A single autoregressive integrated moving average with exogenous variable outperforms ensembles of autoregressive models for forecasting influenza hospitalizations in the contiguous United States
Year: 2026
Type: article
Abstract: Infectious disease forecasting is important to public health decision-making, particularly for mitigating the burden of seasonal influenza. We propose and evaluate comprehensive autoregressive modelling approaches (autoregressive integrated moving average (ARIMA) and autoregressive integrated moving average with exogenous variables (ARIMAX)) for short-term forecasts of influenza-related hospitalizations across the contiguous United States (US). We used data from the National Healthcare Safety Network to forecast the influenza seasons of the years 2022 to 2024 in 48 states. We compare automatically tuned models (AUTO) and ensembles of different sizes (ES27 or ES64). Base models are ARIMA models or ARIMAX models using forecasting covariates with a one-week lag of the forecast day, including: mean temperature, mean hospitalizations by epidemiological weeks, mean hospitalizations in adjacent states and the average of hospitalizations across the contiguous US. We also investigated the effect of forecasting based on log-transformed hospitalizations data and back-transforming the results (LB). Overall, 29 approaches were compared to the AUTO ARIMA model without logarithmic transformation and the FluSight baseline using weighted interval scores. Our results indicate that using the LB approach, which forecasts based on epidemic growth rate and back-transform the results, most increased model performance. The average of hospitalizations across the contiguous US (AVG) forecasting feature was the most useful covariate. Ensemble approaches typically did not improve performance. The best model was the AUTO ARIMAX using the AVG covariate with log-back transformation.
Institutions: University of Georgia, University of Oxford
Language: English
Subfield: Epidemiology
Field: Medicine
Domain: Health Sciences
Open Access status: hybrid
Awards: G1071703045, G3773782423, G3828582264, G4542157032, GENOMIC CENTER OF EXCELLENCE-PUBLIC HEALTH PATHOGENS: TRANSLATION, EDUCATION, RESPONSE BASE, RESPONSE IMPLEMENTATION, LEAD PGCOE EDUCATION - GEORGIA DEPARTMENT OF PUBLIC HEALTH US PUBLIC HEALTH PATHOGENS GENOMICS CENTERS OF EXCELLENCE PURPOSE: THIS PROPOSAL SUPPORTS IMPLEMENTATION OF THE CENTER FOR APPLIED PATHOGEN EPIDEMIOLOGY AND OUTBREAK CONTROL (CAPE) A CENTER OF EXCELLENCE THAT HAS THREE MAIN COMPONENTS, TRANSLATION OF INNOVATIVE RESEARCH TO PUBLIC HEALTH; EDUCATION OF THE PUBLIC HEALTH WORKFORCE; AND PREPARATION AND RESPONSE TO INFECTIOUS DISEASE THREATS. AS A CENTER OF EXCELLENCE, CAPE WILL ADDRESS BOTH NEAR AND LONG-TERM GENOMIC CHALLENGES THROUGH BROADLY APPLICABLE, INNOVATIVE SOLUTIONS FOR THE PUBLIC HEALTH COMMUNITY. OUR INNOVATIVE TEAMING STRATEGY ALLOWS FOR A RESPONSIVE NETWORK THAT CAN DEVELOP CUTTING EDGE SOLUTIONS AND BEST PRACTICES FOR USING PATHOGEN SEQUENCE DATA TO INFORM PUBLIC HEALTH RESPONSE AND POLICY. THIS APPLICATION IS FOR THE MANDATORY COMPONENTS (CORE, RESPONSE IMPLEMENTATION) AND INCLUDES THE OPTIONAL COMPONENT (LEAD PGCOE FOR EDUCATION). STRATEGY: THE GEORGIA DEPARTMENT OF PUBLIC HEALTH (GDPH) IN COLLABORATION WITH SIX MAJOR ACADEMIC INSTITUTIONS (AUGUSTA UNIVERSITY, EMORY UNIVERSITY, GEORGIA INSTITUTE OF TECHNOLOGY, UNIVERSITY OF GEORGIA, GEORGIA STATE UNIVERSITY, AND THE UNIVERSITY OF TEXAS, WILL JOINTLY SUPPORT CAPE. COORDINATING WITH THE CENTERS FOR DISEASE CONTROL AND PREVENTION AND THE OTHER CENTERS OF EXCELLENCE PARTNERS, THE CAPE PUBLIC HEALTH TEAM WILL CONDUCT A PUBLIC HEALTH NEEDS ASSESSMENT TO GUIDE CENTER INITIATIVES. THE ASSESSMENT WILL PROVIDE THE FOUNDATION FOR SUBSEQUENT PLANNING AND PRIORITIZATION FOR THE NETWORK. THIS APPROACH REPRESENTS THE WAY WE WILL ADDRESS SOLUTION DEVELOPMENT, OPERATIONS, AND IMPLEMENTATION. THE CAPE PRINCIPAL AND CO-PRINCIPAL INVESTIGATORS WILL LEVERAGE THE STRENGTH AND EXPERTISE OF THE OTHER CAPE NETWORK MEMBERS TO ADDRESS THE CORE ACTIVITIES. CORE 1.1 FOCUSES ON DEVELOPING NEW SEQUENCING APPROACHES WITH STREAMLINED METHODOLOGIES THAT ARE PATHOGEN AGNOSTIC. CORE 1.2 FOCUSES ON BIOINFORMATIC SOLUTIONS THAT CAN IMMEDIATELY SUPPORT PUBLIC HEALTH DEPARTMENT REPORTING NEEDS. CORE 1.3 FOCUSES ON DEVELOPMENT OF AN INFORMATIC INFRASTRUCTURE THAT CAN LINK MULTIPLE DATA STREAMS, INTEGRATE BIOINFORMATIC SOLUTIONS WHILE PRIORITIZING DATA SECURITY AND INTEGRITY. PRIOR TO DEVELOPING THIS PROPOSAL, AS A SOLUTIONING TEAM, WE SURVEYED EXISTING WORK TO PROPOSE ACTIVITIES IN CORE 1.4 THAT COULD ADDRESS PRIORITY RESEARCH AREAS AND CHALLENGES ENCOUNTERED IN THE SARS-COV-2 RESPONSE. CORE 1.5 EXPANDS ON CURRENT SURVEILLANCE ACTIVITIES BY EXPLOITING WASTEWATER / ENVIRONMENTAL SAMPLING AND DEVELOPING PROTOCOLS FOR DATA AND SAMPLE SHARING TO INTEGRATE HOSPITAL-BASED SURVEILLANCE AND THE CLINICAL PATHOLOGY NETWORK INTO THE OVERALL PUBLIC HEALTH SURVEILLANCE NETWORK. CORE 1.6 DEVELOPS TRAINING MATERIALS, INCLUDING ONLINE ON-DEMAND MATERIALS FOR WORKFORCE DEVELOPMENT, INTERNSHIPS FOR APPLIED PUBLIC HEALTH PRACTICE EXPERIENCE AND TRANSLATION FROM CAPE TO OUR PUBLIC HEALTH PARTNERS AND WORKSHOPS ON MOLECULAR EPIDEMIOLOGY. CORE 1.7 PROPOSES SPECIFIC ACTIVITIES FOR RESPONSE IMPLEMENTATION. ACTIVITIES WILL INCLUDE ENSURING ALL ADMINISTRATIVE AGREEMENTS ARE IN PLACE TO ENSURE SEAMLESS TRANSFER OF DATA, REAGENTS OR SAMPLE MATERIALS IS POSSIBLE BETWEEN PARTICIPATING LABS. IN CORE 2, CAPE WILL LEVERAGE OUR EXPERTISE TO COORDINATE RESPONSE ACTIVITIES, INCLUDING ANALYTICAL AND MODELING CAPABILITIES. OUTCOME: CAPE WILL WORK WITH CDC AND OTHER CENTERS OF EXCELLENCE, ACADEMIC CENTERS, CLINICAL SITES, AND PUBLIC HEALTH LABORATORIES TO SUPPORT INNOVATIONS IN SEQUENCING, BIOINFORMATICS, GENOMIC SURVEILLANCE, DATA INTEGRATION AND VISUALIZATION, AND WORKFORCE DEVELOPMENT, WITH THE OBJECTIVE OF STRENGTHENING THE CAPACITY TO RESPOND TO PUBLIC HEALTH THREATS AND ADVANCE DAY TO DAY OPERATIONS.
OpenAlex ID: W7154409372