Modeling COVID-19 Latent Prevalence to Assess a Public Health Intervention at a State and Regional Scale: Retrospective Cohort Study

A regionally tailored epidemiological forecast and monitoring program to guide a healthcare system in the COVID-19 pandemic

BACKGROUND

During the COVID-19 pandemic, analytics and predictive models built on regional data provided timely, accurate monitoring of epidemiological behavior, informing critical planning and decision-making for health system leaders. At Atrium Health, a large, integrated healthcare system in the southeastern United States, a team of statisticians and physicians created a comprehensive forecast and monitoring program that leveraged an array of statistical methods.

METHODS

The program utilized the following methodological approaches: (i) exploratory graphics, including time plots of epidemiological metrics with smoothers; (ii) infection prevalence forecasting using a Bayesian epidemiological model with time-varying infection rate; (iii) doubling and halving times computed using changepoints in local linear trend; (iv) death monitoring using combination forecasting with an ensemble of models; (v) effective reproduction number estimation with a Bayesian approach; (vi) COVID-19 patients hospital census monitored via time series models; and (vii) quantified forecast performance.

RESULTS

A consolidated forecast and monitoring report was produced weekly and proved to be an effective, vital source of information and guidance as the healthcare system navigated the inherent uncertainty of the pandemic. Forecasts provided accurate and precise information that informed critical decisions on resource planning, bed capacity and staffing management, and infection prevention strategies.

CONCLUSIONS

In this paper, we have presented the framework used in our epidemiological forecast and monitoring program at Atrium Health, as well as provided recommendations for implementation by other healthcare systems and institutions to facilitate use in future pandemics.

Dynamically adjustable SVEIR(MH) model of multiwave epidemics: Estimating the effects of public health measures against COVID-19

The COVID-19 pandemic was characterized by multiple subsequent, overlapping outbreaks, as well as extremely rapid changes in viral genomes. The information about local epidemics spread and the epidemic control measures was shared on a daily basis (number of cases and deaths) via centralized repositories. The vaccines were developed within the first year of the pandemic. New modes of monitoring and sharing of epidemic data were implemented using Internet resources. We modified the basic SEIR compartmental model to include public health measures, multiwave scenarios, and the variation of viral infectivity and transmissibility reflected by the basic reproduction number R0 of emerging viral variants. SVEIR(MH) model considers the capacity of the medical system, lockdowns, vaccination, and changes in viral reproduction rate on the epidemic spread. The developed model uses daily infection reports for assessing the epidemic dynamics, and daily changes of mobility data from mobile phone networks to assess the lockdown effectiveness. This model was deployed to six European regions Baden-Württemberg (Germany), Belgium, Czechia, Lombardy (Italy), Sweden, and Switzerland for the first 2 years of the pandemic. The correlation coefficients between observed and reported infection data showed good concordance for both years of the pandemic (ρ = 0.84-0.94 for the raw data and ρ = 0.91-0.98 for smoothed 7-day averages). The results show stability across the regions and the different epidemic waves. Optimal control of epidemic waves can be achieved by dynamically adjusting epidemic control measures in real-time. SVEIR(MH) model can simulate different scenarios and inform adjustments to the public health policies to achieve the target outcomes. Because this model is highly representative of actual epidemic situations, it can be used to assess both the public health and socioeconomic effects of the public health measures within the first 7 days of the outbreak.

Modeling infection from SARS-CoV-2 wastewater concentrations: promise, limitations, and future directions

Estimating total infection levels, including unreported and asymptomatic infections, is important for understanding community disease transmission. Wastewater can provide a pooled community sample to estimate total infections that is independent of case reporting biases toward individuals with moderate to severe symptoms and by test-seeking behavior and access. We derive three mechanistic models for estimating community infection levels from wastewater measurements based on a description of the processes that generate SARS-CoV-2 RNA signals in wastewater and accounting for the fecal strength of wastewater through endogenous microbial markers, daily flow, and per-capita wastewater generation estimates. The models are illustrated through two case studies of wastewater data collected during 2020-2021 in Virginia Beach, VA, and Santa Clara County, CA. Median simulated infection levels generally were higher than reported cases, but at times, were lower, suggesting a discrepancy between the reported cases and wastewater data, or inaccurate modeling results. Daily simulated infection estimates showed large ranges, in part due to dependence on highly variable clinical viral fecal shedding data. Overall, the wastewater-based mechanistic models are useful for normalization of wastewater measurements and for understanding wastewater-based surveillance data for public health decision-making but are currently limited by lack of robust SARS-CoV-2 fecal shedding data.

Establishing an SEIR-based framework for local modelling of COVID-19 infections, hospitalisations and deaths

Without timely assessments of the number of COVID-19 cases requiring hospitalisation, healthcare providers will struggle to ensure an appropriate number of beds are made available. Too few could cause excess deaths while too many could result in additional waits for elective treatment. As well as supporting capacity considerations, reliably projecting future "waves" is important to inform the nature, timing and magnitude of any localised restrictions to reduce transmission. In making the case for locally owned and locally configurable models, this paper details the approach taken by one major healthcare system in founding a multi-disciplinary "Scenario Review Working Group", comprising commissioners, public health officials and academic epidemiologists. The role of this group, which met weekly during the pandemic, was to define and maintain an evolving library of plausible scenarios to underpin projections obtained through an SEIR-based compartmental model. Outputs have informed decision-making at the system's major incident Bronze, Silver and Gold Commands. This paper presents illustrated examples of use and offers practical considerations for other healthcare systems that may benefit from such a framework.

Systematic review of predictive mathematical models of COVID-19 epidemic

BACKGROUND

Various mathematical models were published to predict the epidemiological consequences of the COVID-19 pandemic. This systematic review has studied the initial epidemiological models.

METHODS

Articles published from January to June 2020 were extracted from databases using search strings and those peer-reviewed with full text in English were included in the study. They were analysed as to whether they made definite predictions in terms of time and numbers, or contained only mathematical assumptions and open-ended predictions. Factors such as early vs. late prediction models, long-term vs. curve-fitting models and comparisons based on modelling techniques were analysed in detail.

RESULTS

Among 56,922 hits in 05 databases, screening yielded 434 abstracts, of which 72 articles were included. Predictive models comprised over 70% (51/72) of the articles, with susceptible, exposed, infectious and recovered (SEIR) being the commonest type (mean duration of prediction being 3 months). Common predictions were regarding cumulative cases (44/72, 61.1%), time to reach total numbers (41/72, 56.9%), peak numbers (22/72, 30.5%), time to peak (24/72, 33.3%), hospital utilisation (7/72, 9.7%) and effect of lockdown and NPIs (50/72, 69.4%). The commonest countries for which models were predicted were China followed by USA, South Korea, Japan and India. Models were published by various professionals including Engineers (12.5%), Mathematicians (9.7%), Epidemiologists (11.1%) and Physicians (9.7%) with a third (32.9%) being the result of collaborative efforts between two or more professions.

CONCLUSION

There was a wide diversity in the type of models, duration of prediction and the variable that they predicted, with SEIR model being the commonest type.

Adaptive Susceptible-Infectious-Removed Model for Continuous Estimation of the COVID-19 Infection Rate and Reproduction Number in the United States: Modeling Study

BACKGROUND

The dynamics of the COVID-19 pandemic vary owing to local population density and policy measures. During decision-making, policymakers consider an estimate of the effective reproduction number Rt, which is the expected number of secondary infections spread by a single infected individual.

OBJECTIVE

We propose a simple method for estimating the time-varying infection rate and the Rt.

METHODS

We used a sliding window approach with a Susceptible-Infectious-Removed (SIR) model. We estimated the infection rate from the reported cases over a 7-day window to obtain a continuous estimation of Rt. A proposed adaptive SIR (aSIR) model was applied to analyze the data at the state and county levels.

RESULTS

The aSIR model showed an excellent fit for the number of reported COVID-19 cases, and the 1-day forecast mean absolute prediction error was <2.6% across all states. However, the 7-day forecast mean absolute prediction error approached 16.2% and strongly overestimated the number of cases when the Rt was rapidly decreasing. The maximal Rt displayed a wide range of 2.0 to 4.5 across all states, with the highest values for New York (4.4) and Michigan (4.5). We found that the aSIR model can rapidly adapt to an increase in the number of tests and an associated increase in the reported cases of infection. Our results also suggest that intensive testing may be an effective method of reducing Rt.

CONCLUSIONS

The aSIR model provides a simple and accurate computational tool for continuous Rt estimation and evaluation of the efficacy of mitigation measures.

Short-Range Forecasting of COVID-19 During Early Onset at County, Health District, and State Geographic Levels Using Seven Methods: Comparative Forecasting Study

BACKGROUND

Forecasting methods rely on trends and averages of prior observations to forecast COVID-19 case counts. COVID-19 forecasts have received much media attention, and numerous platforms have been created to inform the public. However, forecasting effectiveness varies by geographic scope and is affected by changing assumptions in behaviors and preventative measures in response to the pandemic. Due to time requirements for developing a COVID-19 vaccine, evidence is needed to inform short-term forecasting method selection at county, health district, and state levels.

OBJECTIVE

COVID-19 forecasts keep the public informed and contribute to public policy. As such, proper understanding of forecasting purposes and outcomes is needed to advance knowledge of health statistics for policy makers and the public. Using publicly available real-time data provided online, we aimed to evaluate the performance of seven forecasting methods utilized to forecast cumulative COVID-19 case counts. Forecasts were evaluated based on how well they forecast 1, 3, and 7 days forward when utilizing 1-, 3-, 7-, or all prior-day cumulative case counts during early virus onset. This study provides an objective evaluation of the forecasting methods to identify forecasting model assumptions that contribute to lower error in forecasting COVID-19 cumulative case growth. This information benefits professionals, decision makers, and the public relying on the data provided by short-term case count estimates at varied geographic levels.

METHODS

We created 1-, 3-, and 7-day forecasts at the county, health district, and state levels using (1) a naïve approach, (2) Holt-Winters (HW) exponential smoothing, (3) a growth rate approach, (4) a moving average (MA) approach, (5) an autoregressive (AR) approach, (6) an autoregressive moving average (ARMA) approach, and (7) an autoregressive integrated moving average (ARIMA) approach. Forecasts relied on Virginia's 3464 historical county-level cumulative case counts from March 7 to April 22, 2020, as reported by The New York Times. Statistically significant results were identified using 95% CIs of median absolute error (MdAE) and median absolute percentage error (MdAPE) metrics of the resulting 216,698 forecasts.

RESULTS

The next-day MA forecast with 3-day look-back length obtained the lowest MdAE (median 0.67, 95% CI 0.49-0.84, P<.001) and statistically significantly differed from 39 out of 59 alternatives (66%) to 53 out of 59 alternatives (90%) at each geographic level at a significance level of .01. For short-range forecasting, methods assuming stationary means of prior days' counts outperformed methods with assumptions of weak stationarity or nonstationarity means. MdAPE results revealed statistically significant differences across geographic levels.

CONCLUSIONS

For short-range COVID-19 cumulative case count forecasting at the county, health district, and state levels during early onset, the following were found: (1) the MA method was effective for forecasting 1-, 3-, and 7-day cumulative case counts; (2) exponential growth was not the best representation of case growth during early virus onset when the public was aware of the virus; and (3) geographic resolution was a factor in the selection of forecasting methods.

Flexible, Freely Available Stochastic Individual Contact Model for Exploring COVID-19 Intervention and Control Strategies: Development and Simulation

BACKGROUND

Throughout March 2020, leaders in countries across the world were making crucial decisions about how and when to implement public health interventions to combat the coronavirus disease (COVID-19). They urgently needed tools to help them to explore what will work best in their specific circumstances of epidemic size and spread, and feasible intervention scenarios.

OBJECTIVE

We sought to rapidly develop a flexible, freely available simulation model for use by modelers and researchers to allow investigation of how various public health interventions implemented at various time points might change the shape of the COVID-19 epidemic curve.

METHODS

"COVOID" (COVID-19 Open-Source Infection Dynamics) is a stochastic individual contact model (ICM), which extends the ICMs provided by the open-source EpiModel package for the R statistical computing environment. To demonstrate its use and inform urgent decisions on March 30, 2020, we modeled similar intervention scenarios to those reported by other investigators using various model types, as well as novel scenarios. The scenarios involved isolation of cases, moderate social distancing, and stricter population "lockdowns" enacted over varying time periods in a hypothetical population of 100,000 people. On April 30, 2020, we simulated the epidemic curve for the three contiguous local areas (population 287,344) in eastern Sydney, Australia that recorded 5.3% of Australian cases of COVID-19 through to April 30, 2020, under five different intervention scenarios and compared the modeled predictions with the observed epidemic curve for these areas.

RESULTS

COVOID allocates each member of a population to one of seven compartments. The number of times individuals in the various compartments interact with each other and their probability of transmitting infection at each interaction can be varied to simulate the effects of interventions. Using COVOID on March 30, 2020, we were able to replicate the epidemic response patterns to specific social distancing intervention scenarios reported by others. The simulated curve for three local areas of Sydney from March 1 to April 30, 2020, was similar to the observed epidemic curve in terms of peak numbers of cases, total numbers of cases, and duration under a scenario representing the public health measures that were actually enacted, including case isolation and ramp-up of testing and social distancing measures.

CONCLUSIONS

COVOID allows rapid modeling of many potential intervention scenarios, can be tailored to diverse settings, and requires only standard computing infrastructure. It replicates the epidemic curves produced by other models that require highly detailed population-level data, and its predicted epidemic curve, using parameters simulating the public health measures that were enacted, was similar in form to that actually observed in Sydney, Australia. Our team and collaborators are currently developing an extended open-source COVOID package comprising of a suite of tools to explore intervention scenarios using several categories of models.