Coronavirus: Covid-19 spread in Abu Dhabi unlikely to accelerate, AI modelling shows

Coronavirus infection rates in Abu Dhabi are “unlikely to accelerate" over the next six weeks due to the effective measures in place, artificial intelligence modelling has shown.

Seha, the government health organisation that owns and operates Abu Dhabi’s public hospitals, is using AI from a UK company to understand capacity needs and plan its response to the pandemic. The results have shown that the current rate of infection is not expected to accelerate.

Seha has been on the front lines of the UAE’s coronavirus offence, with 60 outpatient clinics and 12 hospitals throughout Abu Dhabi, as well as operating the 14 drive-through testing centres throughout the UAE. Seha is also now building three field hospitals in Abu Dhabi and Dubai. The UAE has a total of 8,238 confirmed cases of Covid-19 and 1,546 recoveries as of Wednesday.

"We felt that [AI] would provide important insights into the Covid-19 virus and also help predict the virus trend and enable us to better prepare our response strategy," said George Kapitelli, the Seha Group chief financial officer. "This is being used to enable Seha to get in front of the required demand for beds, in particular ICU beds, for the more seriously impacted patient.“

A lack of protective equipment, shortage of hospital beds and overloaded ICUs are global challenges. With AI-based forecasting tools, hospitals can better manage their resources, according to an assessment by Dr Bertalan Mesko, a medical futurist.

Draper & Dash Predictive Healthcare Analytics’ Covid-19 impact assessment tool, which is also being used by the UK’s National Health Service (NHS), takes data from individual hospitals to understand usual levels of demand for beds or emergency care, then combines this data with what has already been seen from the spread of the virus across the globe.

Abu Dhabi's health care system has enough capacity to cope with considerably higher volumes of the outbreak, Orlando Agrippa, the chief executive of Draper & Dash, told The National. But looking ahead six weeks, "there is nothing that concerns us".

“We don't expect to see considerably higher volume,” he said.

In late February, as the pandemic picked up speed worldwide, the company took six weeks and 10 data scientists and employees with clinical backgrounds to repurpose their existing AI health care platform to address Covid-19.

A forecast of upcoming demand on a given health care facility is created using each provider’s data, run through several different forecasting models. The results provide the basis upon which healthcare teams can plan their next steps in combating the effect of the virus, working to ensure that patients have a place to go and are able to receive the care they need.

“Abu Dhabi is one of the unsung heroes in my opinion. The growth of the virus and the acceleration of it is pretty slow,” Mr Agrippa said. He highlighted widespread testing for the virus, tight and early border controls and a reliance on clinical insights and data as the biggest factors.

“Leadership around this has been pretty next to none,” he said.

“What we've seen is [Abu Dhabi] has embraced technology. The countries who have done that, wherever there is a lot of data leveraged to manage the spread of this virus, those countries have done well.”

There is a level of complexity in dealing with the virus, given asymptomatic patients, mildly infected patients, moderately infected patients and severely infected patients all need care and to be removed from the population to avoid infecting others.

Asymptomatic or very mild cases are being managed at homes or in hotels, while more serious cases are treated in hospitals.

As for Abu Dhabi's readiness, using the insights from the Draper and Dash assessment tool is helping Seha to increase bed capacity at different levels to support quarantined, acute and critical care patients.

"The massive challenge to fight the Covid-19 virus continues," Mr Kapitelli said. "We are certainly more prepared."