Optimizing Use of Multistream Influenza Sentinel Surveillance Data
2008

Improving Influenza Surveillance in Hong Kong

Sample size: 1555 publication Evidence: moderate

Author Information

Author(s): Lau Eric H. Y., Cowling Benjamin J., Ho Lai-Ming, Leung Gabriel M.

Primary Institution: University of Hong Kong

Hypothesis

Can simultaneous monitoring of multiple streams of influenza surveillance data improve the accuracy and timeliness of alerts?

Conclusion

Simultaneous monitoring of multiple streams of influenza surveillance data significantly improves the accuracy and timeliness of alerts compared to monitoring aggregate data or any single stream alone.

Supporting Evidence

  • Simultaneous monitoring of 8 data streams improved alert performance.
  • Univariate models generally outperformed multivariate models.
  • Alerts from the GP network had better timeliness than those from the GOPC network.

Takeaway

This study shows that watching multiple sources of information about flu can help us notice when the flu is spreading faster, which is important for keeping people healthy.

Methodology

Time-series methods were applied to multivariate sentinel surveillance data from 50 private and 62 public clinics over 9 influenza seasons.

Potential Biases

Potential biases may arise from differences in healthcare access and reporting behavior among sentinel practices.

Limitations

The analysis was limited by the small number of annual cycles of sentinel data and the lack of a standard definition for peak influenza season.

Participant Demographics

Data were collected from private-sector general practitioners and public-sector outpatient clinics in Hong Kong.

Digital Object Identifier (DOI)

10.3201/eid1407.080060

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication