ABSTRACT
INTRODUCTION: the level five (L5) lockdown was a very stringent social distancing measure taken to reduce the spread of COVID-19 infections. This study assessed the impact of the L5 lockdown and its association with the incidence of COVID-19 cases in South Africa (SA). METHODS: data was obtained from the National Department of Health (NDoH) from the 5th March to the 30th April 2020. A basic reproductive number (R0) and a serial interval were used to calculate estimated cases (EC). A double exponential smoothing model was used to forecast the number of cases during the L5 lockdown period. A Poisson regression model was fitted to describe the association between L5 lockdown status and incident cases. RESULTS: a total of 5,737 laboratory-confirmed cases (LCC) were reported by 30th April 2020, 4,785 (83%) occurred during L5 lockdown. Our model forecasted 30,629 cases of COVID-19 assuming L5 lockdown was not imposed. High incidence rates of COVID-19 were recorded in KwaZulu-Natal and Mpumalanga Provinces during the L5 lockdown compared to the other provinces. Nationally, the incident rate of COVID-19 was 68.00% higher in L5 lockdown than pre-lockdown for LCC. CONCLUSION: the L5 lockdown was very effective in reducing the incidence of COVID-19 cases. However, the incident rates of LCC and EC were higher nationally, and in some provinces during the L5 lockdown.