Testing and Isolation Efficacy: Insights from a Simple Epidemic Model (preprint)

Testing individuals for pathogens can affect the spread of epidemics. Understanding how individual-level processes of sampling and reporting test results can affect community- or population-level spread is a dynamical modeling question. The effect of testing processes on epidemic dynamics depends on factors underlying implementation, particularly testing intensity and on whom testing is focused. Here, we use a simple model to explore how the individual-level effects of testing might directly impact population-level spread. Our model development was motivated by the COVID-19 epidemic, but has generic epidemiological and testing structures. To the classic SIR framework we have added a per capita testing intensity, and compartment-specific testing weights, which can be adjusted to reflect different testing emphases -- surveillance, diagnosis, or control. We derive an analytic expression for the relative reduction in the basic reproductive number due to testing, test-reporting and related isolation behaviours. Intensive testing and fast test reporting are expected to be beneficial at the community level because they can provide a rapid assessment of the situation, identify hot spots, and may enable rapid contact-tracing. Direct effects of fast testing at the individual level are less clear, and may depend on how individuals' behaviour is affected by testing information. Our simple model shows that under some circumstances both increased testing intensity and faster test reporting can reduce the effectiveness of control, and allows us to explore the conditions under which this occurs. Conversely, we find that focusing testing on infected individuals always acts to increase effectiveness of control.

Roles of generation-interval distributions in shaping relative epidemic strength, speed, and control of new SARS-CoV-2 variants (preprint)

Inferring the relative strength (i.e., the ratio of reproduction numbers, Rvar/Rwt) and relative speed (i.e., the difference between growth rates, rvar-rwt) of new SARS-CoV-2 variants compared to their wild types is critical to predicting and controlling the course of the current pandemic. Multiple studies have estimated the relative strength of new variants from the observed relative speed, but they typically neglect the possibility that the new variants have different generation intervals (i.e., time between infection and transmission), which determines the relationship between relative strength and speed. Notably, the increasingly predominant B.1.1.7 variant may have a longer infectious period (and therefore, a longer generation interval) than prior dominant lineages. Here, we explore how differences in generation intervals between a new variant and the wild type affect the relationship between relative strength and speed. We use simulations to show how neglecting these differences can lead to biases in estimates of relative strength in practice and to illustrate how such biases can be assessed. Finally, we discuss implications for control: if new variants have longer generation intervals then speed-like interventions such as contact tracing become more effective, whereas strength-like interventions such as social distancing become less effective.

Age-dependence of healthcare interventions for SARS-CoV-2 infection in Ontario, Canada (preprint)

BackgroundPatient age is the most salient clinical indicator of risk from COVID-19. Age-specific distributions of known SARS-CoV-2 infections and COVID-19-related deaths are available for most countries. However, relatively little attention has been given to the age distributions of hospitalizations and serious healthcare interventions administered to COVID-19 patients. We examined these distributions in Ontario, Canada, in order to quantify the age-related impacts of COVID-19, and to identify potential risks should the healthcare system become overwhelmed with COVID-19 patients in the future. MethodsWe analysed known SARS-CoV-2 infection records from the integrated Public Health Information System (iPHIS) and the Toronto Public Health Coronavirus Rapid Entry System (CORES) between 23 January 2020 and 17 June 2020 (N = 30,546), and estimated the age distributions of hospitalizations, ICU admissions, intubations, and ventilations. We quantified the probability of hospitalization given known SARS-CoV-2 infection, and of survival given COVID-19-related hospitalization. ResultsThe distribution of COVID-19-related hospitalizations peaks with a wide plateau covering ages 54-90, whereas deaths are sharply concentrated in very old ages, with a maximum at age 90. The estimated probability of hospitalization given known SARS-CoV-2 infection reaches a maximum of 32.0% at age 75 (95% CI 27.5%-36.7%). The probability of survival given COVID-19-related hospitalization is uncertain for children (due to small sample size), and near 100% for adults younger than 40. After age 40, survival of hospitalized COVID-19 patients declines substantially; for example, a hospitalized 50-year-old patient has a 90.4% chance of surviving COVID-19 (95% CI 81.9%-95.7%). InterpretationConcerted efforts to control the spread of SARS-CoV-2 have kept prevalence of the virus low in the population of Ontario. The healthcare system has not been overstretched, yet the probability of survival given hospitalization for COVID-19 has been lower than is generally recognized for patients over 40. If prevalence of the virus were to increase and healthcare capacities were to be exceeded, survival of individuals in the broad age range requiring acute care would be expected to decrease, potentially expanding the distribution of COVID-19-related deaths toward younger ages.

Reconciling early-outbreak preliminary estimates of the basic reproductive number and its uncertainty: a new framework and applications to the novel coronavirus (2019-nCoV) outbreak (preprint)

A novel coronavirus (SARS-CoV-2) has recently emerged as a global threat. As the epidemic progresses, many disease modelers have focused on estimating the basic reproductive number Ro -- the average number of secondary cases caused by a primary case in an otherwise susceptible population. The modeling approaches and resulting estimates of Ro vary widely, despite relying on similar data sources. Here, we present a novel statistical framework for comparing and combining different estimates of Ro across a wide range of models by decomposing the basic reproductive number into three key quantities: the exponential growth rate $r$, the mean generation interval $\bar G$, and the generation-interval dispersion $\kappa$. We then apply our framework to early estimates of Ro for the SARS-CoV-2 outbreak. We show that many early Ro estimates are overly confident. Our results emphasize the importance of propagating uncertainties in all components of Ro, including the shape of the generation-interval distribution, in efforts to estimate Ro at the outset of an epidemic.