Every year, the RSS president personally invites someone to give the Campion lecture at the RSS conference and this year, Peter Diggle invited Professor Christl Donnelly to give a talk regarding the role of statistics in responding to the recent Ebola outbreak in west Africa.
The recent Ebola outbreak was the biggest the world had ever witnessed. To date, more than 28,000 cases have been reported, of which there were more than 11,000 deaths.
In August 2014 August, the World Health Organisation (WHO) declared a public health emergency. This is when Christl Donnelly, Professor of Statistical Epidemiology in Imperial College and leading member of the WHO Ebola response team, began to get data on the outbreak.
Athough there was a considerable response team, events conspired to let the outbreak gain momentum. In September 2014, the basic reproduction number (the average number of additional cases one case generates) was between 1.5 and 2.3. If reproduction figures are more than 1, it means the epidemic is still expanding. ‘We estimated in September 2014, that the number of outbreaks may exceed 20,000 if nothing improved,’ Christl says.
Reporting the progress of the disease spread presented considerable challenges for the response team. There were country-specific databases which had to be cleaned and transferred into regularised formats so they could be combined. Data was collected frantically which meant dates were sometimes wrong. However, Christl was keen to pay tribute to those collecting data in the field. ‘We compiled a survey for people to fill out,’ she said. ‘People went to great lengths to collect this data. We were lucky to get as much data as we did.’
Databases were coming in every few days which enabled the team to send regular reports to WHO. Initially, Christl said, it wasn’t known how the disease affected different age groups, but it’s been established that most patients were aged 15-45. Around 5% of those affected were healthcare workers which threatened the response teams on the ground.
Tracking and estimating future disease spread was important to work out, for example, how many beds were required. ‘All beds were overwhelmed at some point so we tried to anticipate need,’ Christl recalls. ‘We used projection scales if growth looked exponential.’
The data helped ascertain where those infected may have caught it. It also helped to inform decisions around personal protective equipment and preventative measures. A rapid diagnostic test was developed to prevent people coming in to centres to be diagnosed, as it was found that some people were actually getting infected coming in.
By July 2015 the reproduction number had dropped to 0.7-0.9. However, ‘flare ups’ still occured, such as one in Liberia after it was declared Ebola-free. Some Ebola survivors have it in their blood and every new case has potential to spread. Post viral symptoms are now better detected, however – and now there is more follow-up.
The story ends on a positive note – in June WHO declared Liberia Ebola-free, although there is still close monitoring of the situation to ensure that any new cases are identified and contained quickly.
After her lecture, Christl was presented with the William Guy lecturer certificate by Peter Diggle (pictured right) as she takes the mantle for the coming year to give talks in schools.