Finally, the shortage of training and expertise must be addressed.

The pandemic has ushered Bangladesh into a unique situation. The choice between life and livelihood is a complex issue. Economy falters when a country remains in lockdown for long. Reopen too soon and you will face deadly backlash. With 36 million vulnerable people below the poverty line and an already overburdened healthcare system, we cannot really afford decisions based on gut feeling. We must trust unequivocal science-driven methods and take those into account. Proper collection and analysis of data will arm healthcare professionals and policy-makers to tackle the resurgence of COVID19 or any future healthcare crisis. If implemented successfully, the initiative will complement National Digital Health Strategy'19 and National Plan for Disaster Management.

 

himu_385@yahoo.com

 

Share if you like