The objective of the project is to provide a statistical model for the analysis of longitudinal studies of infant and toddler mortality in order to better estimate the effects of specific health interventions in the presence of multiple dependent causes of death and heterogeneity (in terms of socio-economic status, measured biological frailty and unobserved heterogeneity). The work will decompose mortality into disease incidence and case fatality since interventions for specific diseases often target on or the other of these. The methods will expand statistical work on random effects and generalized estimating equation models to the case of multinomial outcomes. An important outcome will be means to quantify disease dependence. The work will draw heavily on the verbal autopsy questionnaire that has been recently developed. The theoretical work will be applied to a longitudinal data set on infant and toddler morbidity and mortality in Cebu, The Philippines.
| Becker, S; Peters, D H; Gray, R H et al. (1993) The determinants of use of maternal and child health services in Metro Cebu, the Philippines. Health Transit Rev 3:77-89 |
