key: cord-0759674-pdpot1wo
authors: Madhi, Shabir A; Nel, Jeremy
title: Epidemiology of severe COVID-19 from South Africa
date: 2021-08-04
journal: Lancet HIV
DOI: 10.1016/s2352-3018(21)00183-1
sha: 42c44365abb20db267c1906dcc3c858f0f2b873c
doc_id: 759674
cord_uid: pdpot1wo

nan

Despite the global importance of the COVID-19 pandemic, its impact remains poorly characterised in low-income and middle-income countries, including most of those in Africa. As of Aug 1, 2021, only 3·4% of recorded COVID-19 cases and 4·0% of COVID-19-related deaths were from Africa, where 17% of the global popu lation live. South Africa, where 4·4% of Africa's population live, accounted for 36·7% of COVID-19 cases and 42·3% of COVID-19 deaths recorded on the continent. Factors contributing to inadequate characterisation of the burden of COVID-19 in Africa include insufficient diagnostic capabilities for testing of SARS-CoV-2, with South Africa reporting the highest testing rate (248 per 1000 population), which nevertheless is one-fourteenth of testing done in the UK. 1 Furthermore, inadequacy in surveillance infrastructure is major challenge in most African countries, limiting the availability of robust readily accessible data. Consequently, many uncertainties about SARS-CoV-2 epidemiology remain in Africa.

The analysis of people admitted to hospital with COVID-19 by Waasila Jassat and colleagues in The Lancet HIV addresses key knowledge gaps in the epidemiology of COVID-19 in an African setting. 2 In South Africa, 9·1% of the 59·6 million population are older than 60 years, and there is a high prevalence of underlying medical conditions among adults, including HIV (19%), diabetes (8-13%), hypertension (44-46%), and obesity (11-41%). 3 Impressively, the analyses by Jassat and colleagues were based on expeditiously establishing a national electronic surveillance database (DATCOV), soon after the COVID-19 pandemic was declared in March, 2020. DATCOV eventually encompassed data on COVID-19 hospital admissions and outcomes from all hospitals in South Africa. A question probed was the role of underlying HIV and past or present tuberculosis in COVID-19 deaths during the first two waves of COVID-19 in South Africa.

The findings indicated that underlying HIV and past and present tuberculosis were independently associated with 1·34-fold and 1·48-fold higher odds of death, respectively, following hospital admission with COVID-19. Among people living with HIV, those not receiving antiretroviral therapy (ART), or with a history of a HIV viral load of 1000 copies or more per mL or CD4 counts of less than 200 cells per µL in the past year had a higher odds of in-hospital COVID-19-associated death. Nevertheless, the prevalence of HIV among patients admitted to hospital with COVID-19 even when restricting analysis to only public hospitals (20·4%), was similar to the population prevalence of HIV (19%). This finding suggests that, similar to population-based observational studies from the UK and a few other places, underlying HIV is not necessarily a risk factor for COVID-19 hospitalisation or severe disease. 4 Missing data on key covariates were a recognised limitation of the DATCOV dataset. Nevertheless, a high prevalence of mainly non-communicable underlying medical conditions was evident in people with HIV (57·5%) and HIV-uninfected individuals (65·2%), the presence of which was independently associated with increased odds of death. Notably, for individuals admitted to hospital with available data, the absolute numbers of COVID-19 deaths associated with underlying diabetes (n=14 707) and hypertension (n=19 668), which were independently associated with higher risk of COVID-19 death, far exceeded deaths in people with HIV (n=3407) or past and present tuberculosis (n=307). Furthermore, older age group was independently associ ated with greater likelihood of COVID-19 death, including 5·32 higher odds among people aged 40-59 years (16 285 deaths) and 20·67 higher odds among people aged 80 years or older (6015 deaths) compared with people aged younger than 20 years. These data are consistent with the global experience, indicating that even in an African setting with high HIV and tuberculosis prevalence, advanced age and non-communicable disease are the major risk factors for COVID-19 deaths. The data provide guidance on which individuals need to be prioritised for COVID-19 vaccination in settings such as South Africa, where access to COVID-19 vaccines remains constrained despite more than half of adult populations in high-income countries already having been vaccinated. 1 Despite the laudable efforts by Jassat and colleagues, another major issue relates to extrapolating the findings of this analysis to the general population. Modelling for excess mortality from May 3, 2020, to March 27, 2021, by the South African Medical Research Council reported 150 271 excess deaths in individuals older than 1 year (including compared with 52 648 [35·0%] COVID-19 deaths officially recorded in South Africa), which was strikingly similar to the 51 037 deaths reported by Jassat and colleagues at the same timepoint. 5 Notably, the trajectory of excess deaths reported in South Africa during the first and second COVID-19 waves is almost completely synchronous with the reported COVID-19 deaths, indicating the majority of excess deaths are probably due to COVID-19. Reasons for the discrepancy between COVID-19 deaths imputed with the excess mortality estimate and recorded deaths could include deficiencies in investigation of possible cases and reporting of COVID-19 deaths. Furthermore, individuals might choose not to access or be unable to gain timely access to health care, particularly when healthcare facilities are overwhelmed at the time when the COVID-19 waves peak. All indications are that the discrepancy between reported and imputed COVID-19 deaths using excess mortality estimates are widening in the current wave being experienced in South Africa largely because of infections by the delta variant, compared with the earlier two waves. This poses further challenges in fully characterising the epidemiology of COVID-19 even in South Africa, despite the admirable efforts by its scientists. SAM has been the lead investigator on COVID-19 epidemiology and vaccine trials that are funded by BMGF and Novavax. JN is a lead South African investigator on COVID-19 treatment trials undertaken by WHO Solidarity accelerated COVID-19 treatment trials. All funding goes to their institutions.. 

Coronavirus pandemic (COVID-19)

Risk factors for COVID-19-related in-hospital mortality in a high HIV and tuberculosis prevalence setting in South Africa: a cohort study

South African Medical Research Council, and ICF. South Africa Demographic and Health Survey

HIV infection and COVID-19 death: a population-based cohort analysis of UK primary care data and linked national death registrations within the OpenSAFELY platform

Report on weekly deaths in South Africa