A Novel Surveillance Evaluation Approach Using Clinical Text Extraction from the Hospital Information System: Case Study of Somdejpraboromrachineenart Natawee Hospital, Songkhla Province, Southern Thailand for Influenza in 2024

In 2024, Na Thawi District, Songkhla Province, reported the highest influenza cases in Southern Thailand. This cross-sectional study evaluated the influenza surveillance system (R506) at Somdejpraboromrachineenart Natawee Hospital in 2024 using both quantitative and qualitative methods. Stakeholders involved in the epidemiological surveillance system were interviewed to describe the system qualitatively. Quantitatively, 8,758 medical records from the hospital information system (HIS) and 358 R506 reports were reviewed to assess sensitivity, positive predictive value (PPV), completeness, accuracy, and timeliness. The female-to-male ratios were 1.17:1 in HIS meeting the R506 definition and 1.08:1 in R506. Most cases were in the 25–60-year age group in HIS and 5–9-year group in R506. The lowest proportions were among those aged 60 years or more. Cases peaked in July; HIS showed a gradual rise from May, while R506 surged from June. Most cases occurred in Na Thawi and Chana districts. Subdistrict-level data showed consistent hotspots in Na Thawi, Sathon, and surrounding areas. The overall incidence in the area was higher than in the reporting system. The sensitivity was 8.52% and the PPV was 84.92% with R506 showing 100% completeness and accuracy, except for onset date (21.79%). Timeliness was high: 98.88% within 3 days and 99.72% within 7 days. From the qualitative study, the stakeholders accepted the surveillance system, describing it as simple, flexible, stable, and useful for planning and resource allocation. Clinical text extraction enabled full review without the need for sampling.