Lack of practical identifiability may hamper reliable predictions in COVID-19 epidemic models

TL;DR

This paper presents a framework to assess how data uncertainty affects parameter estimation and unmeasured variables in COVID-19 models, highlighting potential issues with model reliability and forecasting accuracy.

Contribution

It introduces a novel method to quantify identifiability regimes and demonstrates how lack of practical identifiability can hinder reliable epidemic predictions.

Findings

Framework characterizes different identifiability regimes

Lack of identifiability can prevent reliable COVID-19 forecasts

Method applicable to models with few compartments

Abstract

Compartmental models are widely adopted to describe and predict the spreading of infectious diseases. The unknown parameters of such models need to be estimated from the data. Furthermore, when some of the model variables are not empirically accessible, as in the case of asymptomatic carriers of COVID-19, they have to be obtained as an outcome of the model. Here, we introduce a framework to quantify how the uncertainty in the data impacts the determination of the parameters and the evolution of the unmeasured variables of a given model. We illustrate how the method is able to characterize different regimes of identifiability, even in models with few compartments. Finally, we discuss how the lack of identifiability in a realistic model for COVID-19 may prevent reliable forecasting of the epidemic dynamics.