Estimating Prevalence Ratios in Clustered Epidemiological Studies
Author Information
Author(s): Santos Carlos Antônio, Fiaccone Rosemeire L, Oliveira Nelson F, Cunha Sérgio, Barreto Maurício L, do Carmo Maria Beatriz B, Moncayo Ana-Lucia, Rodrigues Laura C, Cooper Philip J, Amorim Leila D
Primary Institution: State University of Feira de Santana
Hypothesis
How can adjusted prevalence ratios be accurately estimated in clustered cross-sectional studies?
Conclusion
The study recommends using logistic models with random effects for analyzing clustered data and suggests choosing the method for estimating confidence intervals based on the study design.
Supporting Evidence
- The study analyzed data from two epidemiological studies with health-related outcomes in children.
- Results indicated major differences between estimated odds ratios and prevalence ratios.
- The delta method showed improved performance compared to bootstrap in certain scenarios.
Takeaway
This study helps researchers understand how to better estimate the prevalence of diseases in groups of people living close together, like in neighborhoods.
Methodology
Logistic models with random effects were used to estimate prevalence ratios, and confidence intervals were obtained using delta and bootstrap methods.
Limitations
The study's findings may not be generalizable to all types of clustered data or different epidemiological contexts.
Participant Demographics
Children aged 6 to 16 years from rural Ecuadorian communities and children aged 4 to 12 years from Salvador, Brazil.
Statistical Information
Confidence Interval
(1.11;1.79)
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website