Evaluation of the Influenza Surveillance System at Kosumphisai Hospital, Maha Sarakham Province

Abstract

Background: Influenza is a significant public health problem with an increasing trend in morbidity. In 2024, the morbidity rate was 343.10 per 100,000 population, making it one of the top ten diseases in Maha Sarakham Province. Kosumpisai Hospital had not previously evaluated its influenza surveillance system, and discrepancies were found between patient records and the D506 disease reporting system. This study aimed to describe the reporting process, assess the quantitative and qualitative attributes of the influenza surveillance system, and propose recommendations to improve system efficiency.

Methods: This cross-sectional study reviewed medical records of influenza and related diseases based on the Bureau of Epidemiology’s case definition. The study included both outpatients and inpatients at Kosumpisai Hospital from January 1 to December 31, 2024. Quantitative attributes assessed included completeness, positive predictive value, representativeness, data quality, and timeliness. Qualitative attributes were evaluated through interviews with key stakeholders involved in the surveillance process.

Results: The influenza surveillance system at Kosumphisai Hospital demonstrated low sensitivity (6.14%), primarily because reporting relied solely on ICD-10 diagnostic codes, resulting in underreporting of cases that met the surveillance case definition but were not coded accordingly. The positive predictive value was moderate (76.35%), supported by the integration between the HOSxP system and the D506 reporting system, which helped reduce data entry errors. The system showed high timeliness (96.43%) and good representativeness, with patient patterns consistent with seasonal influenza trends. Overall data quality was good, although some variables showed lower accuracy. Qualitative assessment indicated that the system is flexible, easy to use, and stable; however, in-depth data analysis to support decision-making and disease control remains limited. This is partly due to insufficient understanding among responsible staff on how to analyze surveillance data to detect potential clusters, resulting in the absence of systematic reporting of findings to hospital administrators.

Conclusion: The influenza surveillance system at Kosumpisai Hospital was considered fair but should be improved in terms of completeness, as reporting relied solely on physician-assigned ICD-10 codes. Some cases meeting the case definition were not reported due to code discrepancies. Nevertheless, the D506 reporting system linked with the HOSxP program facilitated rapid disease reporting and ensured good data quality.