Data Linkage for Pharmacovigilance: Views of Healthcare Professionals on the secondary use of administrative NHS data

The Problem
Maximising patient safety, including through pharmacovigilance, is essential. Paediatric pharmacovigilance is a recognised priority. There is widespread use of medicines off-label in children and a recognised vulnerability to adverse drug reactions. Current systems of pharmacovigilance have deficiencies. Improvements have been suggested, such as facilitating the detection of adverse drug reactions through data linkage. Increasingly, routinely collected clinical data are kept electronically and the use of CHI numbers (unique patient identifiers used throughout the Scottish NHS) provides a mechanism for linkage. The views of healthcare professionals on the linkage of NHS data for pharmacovigilance purposes in children have not been explored; this work explores the acceptability of linking routinely collected healthcare data to inform the design of a new system for pharmacovigilance in children.
The Approach
A mixed methods study was conducted involving interviews with a purposive sample of professional stakeholders, focus groups with purposively selected frontline healthcare professionals (analysed using Framework analysis and a three-round Delphi survey with a random sample of healthcare professionals (nurses, pharmacists and doctors with an interest in paediatric medicine) in Scotland. The survey was structured using the Theoretical Domains Framework. Results were triangulated across the three arms of the study. Ethical approval was granted by the North of Scotland Research Ethics Service.
Findings
Interviews (n=23) identified issues with security, anonymisation and legal challenges that should be addressed prior to implementation. Recommendations included information to and involvement of the public as well as strict control of access to the data by vetting and imposing meaningful sanctions. Focus group participants (n=22, 6 groups) additionally identified potential issues with the feasibility of the planned data linkage as well as latent liability issues from results dissemination. Participants focused less on legal issues anticipating that standard governance protocols and current laws would be satisfied. The Delphi survey initially covered all the issues identified in the preceding work but after three rounds that retained items on which there was consensus, interest focused on professional standards, requirements for linkage and the use and form of potential feedback. The proposed data linkage was perceived as positive and necessary, with participants highlighting the benefits for research and for patients.
Consequences
The development of a paediatric linked database has support from professional stakeholders and healthcare professionals in Scotland. The proposed data linkage was perceived as addressing a gap in current knowledge. No insurmountable issues were identified, but a range of issues and concerns should be addressed prior to implementation. Front-line health professionals, ultimately responsible for collecting the data, saw feedback as a strong motivator for engaging with the project. The data from this study will be supported by and cross-referenced to a parallel study investigating the views of the public.