Reusing routinely collected clinical data for medical device surveillance



Hickey, Graeme ORCID: 0000-0002-4989-0054, Dunning, Joel, Buchan, Iain and Bridgewater, Ben
(2014) Reusing routinely collected clinical data for medical device surveillance. [Poster]

[img] Text
EPH Poster.pdf - Unspecified

Download (3MB)

Abstract

Background Following the public healthcare scandal surrounding Poly Implant Prothèse breast implants, there is increased focus on the surveillance of medical devices. A number of clinical specialties in the UK prospectively collect clinical data on procedures performed. We explore a surveillance programme in the case study of prosthetic aortic valve heart implants, reusing routinely collected data. Methods Demographic, comorbidity and operative pseudonymised data from the UK National Adult Cardiac Surgery Audit registry were extracted for all patients undergoing an aortic valve replacement (AVR) operation from 1998-onwards. Rules were developed to classify implants, recorded as free-text, by manufacturer, series, model and prosthesis type, and cleaning algorithms applied to the dataset. Patient outcomes are assessed across implants. Long-term mortality follow-up was tracked by record linkage to the Office for National Statistics death register, and surgical re-intervention tracked by reoccurrence in the registry. Results Data on 95,000 AVR operations were extracted. Prosthetic implants were classified into 97 models from ten manufacturers. There were substantial differences in implant volumes by manufacturers, deconstructed into temporal trends, prosthesis type and models, and healthcare providers. Significant differences were observed in outcomes between models. These differences are influenced by case-mix selection bias. Conclusion Reuse of routinely collected clinical data for medical device surveillance is viable and economically effective. Data collected, when properly analysed, can potentially be used to detect inferior devices, inform manufacturers and clinicians of device quality, supplement research, facilitate development of (inter-) national clinical guidelines for implant choice and inform businesses and healthcare procurement officers about market access. Linkage to other routinely collected data, including Hospital Episode Statistics, product data and other audits, offer richer surveillance capabilities.

Item Type: Poster
Depositing User: Symplectic Admin
Date Deposited: 27 May 2015 08:07
Last Modified: 02 Apr 2021 07:17
URI: https://livrepository.liverpool.ac.uk/id/eprint/2012062