Skip to main content

Table 2 Mixture Model and Mortality Functions for Simulation Study

From: Thinking outside the curve, part II: modeling fetal-infant mortality

Model feature

Specification for simulation study

Probability density for mixture model

.007 f(x;832,210) +.182 f(x;2772,740) +.758 f(x;3170,417) +.052 f(x;3804,413)

Risk within component 1

r 1(x) = logit -1(-4.6975 -0.2362 z + 0.3994 z 2 + 0.1690 z 3 + 0.1328 z 4)

Risk within component 2

r 2(x) = logit -1(-4.0962 -0.7496 z - 0.0289 z 2 - 0.1094 z 3 + 0.0918 z 4)

Risk within component 3

r 3(x) = logit -1(-5.7538 -1.7275 z + 1.6269 z 2 + 0.1897 z 3 - 0.0249 z 4)

Risk within component 4

r 4(x) = logit -1(-5.3285 -0.2786 z - 0.1979 z 2 + 0.0535 z 3 + 0.0773 z 4)

  1. The probability density for the mixture model used in our simulation study is specified, as are the mortality risk functions associated with the mixture model components. Above, z is defined as (x - 3000)/1000, where x is birthweight in grams.