Estimating actual COVID-19 case numbers using cumulative death count-A method of measuring effectiveness of lockdown of non-essential activities : a South African case study

Abstract

INTRODUCTION : estimating the number of SARS-CoV-2 infected individuals at any specific time point is always a challenge due to asymptomatic cases, the incubation period and testing delays. Here we use an empirical analysis of cumulative death count, transmission-to-death time lag, and infection fatality rate (IFR) to evaluate and estimate the actual cases at a specific time point as a strategy of tracking the spread of COVID-19. METHODS : this method mainly uses death count, as COVID-19 related deaths are arguably more reliably reported than infection case numbers. Using an IFR estimate of 0.66%, we back-calculate the number of cases that would result in the cumulative number of deaths at a given time point in South Africa between 27 February and 14 April. We added the mean incubation period (6.4 days) and the onset-to-death time lag (17.8 days) to identify the estimated time lag between transmission and death (25 days, rounded up). We use the statistical programming language R to analyze the data and produce plots. RESULTS : we estimate 28,182 cases as of 14 April, compared with 3,465 reported cases. Weekly growth rate of actual cases dropped immediately after lockdown implementation and has remained steady, measuring at 51.2% as of 14 April. The timing of drop in growth rate suggests that South Africa’s infection prevention strategy may have been effective at reducing viral transmission. CONCLUSION : estimating the actual number of cases at a specific time point can support evidence-based policies to reduce and prevent the spread of COVID-19. Non-reported, asymptomatic, hard to reach and, mild cases are possible sources of outbreaks that could emerge after lockdown. Therefore, close monitoring, optimized screening strategy and prompt response to COVID-19 could help in stopping the spread of the virus.