A novel method to monitor COVID-19 fatality rate in real-time, a key metric to guide public health policy

Abstract

An accurate estimator of the real-time fatality rate is warranted to monitor the progress of ongoing epidemics, hence facilitating the policy-making process. However, most of the existing estimators fail to capture the time-varying nature of the fatality rate and are often biased in practice. A simple real-time fatality rate estimator with adjustment for reporting delays is proposed in this paper using the fused lasso technique. This approach is easy to use and can be broadly applied to public health practice as only basic epidemiological data are required. A large-scale simulation study suggests that the proposed estimator is a reliable benchmark for formulating public health policies during an epidemic with high accuracy and sensitivity in capturing the changes in the fatality rate over time, while the other two commonly-used case fatality rate estimators may convey delayed or even misleading signals of the true situation. The application to the COVID-19 data in Germany between January 2020 and January 2022 demonstrates the importance of the social restrictions in the early phase of the pandemic when vaccines were not available, and the beneficial effects of vaccination in suppressing the fatality rate to a low level since August 2021 irrespective of the rebound in infections driven by the more infectious Delta and Omicron variants during the fourth wave.

Estimation of standardized real-time fatality rate for ongoing epidemics

“…A real-time fatality rate estimator taking into account reporting delay in death has been recently proposed by Qu et al [ 6 ], where minimal epidemiological data, namely the observed time series of c t and d t , and the distribution F , are required as inputs. We will briefly mention the existing estimator to facilitate the discussion of its standardized version.…”

“…By setting p * = p 1 = ⋯ = p τ in ( 1 ), the above framework is reduced to that of the time-delay adjusted case fatality rate estimator studied in Nishiura et al [ 18 ], denoted as p * in this article. Given the strong serial correlation in the time series of p t and the fact that typically only the trend of the time series is of interest, Qu et al [ 6 ] proposed to apply the fused lasso technique [ 19 , 20 ] to obtain a smoothed estimator of p , namely the real-time delay-adjusted case fatality rate, henceforth rtaCFR t for t = 1, …, τ . Let d = ( d 1 , …, d τ ) T and be a ( τ − 1) × τ penalty matrix.…”

“…This metric often yields more valuable information that can inform public health policies, including shifts in disease virulence, the efficacy of response strategies, and the quality of healthcare services. To this end, Qu et al [ 6 ] proposed a real-time fatality rate estimator adjusted for reporting delay (rtaCFR) based on the fused lasso technique. The rtaCFR was shown to outperform other fatality rate estimators in tracking disease severity throughout the COVID-19 pandemic in Germany.…”

“…The proposed method corrects the bias of the commonly used CFR by adjusting for the time lags between illness onset and death, and is more sensitive in picking up the changes in fatality rate during an ongoing epidemic. A distinguished feature of the new estimator compared to the existing rtaCFR in [ 6 ] is its ability to remove the effects of some time-varying confounding factors attributed to the changes in disease severity when the relevant information is available. One typical application of the proposed method is to adjust for the impact induced by the changes in the age distribution of confirmed cases when monitoring changes in disease virulence over time.…”

An interactive national digital surveillance system to fight against COVID-19 in Bangladesh

“…Implementing AI techniques like machine learning and deep learning can be crucial for rapid diagnosis and screening ( 29 ). This AI techniques may be beneficial for contact tracking and continuous monitoring of COVID-19 affected individuals ( 30 – 34 ). However, most advanced technology-enabled surveillance systems have been developed and deployed in high-income countries, leaving Bangladesh behind.…”