Multistate models of postpartum infecundity, fecundability and sterility by age and parity: Methodological issues
How do hidden physiological processes influence estimates of fecundability and sterility? Does unobserved heterogeneity play a role in these estimates? To address these questions mathematical models of the reproductive process are needed. It is not well known how to evaluate characteristics of reproductive models based on observed reproductive history data, and such models may not be identifiable without ancillary information. However, little is known about how to introduce ancillary information into reproductive models. Furthermore, even if such information was involved, the use of standard software packages for maximization of the likelihood function is often not feasible, because the function cannot be represented in an explicit parametric form. In this paper we propose an approach which represents the likelihood function in a form useful for further analysis. This approach is based on multistate models of the basic physiological processes that influence reproductive outcomes, and it is suitable in applications where ancillary information is given in the form of hazard rates. As an alternative, a competing risks model with incomplete information is discussed. © 1993 OPA (Overseas Publishers Association) Amsterdam B.V. Published under license under the Gordon and Breach Publishers imprint.
Yashin, AI; Iachine, IA; Andreev, KF; Larsen, U
Volume / Issue
Start / End Page
International Standard Serial Number (ISSN)
Digital Object Identifier (DOI)