Developing a Dynamic Microsimulation Model of the Australian Health System: A Means to Explore Impacts of Obesity over the Next 50 Years

Table 3

Models for transition probabilities within health module.

Outcome	Years	Statistical model	Explanatory variables
Inadequate physical activity	2001–2007	Pooled dynamic logistic regression	Age, couple, bachelor degree, employed, kids in household, adults in household, household income, smoker in previous year, at-risk alcohol consumption in previous year, inadequate physical activity in the previous year, and health status in previous year
Smoker	2002–2007	Pooled dynamic logistic regression	Age, couple, bachelor degree, employed, kids in household, adults in household, household income, at-risk alcohol consumption in previous year, inadequate physical activity in the previous year, and health status in previous year
At-risk alcohol consumption	2002–2007	Pooled dynamic logistic regression	Age, couple, bachelor degree, employed, kids in household, adults in household, household income, at-risk alcohol consumption in previous year, inadequate physical activity in the previous year, and health status in previous year
Obese	2006–2008	Pooled dynamic logistic regression	Age, education, current smoker, physical activity, at-risk alcohol consumption, health status, and obesity t−1
Health status	2006–2008	Pooled dynamic generalised ordinal logistic regression	Age, couple, bachelor degree, employed, kids in household, adults in household, household income, smoking status, at-risk alcohol consumption, physical activity, obese, and health status, t−1
Private health insurance	2005–2007	Multinomial logistic regression	Age, gender, and change in status of (education level, labour force status, income, household structure, and marital status)
Medical services used	2001–2008	Logistic regression model and probability tables	Age, sex, education, labour force status, health status, and private health insurance
Prescriptions used	2001–2008	Logistic regression and negative binomial regression models	Age, sex, education, labour force status, income, health status, and age-sex interactions
Hospital admissions		Logistic regression and negative binomial regression models	Age, sex, labour force status, and private health insurance
Medical costs		Tables of costs per service	Age and sex
Prescription costs		Tables of costs per script	Age and sex
Hospital costs		Tables of cost per bed day	Age