Comparing pharmaceutical company payments in the four UK countries: a cross-sectional and social network analysis

To examine the characteristics of pharmaceutical payments to healthcare and patient organisations in the four UK countries. Compare companies spending the most; types of organisations receiving payments and types of payments in the four countries. Measure the extent to which companies target payments at the same recipients in each country and whether it differs depending on the type of recipient.

Cross-sectional comparative and social network analysis.

England, Scotland, Wales, Northern Ireland.

100 donors (pharmaceutical companies) reporting payments to 4229 recipients (healthcare organisations and patient organisations) in 2015.

For each country: payment totals and distribution; average number of common recipients between companies; share of payments to organisations fulfilling different roles in the health ecosystem and payments for different activities.

Companies prioritised different types of recipient and different types of activity in each country. There were significant differences in the distribution of payments across the four countries, even for similar types of recipients. Recipients in England and Wales received smaller individual payments than in Scotland and Northern Ireland. Overall, targeting shared recipients occurred most frequently in England, but was also common in certain pockets of each country's health ecosystem. We found evidence of reporting errors in Disclosure UK.

Our findings suggest a strategic approach to payments tailored to countries' policy and decision-making context, indicating there may be specific vulnerabilities to financial conflicts of interest at subnational level. Payment differences between countries may be occurring in other countries, particularly those with decentralised health systems and/or high levels of independence across its decision-making authorities. We call for a single database containing all recipient types, full location details and published with associated descriptive and network statistics.