|
Activity Number:
|
217
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #303683 |
|
Title:
|
Nonparametric Incidence Estimation from Prevalent Cohort Data
|
|
Author(s):
|
Marco Carone*+ and Masoud Asgharian and Mei-Cheng Wang and Daniel Scharfstein
|
|
Companies:
|
Johns Hopkins Bloomberg School of Public Health and McGill University and Johns Hopkins Bloomberg School of Public Health and Johns Hopkins Bloomberg School of Public Health
|
|
Address:
|
615 N. Wolfe St., Baltimore, MD, 21205,
|
|
Keywords:
|
incidence ; prevalent cohort ; cross-sectional sampling ; nonparametric maximum likelihood
|
|
Abstract:
|
Incidence is an important epidemiologic concept particularly useful in assessing an intervention, quantifying disease risk, and planning health resources. Incident cohort studies constitute the gold-standard in estimating disease incidence. However, due to material constraints, data are often collected from prevalent cohort studies whereby diseased individuals are recruited through a cross-sectional survey and followed forward in time. We discuss the identifiability of measures of incidence in the context of prevalent cohort studies and derive nonparametric maximum likelihood estimators and their asymptotic properties. We also discuss age-specific incidence and adjustments for the sampling scheme. We apply our methodology to data from the Canadian Study of Health and Aging and provide insight into temporal trends in the incidence of dementia in the Canadian elderly population.
|
