The number of children who are overweight is increasing in many countries. Being overweight as a child can be detrimental to health. Overweight children tend to become overweight adults. Furthermore, there is a link between adult overweight and many long-term diseases. Treating preventable adult diseases is costly to the health service.
But we don't know a lot about how much childhood overweight costs the health service. This is important because new treatments must show value for money. In this project we will capture the costs linked to being overweight as a child. We will use the UK’s largest collection of GP and hospital records. We will gather information on the number and type of contacts based on appointments, prescriptions and hospital stays. We will then calculate the total cost of treating each child based on the prices for each type of contact.
This information can be used to compare the costs of treating children living with overweight and obesity and compared with healthy weight children. Knowing the cost of childhood overweight will help the people who decide on which treatments should be made available select treatments that are value of money.
Our overarching aim is to estimate annual incremental health care costs associated with childhood excess weight and to disaggregate excess costs by weight status (overweight or obesity); gender; age groups; and cost/resource use category.
Our proposed study design for estimating excess costs is a retrospective cross-sectional analysis of Clinical Practice Research Datalink (CPRD) Aurum linked to Hospital Episode Statistics (HES) data, specifically, Accident and Emergency (A&E), admitted patient care (APC), and outpatient (OP) activities. In addition, we will use data on patient-level Index of Multiple Deprivation (IMD) to explore the potential effect that differences in socioeconomic status may have on the association between total health care costs and childhood excess weight.
The primary exposure for our study will be childhood weight status and outcomes will be incremental resource use and cost outcomes.
Using descriptive statistics, we will summarise the data in terms of percentages of children in each weight category, demographics (age, sex of infant, and ethnicity), and other relevant covariates such as neonatal factors and family socioeconomic circumstances. Cost outcomes estimated from resource categories of health care services such as primary care prescriptions, which are likely to have a high proportion of zero costs, will be analysed using a two-part regression specification. The two-part model is recommended as the choice statistical model for handling the substantial skewness and mass of zeros often seen in healthcare expenditure data. The first part of the two-part model will be specified using a regression model for binary outcomes such as the logit. We will select the second part based on the performance and goodness of fit of either the ordinary least squares (OLS) or generalized linear models (GLM) models. After specifying the models for the two parts, predictions from the first and second parts will be multiplied to generate the total predicted costs.
The primary outcome measure will be annual incremental health care costs associated with childhood overweight and childhood obesity relative to healthy weight and childhood obesity relative to childhood overweight. The secondary outcome measure will be annual incremental resource use associated with childhood overweight and childhood obesity relative to healthy weight and childhood obesity relative to childhood overweight.
Olu Onyimadu - Chief Investigator - University of Oxford
Olu Onyimadu - Corresponding Applicant - University of Oxford
Alison Hayes - Collaborator - University Of Sydney
Cynthia Wright Drakesmith - Collaborator - University of Oxford
Mara Violato - Collaborator - University of Oxford
Margaret Smith - Collaborator - University of Oxford
Nerys Astbury - Collaborator - University of Oxford
Stavros Petrou - Collaborator - University of Oxford
Alison Hayes - Collaborator - University Of Sydney
HES Accident and Emergency;HES Admitted Patient Care;HES Outpatient;Patient Level Index of Multiple Deprivation