Use of real-world registry data to support the delivery of personalised and value-based healthcare

Use of real-world registry data to support the delivery of personalised and value-based healthcare

There is global recognition that real-world data about patterns of care and real-world evidence on treatment outcomes is an under-used source for health services research. Earlier work has shown that RWE/RWD may provide additional insights in the reasons for practice variation, the understanding of diffusion and access to health services as well as provide better estimates of outcomes in the real-world. Our CHRS unit has several projects analyzing (local) clinical repositories as well as linked data-sets to understand and improve variation in clinical practice.

Projects

Building Analytical Capability for Data-Driven Cancer Health Services Research (VCCC)

The program aims to make better use of health data that is already collected and stored. Work to be done is the identification of clinical registries and the linkage to hospital ad min data, primary care data and VAEDs, VEMDs and PBS. For this project, Biogrid Australia, will be the data-linkage platform. This strategically important VCCC program follows from several Australian reports pointing at the need to build capacity in health services research. The McKean report (2012) states that " ... simply increasing healthcare expenditure will not necessarily lead to improved outcomes." and recommends building HSR health economics capacity to support policy makers.

Professor Maarten IJzerman co-chairs the "Building Analytical Capability for Data-Driven Research" program with Professor Jon Emery (primary cancer care) and Doctor Meredith Layton at the VCCC.

Data linkage for examining treatment sequences and costs in metastatic castration-resistant prostate cancer in Australia (ePAD registry)

Clinical management of prostate cancer (PC) is challenging because of the challenges associated with defining the optimal treatment in early and advanced disease involving multiple lines of drug treatment. While some variation in care is expected, unwanted and unnecessary variation in service delivery may significantly impact overall survival and Quality of Life. Additionally, unwanted variation can also have major health economic implications, particularly given the cost associated with treatments considered to be equally effective can vary from AUD$500 per month (Docetaxel chemotherapy) to AUD$4,000 per month (Abiraterone or Enzalutamide). In this project we will use large registry data-sets (e.g. ePAD - electronic Castration Resistant Prostate Cancer Australian Database) to analyse treatment costs of advanced prostate cancer (aPC) and to investigate the role of new and existing biomarkers. Results will be presented to experts and patients to enable interpretation of the observed variation and to optimise cost-effective patient management for future patients based on better risk-stratification, incorporating existing and promising new biomarkers. The work carried out is an active collaboration with Doctor Ben Tran at Peter Maccallum Cancer Centre and Professor Peter Gibbs (WEHI and Western Health).

Simulation Modeling of biomarker-based metastatic Colorectal Cancer treatment in Australia (TRACC registry)

Traditionally, treatment strategies for a group of similar patients, e.g. metastatic Colorectal Cancer patients with a KRAS-mutation, were based on population-level results from clinical trials. Although such treatment strategies may result in an average improved outcome for the whole population compared to alternative strategies, there still may be an alternative strategy that is more effective for a sub-population of patients. Ultimately, the use of biomarkers can be optimized by combining decision modelling and operations research methods based on individual patient data, to provide patient-specific guidance on treatment sequencing decisions, rather than population-based guidance. As a first step towards patient-specific guidance on treatment decisions in metastatic Colorectal Cancer based on translation of individual patient data using decision modelling and operations research methods, the proposed research project aims to model current treatment sequences and patient flows within Australian hospitals based on patient-level data on treatment sequences, outcomes, biomarker expressions, and general characteristics (TRACC). Insights obtained from the resulting simulation model, and experiments that can be performed using this model, can be used to obtain an improved understanding of metastatic Colorectal Cancer treatment in Australia on a systems level, and identify opportunities for improvement, both in patient outcomes and system performance. This project is a collaboration with Professor Peter Gibbs (WEHI, Western Health), Doctor Hui-Li Wong (Melbourne Health).