COVID-19 Experience in the Philippines: Response, Surveillance and Monitoring Using the FASSSTER Platform (Disaster Risk Reduction)

This book provides an overview of the extensive work that has been done on the design and implementation of the COVID-19 Philippines Local Government Unit Monitoring Platform, more commonly known as Feasibility Analysis of Syndromic Surveillance Using Spatio-Temporal Epidemiological Modeler for Early Detection of Diseases (FASSSTER). The project began in 2016 as a pilot study in developing a multidimensional approach in disease modeling requiring the development of an interoperable platform to accommodate input of data from various sources including electronic medical records, various disease surveillance systems, social media, online news, and weather data. In 2020, the FASSSTER platform was reconfigured for use in the COVID-19 pandemic. Using lessons learned from the previous design and implementation of the platform toward its full adoption by the Department of Health of the Philippines, this book narrates the story of FASSSTER in two main parts.Part I provides a historical perspective of the FASSSTER platform as a modeling and disease surveillance system for dengue, measles and typhoid, followed by the origins of the FASSSTER framework and how it was reconfigured for the management of COVID-19 information for the Philippines. Part I also explains the different technologies and system components of FASSSTER that paved the way to the operationalization of the FASSSTER model and allowed for seamless rendering of projections and analytics. Part II describes the FASSSTER analytics and models including the Susceptible-Exposed-Infected-Recovered (SEIR) model, the model for time-varying reproduction number, spatiotemporal models and contact tracing models, which became the basis for the imposition of restrictions in mobility translated into localized lockdowns. Foreword by Dr. Jaime C. Montoya Foreword by Dr. Alethea R. De Guzman Foreword by Dr. Selva Ramachandran Preface Acknowledgements Contents Contributors Part I COVID-19 Disease Surveillance System in the Philippines 1 Origins of FASSSTER 1.1 Building Blocks of Syndromic Surveillance 1.1.1 The FASSSTER Framework 1.1.2 Spatio-Temporal Epidemiological Modeler (STEM) 1.1.3 Localizing Parameter Estimation 1.2 Use of Electronic Medical Records for Symptoms Monitoring and Reporting 1.3 Infodemiology: Use of Crowdsourced Health-Related Information 1.3.1 Extracting Data from Tweets 1.4 Spatio-Temporal Visualization of Health and Syndromic Surveillance Data 1.5 Insights on the Feasibility Analysis of an Online Syndromic Surveillance Platform Using Spatio-Temporal Visualization References 2 Management of COVID-19 Data for the FASSSTER Platform 2.1 Data Sources 2.1.1 The COVID-19 Line List 2.1.2 DOH Data Collect 2.1.3 Mobility Data 2.2 Data Cleaning 2.2.1 Data Cleaning Process for COVID-19 Line List 2.2.2 Data Imputation 2.3 Utilization of Linelist Data for Analytics and Modeling 2.3.1 Case Statistics 2.3.2 Epidemic Curves 2.3.3 Growth Rate 2.3.4 Risk Classification 2.3.5 Barangay Hotspots 2.3.6 Deaths over Time 2.4 Daily Hospital Report 2.4.1 Health Facility Bed Capacity Utilization Rates (Regional, Provincial/HUC/ICC) 2.4.2 Province/HUC/ICC Risk Classification 2.4.3 Testing Aggregates 3 FASSSTER Data Pipeline and DevOps 3.1 Description of Data Sources for the COVID-19 FASSSTER Platform 3.2 Preprocessing 3.2.1 Producing the Compartmental Model (SEIR) 3.2.2 Producing EpiNow and EpiNow2 3.2.3 Producing the Spatio-Temporal Model 3.3 Interoperability Between R Models and Python Scripts 3.4 Data Visualization Part II FASSSTER Analytics and Models 4 Disease Surveillance Metrics and Statistics 4.1 Risk Classifications and Community Quarantine Protocols 4.2 Phase 1: LGU Epidemic Response Framework 4.2.1 Case Doubling Time and Mortality Doubling Time 4.2.2 Critical Care Utilization Rate 4.3 Phase 2: Average Daily Attack Rate, Two-Week Growth Rate, and Healthcare Utilization Rate 4.3.1 Average Daily Attack Rate 4.3.2 Two-Week Growth Rate 4.3.3 Healthcare Utilization Rate 4.4 Other Metrics 4.4.1 Social Risk Rating and Classification 4.4.2 Economic Risk Classification 4.4.3 Security Risk Classification 4.5 Other Surveillance Visualizations 4.5.1 Epidemic Curve by Location 4.5.2 COVID-19 Positivity Rate 4.5.3 7-Day Moving Average and Growth Factor 4.5.4 Number and Percentage of Barangays with New Cases in Last 14 Days References 5 Effective Reproduction Number Rt 5.1 The COVID-19 Pandemic and the Reproduction Number 5.2 Basic Versus Effective Reproduction Number 5.3 Computing Rt in FASSSTER 5.3.1 EpiEstim 5.3.2 EpiNow2 5.4 Conclusion References 6 The FASSSTER SEIR Model 6.1 The SEIR Equations 6.2 The Basic Reproduction Number 6.3 Model Implementation 6.3.1 Model Parameters from Data 6.3.2 Model Parameters from Literature 6.3.3 Model Fitting to Data 6.3.4 Historical Changes in the Model Parametrization 6.4 Applications of the SEIR Model 6.4.1 Community Quarantine Policies and PDITR 6.4.2 Healthcare Costs and Economic Losses 6.4.3 Healthcare Requirements 6.5 Conclusion References 7 Geospatial and Spatio-Temporal Models 7.1 Introduction 7.2 Hot Spots and Attack Rate 7.3 Local Indicator of Spatial Autocorrelation (LISA) 7.3.1 Motivation and Relevance 7.3.2 Method 7.3.3 Sample Results and Interpretation 7.4 Bayesian Modeling of Spatio-Temporal Risk 7.4.1 Motivation and Relevance 7.4.2 Model Design 7.4.3 Method 7.4.4 Model Selection 7.4.5 Sample Results and Interpretation 7.5 Concluding Remarks References Appendix FASSSTER Data Dictionary