|
Activity Number:
|
312
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #305391 |
|
Title:
|
An Association Model for Longitudinal Intercourse Data with Informative Censoring via Time-to-Conception
|
|
Author(s):
|
Kirsten J. Lum*+ and Rajeshwari Sundaram and Germaine M. Louis
|
|
Companies:
|
Eunice Kennedy Shriver National Institute of Child Health and Human Development, NIH and Eunice Kennedy Shriver National Institute of Child Health and Human Development, NIH and Eunice Kennedy Shriver National Institute of Child Health and Human Development, NIH
|
|
Address:
|
6100 Executive Blvd. Room 7B01, MSC 7510, Bethesda, MD, 20892,
|
|
Keywords:
|
Binary response ; Epidemiology ; Informative censoring ; Longitudinal response ; Survival ; Time-to-conception
|
|
Abstract:
|
This work is motivated by epidemiologists' interest in intercourse behavior during the fertile window, the days in a menstrual cycle when intercourse can result in conception. Intercourse on these days is viewed as a multivariate binary response, observed longitudinally (over multiple cycles), and complicated by informative censoring via time-to-conception. Intercourse and time-to-conception are jointly modeled using an association model and a discrete survival time model respectively. In this joint model, (i) dependency between intercourse acts is captured by a latent variable and is allowed to vary over days, (ii) dependency between intercourse and time-to-conception is captured by a second latent variable, and (iii) the logistic form of the marginal model for intercourse is retained. These approaches are illustrated with data from the New York State Angler Prospective Pregnancy Study.
|
