A demographic adjustment to improve measurement of COVID-19 severity at the developing stage of the pandemic

Abstract

The need for accurate statistics has never been felt so deeply as the novel COVID-19 pathogen spreads around the world and quantifying its severity is a primary clinical and public health issue. In Italy, the magnitude and increasing trend of the case-fatality risk (CFR) is fueling the already high levels of public alarm. In this paper, we highlight that the widely used crude CFR is an inaccurate measure of the disease severity since the pandemic is still unfolding. With the goal to improve its comparability over time and across countries at this stage, we then propose a demographic adjustment of the CFR that addresses the bias arising from differential case ascertainment by age. When applied to publicly released data for Italy, we show that until March 16 our adjusted CFR was similar to that of Wuhan -the most affected Chinese region, where COVID-19 has now been contained. This indicates that our adjusted CFR improves its comparability over time, making an important tool to chart the course of the COVID-19 pandemic across countries. Since March 16, the Italian COVID-19 outbreak has entered a new phase, with the northern and southern regions following different trajectories. As a result, our adjusted CFR has been increasing between March 16 and March 20. Data at the subnational level are needed to correctly assess the disease severity in the country at this stage. SOMMARIOIl diffondersi dell'epidemia COVID-19 ha dimostrato l'importanza di un'informazione corretta, laddove la mancanza di una risposta tempestiva si valuta essenzialmente in termini di vite umane. In Italia, la propagazione del COVID-19 è stata accompagnata da un'informazione imprecisa circa la sua reale entità e gravità. In questo contributo, dimostriamo come il tasso di fatalità, attualmente l'indicatore più utilizzato della severità del virus, sia inaccurato a questo stadio della pandemia e proponiamo un metodo demografico per migliorarne la stima, in particolare durante la fase ascendente della curva epidemica. L'applicazione del nostro metodo ai dati nazionali italiani dimostra che, fino al 16 di Marzo, il tasso di fatalità rimane assimilabile a quello osservato nella regione di Wuhan, ma comincia a conoscere un drastico incremento negli ultimi giorni. I nostri risultati indicano che il metodo proposto migliora la comparabilità del tasso di fatalità nel tempo e tra paesi, dimostrando la sua utilità come strumento per tracciare e comprendere l'evoluzione pandemica attuale.

Monitoring trends and differences in COVID-19 case-fatality rates using decomposition methods: Contributions of age structure and age-specific fatality

“…On April 13, Italy had roughly the same number of cases as Germany on April 28, and a CFR of 12.9%. Thus, the outbreak in Italy is going along with a much higher CFR, which has also increased over time [2,3]. Also shown in Fig 1 are trends for Spain (until May 21) and New York City (until June 30), which fall somewhere between Germany and Italy.…”

Monitoring trends and differences in COVID-19 case-fatality rates using decomposition methods: Contributions of age structure and age-specific fatality

“…On March 16, Italy had roughly the same number of cases as Germany on March 24, and a CFR of 7.7%. Thus, the outbreak in Italy is going along with a much higher CFR, and the CFR increased over time [2,3]. Differences in the CFR could indicate that the risk of dying of COVID-19 among detected cases differs between countries or changes within a population over time.…”

Estimating COVID-19 pandemic severity in Indian context

“…The outcome-based and 14-day lag estimates of CFR were, respectively, ~7 and ~8 times higher than cCFR estimate [2.32 (2.05-2.59)]. In China, the age adjusted CFR (ACFR) was 3.8% as of February 22, 2020, and in Italy, the ACFR ranged 4.7%-7% as of March 20, 2020 [21], both of which were lower than the ACFR value of 15% (1.58-48.42) in India, as estimated in the present study. The outcome-based CFR estimate was applied for SARS by Ghani et al [22], and that was independent of any window period providing point-estimate during the course of epidemic.…”