Costs of Cancer
We are identifying variation in the type and costs of care for various cancers, treatments and patient groups using big data. Based on this information we will work with our experts and stakeholders to identify the cancers, stages and treatments that will benefit most from payment reform.
Objectives: We estimate the excess out of pocket paid by cancer patients compared to people with similar background without a cancer diagnosis. We identify the cancer types with the highest out of pocket costs. Moreover, we show our result across income categories and evaluate the drivers of out-of-pocket costs for cancer patients. We conclude by considering the implications for design and acceptability of value-based policies to enhance the sustainability of Medicare.
Data/Methods: We use the Sax Institute’s 45 and Up study of 260,000 residents living in New South Wales, Australia. We employ a matching model to investigate the excess out of pocket costs of those with cancer.
The problem: The direct costs of cancer treatment delivered out-of-hospital or in private hospitals, are shared between patients and the government and, despite the Medicare safety net, many still out-of-pocket can cause financial distress, though there is a lack of robust evidence to determine the extent of this problem. Still less is known about the indirect costs of receiving cancer treatment such as reduced household income if the patient (and their carer) have to stop working, or reduce their hours. Some may continue to feel the negative impact of reduced labour market attachment, and become increasingly reliant on government supports such as disability, carer, and unemployment payments. The extent and duration of these indirect costs are likely to be correlated with socioeconomic factors, particularly in light of the compounding effect of poorer health outcomes observed for cancer patients from lower socioeconomic groups in Australia.
Objectives: The project measures both direct and indirect costs to allow us to estimate the true economic cost of receiving treatment for cancer in Australia on individuals, households and society.
Data/Methods: The study will use the new, population level linked dataset of the Multi-Agency Data Integration Project (MADIP) managed by the Australian Bureau of Statistics. This detailed, individual-level dataset links federal records from: the Australian Tax Office (ATO), Department of Human Services and Centrelink, the Department of Health and national surveys including the Census.
First we will use population level descriptive statistics and data visualization methods to identify the distribution of out of pocket expenses and government expenditure on MBS and PBS items for cancer treatments. This will lay the foundation to use the richness of the MADIP data to implement a difference-in-difference method comparing outcomes between a ‘treatment group’ and a number of plausible ‘control groups’, before and after a cancer diagnoses/treatment.
The problem: The use and costs of healthcare, particularly hospital services, rise substantially in the months prior to death, regardless of the age at death or the cause of death. Although the use of specialist palliative care services has been associated with less costly care at the end of life, there is evidence of gaps and variation in the care available across Australia.
Objectives: This project investigates the costs of care in the last year of life for people dying from cancer and other life limiting conditions and if these differ according to access to specialist palliative care services. The focus will be on identifying variation in costs across geographic areas for similar patients, as well as differences by patient condition and other characteristics. Evidence for changing the payment mechanisms to deliver more appropriate and less costly end-of-life care will be established.
Data/Methods: The study uses the 45 & Up survey data linked to administrative data sets including cause of death, death registry, admitted patient data, emergency department, MBS and PBS. Two analytical approaches are being used: 1) A retrospective analysis examining costs in the final year for those who receive specialist palliative care and those who do not and; 2) a prospective analysis examining the costs trajectories for a cohort identified as having a high risk of death. We examine how costs and survival vary for individuals with similar risk status, and the observable and unobservable risks of having high end of life care costs.