Principal Investigator: Stuart Gansky, MS, DrPH
This study utilized statewide and local cross-sectional data sources to
develop, test refine an ECC risk association model,
subsequently tested with longitudinal data from Project 1. We took a
multifactorial and multilevel approach looking at
various individual, family and community characteristics, including
demographics, socioeconomics, acculturation, health beliefs and behaviors,
access to care, and physiology, using computationally intensive statistical
data mining methods such as multiple additive regression trees (MART). Risk
assessment models to be used in nondental settings
and race/ethniciry specific models were developed.
This has been the first study to simultaneously examine individual-, family-
and community-level characteristics related to ECC
with data mining techniques.
Secondary Data Analyses
of Disease or condition:
Early Childhood Caries
Gansky SA, Cheng NF, Pollick HF: Predicting early childhood caries with individual, family,
and neighborhood factors. J Dent Res 2005;84(Spec Iss A):12.
Gansky SA, Cheng NF, Shain SG, Weintraub JA, Ramos-Gomez F:
Early childhood caries prediction with knowledge discovery data mining tools.
J Dent Res 2006;
85(Spec Iss B):515.
Gansky SA, Cheng NF: Ensemble models for risk prediction with
survey and multilevel data. 2006; Session 47: Joint Statistical Meetings, Seattle
P, Aamodt K, Bui T, Sagier
B, Gansky SA: Evaluating 3 caries risk assessment
tools: CaMBRA, Cariogram
& CAT. J Dent Res
2008;87(Spec Iss A): 720.
Gansky SA: Dental data mining: potential pitfalls and
practical issues. Adv Dent Res 2003;17:109-14.
CH, Gansky SA, Ramos-Gomez F, Ngo L, Isman R, Pollick HF: The association of early childhood caries and
race/ethnicity among California preschool children. J Public Health Dent 2003; 63:38-46.
Dr. Stuart Gansky
UCSF Oral Epidemiology
3333 California Street, Suite 495
San Francisco, CA