Covid-19: What lies ahead for us in winter? Expert modellers weigh in
Epidemiologist in public health surveillance and geospatial modelling Dr Harry Moultrie.
Advertisement
Health Minister Zweli Mkhize has discussed South Africa’s trajectory on Covid-19 with expert modellers weighing in.
During an evening Zoom meeting on 19 May, an epidemiologist in public health surveillance and geospatial modelling Dr Harry Moultrie spoke about the work done by the team of modellers over the last six weeks.
The purpose of modelling and special models is to prepare for needs, such as hospital and ICU beds.
Other models are being created, including budget, cost, and epi models (SEIR).
The SEIR (Susceptible – Exposed – Infectious – Recovered) model is generalised and makes predictions on disease severity (asymptomatic, mild, severe, critical), as well as treatment pathways (outpatients, non-ICU, ICU).
See figure below:
Dr Sheetal Silal, senior lecturer in the University of Cape Town (UCT) department of statistical sciences, explained there are two scenarios – optimistic and pessimistic. These scenarios are used to create trajectories for a variety of issues such as infection rates and ICU bed threshold.
See figure below:
“What we see is that the initial lockdown [and the extension to five weeks] has bought us critical additional time to prepare for the wave of active cases. There will likely be a peak in active cases between early July (pessimistic) and early August (optimistic). This will be affected by further post-lockdown measures,” said Silal.
The elephant in the room was also addressed.
Should bigger metros like Cape Town and Johannesburg be kept in level four?
Silal said it would be irresponsible for modellers to say whether the lockdown should be lifted. Their models can show the numbers, but ultimately it was an inter-governmental decision.
Silal continued, saying there was large uncertainty in their death and other predictions. Since some deaths are reported outside the hospital administration, the team acknowledges this and calibrates their parameters to accommodate uncertainty.
