Using a hepatitis B surveillance system evaluation in Fujian, Hainan, and Gansu provinces to improve data quality and assess program effectiveness, China, 2015.

ABSTRACT

BACKGROUND: Monitoring hepatitis B surveillance data is important for evaluating progress towards global hepatitis B elimination goals. Accurate classification of acute and chronic hepatitis infections is essential for assessing program effectiveness. METHODS: We evaluated hepatitis B case-reporting at six hospitals in Fujian, Hainan and Gansu provinces in 2015 to assess the accuracy of case classification. We linked National Notifiable Disease Reporting System (NNDRS) HBV case-reports with hospital information systems and extracted information on age, gender, admission ward and viral hepatitis diagnosis from medical records. To assess accuracy, we compared NNDRS reported case-classifications with the national HBV case definitions. Multivariable logistic regression was used to identify factors associated with misclassification. RESULTS: Of the 1420 HBV cases reported to NNDRS, 23 (6.5%) of the 352 acute reports and 648 (60.7%) of the 1068 chronic reports were correctly classified. Of the remaining, 318 (22.4%) were misclassified and 431 (30.4%) could not be classified due to the lack of supporting information. Based on the multivariable analysis, HBV cases reported from Hainan (aOR = 1.8; 95% CI 1.3-2.4) and Gansu (aOR = 12.7; 95% CI 7.7-20.1) along with reports from grade 2 hospitals (aOR = 1.6; 95% CI1.2-2.2) and those from non-HBV related departments (aOR = 5.3; 95% CI 4.1-7.0) were independently associated with being 'misclassified' in NNDRS. CONCLUSIONS: We identified discrepancies in the accuracy of HBV case-reporting in the project hospitals. Onsite training on the use of anti-HBc IgM testing as well as on HBV case definitions and reporting procedures are needed to accurately assess program effectiveness and ensure case-patients are referred to appropriate treatment and care. Routine surveillance evaluations such as this can be useful for improving data quality and monitoring program effectiveness.