The influence of funding source on study characteristics in the Australian New Zealand Clinical Trials Registry (ANZCTR)




Poster session 1 Wednesday: Evidence production and synthesis


Wednesday 13 September 2017 - 12:30 to 14:00


All authors in correct order:

Seidler AL1, Hunter K1, Willson M1, Langford A1, Tan-Koay A1, Berber S1, Askie L1
1 NHMRC Clinical Trials Centre, University of Sydney, Australia
Presenting author and contact person

Presenting author:

Lisa Askie

Contact person:

Abstract text
Background: Clinical trials are increasingly funded by industry. The influence of funding source on research conduct and outcomes remains controversial. There is evidence that industry funding influences the information published after trial completion. It is unclear whether differences in study characteristics between industry- and non-industry funded trials can be detected prior to trial commencement at the time of registration.

Objectives: To investigate whether, and how, funding source influences study characteristics, as documented at the time of registration.

Methods: We examined the distribution of primary funding sources across all studies registered on the ANZCTR in 2016. We analysed whether funding source was related to the following study characteristics: target sample size, type of controls used, prospective versus retrospective registration, and the primary purpose of the study ie prevention, diagnosis, education or treatment. University-funded studies were used as the reference group.

Results: Study characteristics differed across funding sources (Figure). Of the 1753 registered studies, 14% were industry-funded. Industry-funded studies were less likely to use active controls eg standard care (Odds Ratio (OR)=0.43, 95% Confidence Interval (CI)=0.31 - 0.60), but they were more likely to be prospectively registered (OR=1.36, 95% CI=1.11 - 1.66). There was no statistically significant difference in target sample size (Median(Interquartile Range)industry=48(72); Median(Interquartile Range)university=60(80), p=0.68). Funding source was related to the primary purpose of the study (χ2(28)= 164.89, p<.001), with 85% of industry-funded studies aiming at treatment while governments and universities were more likely to fund prevention, diagnosis and education studies.

Conclusions: For industry-funded trials, the higher proportion of prospective registration indicates an awareness of the necessity to reduce bias. Yet, their reduced use of active controls may increase effect sizes and thus produce more favourable results. Non-industry funders are crucial to ensure research addresses not only treatment, but also prevention and education questions.