Epidemiology applied to the regulation of medical devices in Australia

Abstract

This thesis presents the results of research projects I completed to meet the competency requirements of the MAE. I completed these projects while based at the Therapeutic Goods Administration (TGA) in Canberra from February 2018 to February 2022. A major MAE requirement is the response to an acute public health threat. In April 2018, I was commissioned to aid in the investigation of a cluster of cases of intraocular lens opacification in Australia detected during routine post-market vigilance activities. The aim was to identify a causative factor to the outbreak and take consequent regulatory action. I decided to implement a novel approach by applying the 10 steps of disease outbreak investigation to a non infectious event. By the end of a 3-year period, 13 cases of IOLO were reported to the TGA. The adverse event notification rate in Australia (141.3 per million) was 7 times higher than the worldwide figure (17.1 per million). Even though it was not possible to determine the root cause of the events, the initial hypotheses were ruled out: a medical device manufacturing related issue and a different pattern of reporting of adverse events by the Australian ophthalmologic community. As a consequence of the investigation, a change in the instructions for use (IFU) of the intraocular lens was put in place to inform the users about this rare but critical adverse event. I evaluated the current recall data collection by the TGA and proposed the implementation of a surveillance system to monitor and detect signals related to medical device recall actions in Australia. Attributes from the CDC guidelines for evaluating surveillance systems were used to frame the collection and analysis of information from several sources. I conducted interviews with key stakeholders, reviewed workflows and varied documentation. The main finding of this research was that the current structure in place for managing and capturing the data from a recall by the TGA is complex and includes several different databases and IT systems. A new system was proposed to improve the existing rudimentary data analysis and provide a framework to gather evidence and assist in dealing efficiently with recalls and implement prompt regulatory actions. Key recommendations included the implementation of an analytic tool for data analysis and improvement of the current online applications and reporting by the companies responsible of the faulty devices. My final core project merged the data analysis and the epidemiological study competency. The aim of this project was to create a predictive model for identifying problematic hip prostheses earlier in their lifecycle. For that purpose, I conducted a retrospective data analysis after linking two public health databases: the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) database, which captures revisions as primary outcome, and the internal adverse event database held by the TGA. Currently, the AOANJRR identifies outliers in its annual reports which are prostheses with a higher than anticipated rate of revision (HTARR). After linking the data collected by the TGA from 2008 to 2020 specifically for hip prostheses incidents, I tested different variables within the internal dataset in order to predict the hip prostheses which were identified as problematic in the 2018 AOANJRR Annual report. Unfortunately, I was not successful in creating that tool. The inability to precisely predict the outcome at the individual hip product level using adverse event data was probably related to poor quality of the reporting of adverse events and underreporting of incidents. Despite of this, some keyworks have the potential to correlate to prostheses with HTAAR and could be useful after improving adverse event reporting to the TGA. In this thesis, I present my MAE journey and illustrate the contribution of my work to strengthen the regulation of medical devices in Australia. Show more