MODELLING THE TREATMENT PATHWAY OF HER2-POSITIVE METASTATIC BREAST CANCER (HER2+MBC) IN AUSTRALIA, 2001 – 2015
Benjamin Daniels, UNSW
Tuesday October 24 12:00pm
Background: Metastatic cancer and its treatment are conceptualised in terms of “stages”. Clinical trials and most observational research are concerned with what is often just one stage out of several. In this project we attempt to use routinely collected Australian administrative health data to associate survival outcomes with the patterns of treatment for patients receiving trastuzumab for HER2+MBC.
Methods: This study is a work in progress that uses linked linked Pharmaceutical Benefits Scheme (PBS), Medicare Benefits Schedule (MBS), and Herceptin Program records for every patient in Australia dispensed trastuzumab through the in Herceptin Program between 2001 – 2015. We determined courses of anticancer therapy from first dispensing date until the last dispensing, plus 30 days; defining a break of ≥90 days between dispensings as a separate treatment course for each medicine. We used PBS records to identify dispensing events for other relevant, non-cancer medicines and MBS records to identify health service events. We will use multidimensional scaling and PAM clustering to identify groups for which we will detail the treatment trajectories and survival outcomes using a hidden markov model approach.
Results: 5,899 patients were dispensed trastuzumab for HER2+MBC between December 2001 – July 2015 (the end of the Herceptin Program). Median overall survival from initiation of trastuzumab was 30.3 months (IQR: 13.4 – 68.7) for the entire cohort.
Conclusions: The aim of this work is to study the entire treatment pathway for HER2+MBC cancer, using the largest cohort of trastuzumab-treated HER2+MBC patients in the world. Through examination of treatment patterns it may be possible to identify non-intuitive patient subgroups; identify points in the treatment timeline where adverse events are more likely to occur or the disease condition is more likely to worsen; provide a more accurate picture of treatment for patients; and to assess whether real-world treatment adheres to accepted guidelines and prescribing restrictions.
Benjamin Daniels is a PhD candidate and research officer with the Medicines Policy Research Unit, Centre for Big Data Research in Health, UNSW. He has worked as a data analyst and statistician for over 10 years in a range of research settings including pharmacoepidemiology, cancer epidemiology, sociology of natural disaster, and criminology. His current research focuses on using linked, routinely collected administrative datasets to explore the outcomes and patterns of care associated with the treatment of HER2-positive breast cancer. He holds an MA in Sociology & Social Statistics from McGill University, Montreal and a BS in Psychology from Drexel University, Philadelphia.