Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Generalized additive models for cancer mapping with incomplete covariates.

Jonathan L French1, Matthew P Wand

  • 1Biostatistics, Global Research and Development, Pfizer, Inc, 50 Pequot Avenue, New London, CT 06320, USA. Jonathan_L_French@groton.pfizer.com

Biostatistics (Oxford, England)
|April 1, 2004
PubMed
Summary

Mapping cancer risk is enhanced by a new method accounting for missing covariate data. This approach improves spatial risk estimates, avoiding biases from incomplete information in public health research.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

A Model-Based Meta-Analysis Framework Quantifying Drivers of Placebo Response in Atopic Dermatitis Trials.

CPT: pharmacometrics & systems pharmacology·2025
Same author

Confounded exposure metrics.

CPT: pharmacometrics & systems pharmacology·2023
Same author

Complex genetic dependencies among growth and neurological phenotypes in healthy children: Towards deciphering developmental mechanisms.

PloS one·2020
Same author

Corrigendum: Exposure-efficacy Relationships for Vedolizumab Induction Therapy in Patients with Ulcerative Colitis or Crohn's Disease.

Journal of Crohn's & colitis·2018
Same author

Exposure-efficacy Relationships for Vedolizumab Induction Therapy in Patients with Ulcerative Colitis or Crohn's Disease.

Journal of Crohn's & colitis·2017
Same author

A linearization approach for the model-based analysis of combined aggregate and individual patient data.

Statistics in medicine·2014

Area of Science:

  • Epidemiology
  • Biostatistics
  • Geographic Information Systems

Background:

  • Cancer incidence maps are vital for public health, traditionally using aggregated area data.
  • Emerging point data and geographic information systems enable individual-level covariate analysis.
  • Missing covariate data poses a challenge for accurate spatial risk assessment.

Purpose of the Study:

  • To propose and validate a novel statistical method for mapping cancer risk with missing covariate data.
  • To address limitations of traditional complete-case analysis in spatial epidemiology.
  • To improve the accuracy of estimating spatial disease risk variations.

Main Methods:

  • Utilized a logistic generalized additive model (GAM) for point-referenced cancer data.

Related Experiment Videos

  • Employed a mixed-effects model representation for estimating covariate effects.
  • Developed an Expectation-Maximization (EM) algorithm with Laplace approximation to handle missing data and random effects.

    • Main Results:

    • Demonstrated that standard complete-case methods can produce biased estimates of spatial cancer risk.
    • The proposed EM algorithm effectively accounts for missing covariate values in GAMs.
    • Accurate spatial variation in cancer risk was estimated using the novel method.

    Conclusions:

    • The developed framework offers a robust solution for handling missing covariate data in spatial epidemiological studies.
    • This method enhances the reliability of cancer incidence mapping and public health surveillance.
    • Accurate spatial risk assessment is crucial for targeted public health interventions.