Pooling overdispersed binomial data to estimate event rate
2008

Estimating Event Rates Using the Beta-Binomial Model

Sample size: 41 publication 10 minutes Evidence: high

Author Information

Author(s): Young-Xu Yinong, Chan K Arnold

Primary Institution: Harvard School of Public Health

Hypothesis

Can the beta-binomial model provide a valid method for combining event rates from overdispersed binomial data?

Conclusion

The beta-binomial model can provide a robust estimate for the summary event rate by pooling overdispersed binomial data from different studies.

Supporting Evidence

  • The beta-binomial model provides a better fit for overdispersed data compared to the binomial model.
  • The pooled event rate estimated using the beta-binomial model was 3.44%.
  • Tests for overdispersion indicated significant overdispersion (p < 0.05).
  • The beta-binomial model allows for the incorporation of study attributes into regression models.

Takeaway

This study shows how to combine results from different studies to get a better idea of how often side effects happen when using certain medications.

Methodology

The study describes the beta-binomial model and illustrates its application using data from multiple studies on oral antifungal treatments.

Limitations

The beta-binomial model may not be widely used in clinical research despite its advantages.

Participant Demographics

Patients aged 18 or above with superficial dermatophytosis or onychomycosis receiving oral antifungal therapy.

Statistical Information

P-Value

< 0.001

Confidence Interval

95% CI: 2.28% to 4.61%

Statistical Significance

p<0.05

Digital Object Identifier (DOI)

10.1186/1471-2288-8-58

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication