TY - JOUR
T1 - An actuarial investigation into maternal out-of-hospital cost risk factors
AU - William, Jananie
AU - Chojenta, Catherine
AU - Martin, Michael A.
AU - Loxton, Deborah
N1 - Publisher Copyright:
© Institute and Faculty of Actuaries 2018.
PY - 2019/3/1
Y1 - 2019/3/1
N2 - This paper adopts an actuarial approach to identify the risk factors of government-funded maternal out-of-hospital costs in Australia, with a focus on women who experience adverse birth outcomes. We use a two-phase modelling methodology incorporating both classification and regression trees and generalised linear models on a data set that links administrative and longitudinal survey data from a large sample of women, to address maternal out-of-hospital costs. We find that adverse births are a statistically significant risk factor of out-of-hospital costs in both the delivery and postnatal periods. Furthermore, other significant cost risk factors are in-vitro fertilisation, specialist use, general practitioner use, area of residence and mental health factors (including anxiety, intense anxiety, postnatal depression and stress about own health) and the results vary by perinatal sub-period and the patient's private health insurance status. We highlight these differences and use the results as an evidence base to inform public policy. Mental health policy is identified as a priority area for further investigation due to the dominance of these factors in many of the fitted models.
AB - This paper adopts an actuarial approach to identify the risk factors of government-funded maternal out-of-hospital costs in Australia, with a focus on women who experience adverse birth outcomes. We use a two-phase modelling methodology incorporating both classification and regression trees and generalised linear models on a data set that links administrative and longitudinal survey data from a large sample of women, to address maternal out-of-hospital costs. We find that adverse births are a statistically significant risk factor of out-of-hospital costs in both the delivery and postnatal periods. Furthermore, other significant cost risk factors are in-vitro fertilisation, specialist use, general practitioner use, area of residence and mental health factors (including anxiety, intense anxiety, postnatal depression and stress about own health) and the results vary by perinatal sub-period and the patient's private health insurance status. We highlight these differences and use the results as an evidence base to inform public policy. Mental health policy is identified as a priority area for further investigation due to the dominance of these factors in many of the fitted models.
KW - Adverse birth
KW - Classification and regression tree
KW - Generalised linear model
KW - Maternal out-of-hospital costs
KW - Mental health policy
UR - http://www.scopus.com/inward/record.url?scp=85060601306&partnerID=8YFLogxK
U2 - 10.1017/S1748499518000015
DO - 10.1017/S1748499518000015
M3 - Article
SN - 1748-4995
VL - 13
SP - 1
EP - 35
JO - Annals of Actuarial Science
JF - Annals of Actuarial Science
IS - 1
ER -