Development of an Influenza-like Illness Detection System Using Google Search Queries in Chiang Mai Province

Abstract

Thailand’s influenza surveillance waits for patients to present at the hospital, producing delayed and information that did not cover the masses. This research-and-development study developed and assessed a prototype to detect influenza-like illness (ILI) using Google Relative Search Value (RSV) in Chiang Mai. Participants are 15 Situation Awareness Team (SAT) members. The study conducted during Sep 2023–Jul 2024 and comprised of phase one: situation analysis through structured interviews, phase two: Agile system development, and phase 3: evaluation with a validated five-point questionnaire. We also analyzed relationships between three keywords’ RSV, “fever,” “cough,” “sore throat”, and Chiang Mai’s 2016-2023 influenza cases using a ±5-week lead–lag timeframe. Interviews favored early signals and usability. The prototype ingested RSV, detected changes, and sent email alerts. In 2016–2019, Google searches led cases by 1–3 weeks with moderate–high correlation e.g. fever r≈0.61–0.74, p-value<0.001. In 2020–2023, relationships weakened to no statistical significance, with partial recovery observed in 2023 for “fever” r≈0.73,
p-value<0.001. Field evaluation indicated “highest” overall feasibility ( =4.44, SD=0.71), supporting use as a supplementary digital signal alongside existing systems