A cross-sectional study of all clinicians’ conflict of interest disclosures to NHS hospital employers in England 2015-2016

OBJECTIVE We set out to document how NHS trusts in the UK record and share disclosures of conflict of interest by their employees. DESIGN Cross-sectional study of responses to a Freedom of Information Act request for Gifts and Hospitality Registers. SETTING NHS Trusts (secondary/tertiary care organisations) in England. PARTICIPANTS 236 Trusts were contacted, of which 217 responded. MAIN OUTCOME MEASURES We assessed all disclosures for completeness and openness, scoring them for achieving each of five measures of transparency. RESULTS 185 Trusts (78%) provided a register. 71 Trusts did not respond within the 28 day time limit required by the FoIA. Most COI registers were incomplete by design, and did not contain the information necessary to assess conflicts of interest. 126/185 (68%) did not record the names of recipients. 47/185 (25%) did not record the cash value of the gift or hospitality. Only 31/185 registers (16%) contained the names of recipients, the names of donors, and the cash amounts received. 18/185 (10%) contained none of: recipient name, donor name, and cash amount. Only 15 Trusts had their disclosure register publicly available online (6%). We generated a transparency index assessing whether each Trust met the following criteria: responded on time; provided a register; had a register with fields identifying donor, recipient, and cash amount; provided a register in a format that allowed further analysis; and had their register publicly available online. Mean attainment was 1.9/5; no NHS trust met all five criteria. CONCLUSION Overall, recording of employees' conflicts of interest by NHS trusts is poor. None of the NHS Trusts in England met all transparency criteria. 19 did not respond to our FoIA requests, 51 did not provide a Gifts and Hospitality Register and only 31 of the registers provided contained enough information to assess employees' conflicts of interest. Despite obligations on healthcare professionals to disclose conflicts of interest, and on organisations to record these, the current system for logging and tracking such disclosures is not functioning adequately. We propose a simple national template for reporting conflicts of interest, modelled on the US 'Sunshine Act'.