Assessing Genetic Risk Factors for Cervical Cancer
Author Information
Author(s): Walter Stephen D, Franco Eduardo L
Primary Institution: McMaster University
Hypothesis
Can latent class models improve the assessment of genetic polymorphisms associated with cervical cancer risk despite inter-laboratory variation?
Conclusion
The study provides a method to evaluate the association of a genetic polymorphism with disease while accounting for laboratory measurement errors.
Supporting Evidence
- Inter-laboratory agreement on genotype was only moderate.
- Latent class models provided maximum likelihood estimates of laboratory accuracy rates.
- The estimated odds ratio for the Arg/Arg genotype was 1.83.
- Using latent class models yielded more precise estimates than empirical methods.
Takeaway
This study shows that different labs can get different results when testing for a genetic risk factor for cervical cancer, but using special models can help us understand the true risk better.
Methodology
The study used latent class models to analyze data from a case-control study of cervical cancer, accounting for inter-laboratory variability in genetic testing.
Potential Biases
Potential biases due to measurement errors and inter-laboratory discrepancies in test results.
Limitations
The study was not originally designed to assess the association between polymorphism and disease risk, which may affect the robustness of the findings.
Participant Demographics
The study included 142 cases of cervical cancer and 162 controls, with participants from multiple countries.
Statistical Information
P-Value
0.09
Confidence Interval
1.02–3.75
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website