Estimating Infection Attack Rates and Severity in Real Time during an Influenza Pandemic
Author Information
Author(s): Wu Joseph T., Ho Andrew, Ma Edward S. K., Lee Cheuk Kwong, Chu Daniel K. W., Ho Po-Lai, Hung Ivan F. N., Ho Lai Ming, Lin Che Kit, Tsang Thomas, Lo Su-Vui, Lau Yu-Lung, Leung Gabriel M., Cowling Benjamin J., Peiris J. S. Malik
Primary Institution: The University of Hong Kong
Hypothesis
Can serological data coupled with clinical surveillance data provide real-time estimates of infection attack rates and severity during an influenza pandemic?
Conclusion
The study concludes that serial cross-sectional serologic data combined with clinical surveillance can provide reliable real-time estimates of infection attack rates and severity during an influenza pandemic.
Supporting Evidence
- Reliable estimates of infection attack rates and hospitalization probabilities could have been obtained if serological data were available weekly in real time.
- The study found that the ratio of infection attack rate to pre-existing seroprevalence was a major determinant for the timeliness of reliable estimates.
- Serological monitoring should be included in future pandemic preparedness plans.
- Had sero-surveillance begun three weeks after community transmission was confirmed, reliable estimates could have been obtained four weeks before the epidemic peak for younger age groups.
- Performance of sero-surveillance deteriorates if test specificity is not near 100% or pre-existing seroprevalence is not near zero.
- Computer simulations indicated that sero-surveillance with 300 specimens per week would yield reliable estimates when infection attack rates reach around 6%-10%.
- Serological data can complement clinical surveillance data to provide timely estimates of pandemic severity.
Takeaway
This study shows that testing blood samples can help us understand how many people get sick from a flu pandemic and how serious it is, which is important for keeping everyone safe.
Methodology
The study used a convolution-based method to analyze serologic and hospitalization data to estimate infection attack rates and infection-hospitalization probabilities.
Potential Biases
Potential biases include the reliance on convenience sampling and the assumption that asymptomatic cases develop antibodies at the same rate as symptomatic cases.
Limitations
The serologic specimens were collected via convenience sampling, which may not represent the general population, and the study assumed that the proportion of cases developing antibodies was similar across different groups.
Participant Demographics
Participants included blood donors aged 16-59, hospital outpatients aged 5-90, and children aged 5-14 from a community cohort.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website