The impact of non-pharmaceutical interventions on the first COVID-19 epidemic wave in South Africa

  1. Thabo Mabuka
  2. Nesisa Ncube
  3. Michael Ross
  4. Andrea Silaji
  5. Willie Macharia
  6. Tinashe Ndemera
  7. Tlaleng Lemeke

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Objective

In this study, we investigated the impact of COVID-19 NPIs in South Africa to understand their effectiveness in the reduction of transmission of COVID-19 in the South African population. This study also investigated the COVID-19 testing, reporting, hospitalised cases, excess deaths and COVID-19 modelling in the first wave of the COVID-19 epidemic in South Africa.

Methods

A semi-reactive stochastic COVID-19 model, the ARI COVID-19 SEIR model, was used to investigate the impact of NPIs in South Africa to understand their effectiveness in the reduction of COVID-19 transmission in the South African population. COVID-19 testing, reporting, hospitalised cases and excess deaths in the first COVID-19 epidemic wave in South Africa were investigated using regressional analysis and descriptive statistics.

Findings

The general trend in population movement in South African locations shows that the COVID-19 NPIs (National Lockdown Alert Levels 5,4,3,2) were approximately 30% more effective in reducing population movement concerning each increase by 1 Alert Level. The translated reduction in the effective SARS-CoV-2 daily contact number (β) was 6.12% to 36.1% concerning increasing Alert Levels. Due to the implemented NPIs, the effective SARS-CoV-2 daily contact number in the first COVID-19 epidemic wave in South Africa was reduced by 58.1–71.1% while the peak was delayed by 84 days. The estimated COVID-19 reproductive number was between 1.98 to 0.40. During South Africa’s first COVID-19 epidemic wave, the mean COVID-19 admission status in South African hospitals was 58.5%, 95% CI [58.1–59.0] in the general ward, 13.4%, 95% CI [13.1–13.7] in the intensive care unit, 13.3%, 95% CI [12.6–14.0] on oxygen, 6.37%, 95% CI [6.23–6.51] in high care, 6.29%, 95% CI [6.02–6.55] on ventilator and 2.13%, 95% CI [1.87–2.43] in isolation ward respectively. The estimated mean South African COVID-19 patient discharge rate was 11.9 days per patient. While the estimated mean of the South African COVID-19 patient case fatality rate (CFR) in hospital and outside the hospital was 2.06%, 95% CI [1.86–2.25] (deaths per admitted patients) and 2.30%, 95% CI [1.12–3.83](deaths per severe and critical cases) respectively. The relatively high coefficient of variance in COVID-19 model outputs observed in this study shows the uncertainty in the accuracy of the reviewed COVID-19 models in predicting the severity of COVID-19. However, the reviewed COVID-19 models were accurate in predicting the progression of the first COVID-19 epidemic wave in South Africa.

Conclusion

The results from this study show that the COVID-19 NPI policies implemented by the Government of South Africa played a significant role in the reduction of COVID-19 active, hospitalised cases and deaths in South Africa’s first COVID-19 epidemic wave. The results also show the use of COVID-19 modelling to understand the COVID-19 pandemic and the impact of regressor variables in an epidemic.

Article activity feed

  1. Version published to 10.1186/s12889-023-16162-0
  2. Version published to 10.21203/rs.3.rs-1795276/v1 on Research Square
  3. ScreenIT

    SciScore for 10.1101/2021.06.29.21259625: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    NIH rigor criteria are not applicable to paper type.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    Infected with Mild and Moderate Symptoms (Ib): Individuals within the population of the model with mild and moderate symptoms who are infectious.
    Mild
    suggested: (MILD, RRID:SCR_003335)
    To allow the goodness of fit of Modelled data to Active Case or Reported COVID-19 Deaths Data, the Average Normalised Error of all Data points used in the Regression Analysis was reduced to 0 by changing the Effective Daily Contact Number (β) using the What-If Analysis Function in MS Excel.
    MS Excel
    suggested: None

    Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    The South African National Lockdown Level Alert 5 as shown in Table 4 was predominantly movement restrictions and limitation of services to essential services. Under this level the South African borders and air space were closed, there was an enforcement of strict non-movement of non-essential personal and a ban on some of the industries such as the alcohol and tobacco industry. To enforce this the South African Defence Force was deployed to oversee the compliance of this measure (Government of South Africa, 2020b). A screening and testing program for COVID-19 was initiated under the South African Department of Health. The initial testing was conducted by the National Institute of Communicable Disease (NICD) and this was expanded to a larger network of private and National Health Laboratory Services (NHLS) (NICD, 2020b). Mobile testing units were also deployed particularly to the hardest-hit provinces of Gauteng, Western Cape and Kwa-Zulu Natal (Government of South Africa, 2020c). The South African economy like in many other countries with similar COVID-19 measures was negatively affected due to limitations in business and trade (Arndt et al., 2020). South Africa gradually eased restrictions to the National Lockdown Level Alerts 4,3,2 and 1 by permitting businesses to trade, easing curfews, gathering capacity and movement restrictions as shown by the difference in the Alert Level policy summaries provided in Table 4. On the 17th of August 2020, the national lockdown alert lev...

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: We did not find any issues relating to colormaps.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    Results from scite Reference Check: We found no unreliable references.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  4. Version published to 10.1101/2021.06.29.21259625v1 on medRxiv