Work

The role of modelling in planning and budgeting for South Africa’s COVID-19 response

Year: 2022

Type: article

Abstract: The unprecedented health response to the South African COVID-19 epidemic saw expert virologists, clinicians, epidemiologists and modellers coming together to provide scientific insight and expertise in the management of the epidemic. One such group was the South African COVID-19 Modelling Consortium, established in late March 2020 to support government’s planning and budgeting for COVID-19-related health care. Here we report on the approach taken to modelling cases, hospital admissions, deaths and costs in a situation of substantial uncertainty, the processes supported by the models, and the lessons learnt for future pandemics. We developed two tools: the National COVID-19 Epi Model (NCEM) and the National COVID-19 Cost Model (NCCM). The NCEM is a stochastic compartmental transmission model developed to project incidence of COVID-19 across the provinces (and later, districts) of South Africa. The model follows a generalised Susceptible-Exposed-Infectious-Removed structure accounting for disease severity (asymptomatic, mild, severe and critical cases) and treatment pathways (out-patient and in-patient [ICU and non-ICU care]). The NCCM uses NCEM outputs and information on likely need for and baseline availability of resources such as human resources at all care levels, oxygen, oxygen delivery devices, SARS CoV-2 tests, infection control and prevention infrastructure, and public-sector costs to project the total COVID-19 resource needs and budget impact. Between March and September 2020, we created 38 sets of updates to the models which were shared with stakeholders in the National and Provincial Departments of Health, National Treasury and other partners. We published four sets of short- and long-term NCEM projections and the model code, and created an online dashboard with data and visualisations of all outputs. NCCM results were used to inform the COVID-19 Special Adjustment Budget. Regularly updated assumptions included contact and mobility rates, case fatality, severity, types of hospital care and length of stay, resource availability, costs, and the impact of behavioural factors such as adherence to restrictions. The South African COVID-19 epidemic provided an opportunity for disease modelling to be a source of scientific evidence for decision-making.

Source: South African Health Review

DOI: 10.61473/001c.71474

Authors: Sheetal P Silal, Juliet Pulliam, Lise Jamieson, Brooke Nichols, Harry Moultrie, Gesine Meyer-Rath