Study: 39% of U.S. clinical trial results are not made public

A study based on a random sample of 400 U.S. clinical trials found that 39% had not made their results public on trial registries or in scientific journals more than three years after they had been completed, with industry-funded trials performing worst.

The findings highlight that the problem of unreported trials, which harm patients and undermine public health, remains unresolved.

Timely availability of clinical trial results is of critical importance for patients and healthcare providers. The study authors call for continued research and efforts to bridge the gaps in result disclosure behavior. By addressing these gaps, vital trial information can be readily accessible to medical professionals, researchers, and the general public, enabling informed decision-making and advancing medical knowledge for the benefit of all.

NIH research waste

In the case of publicly funded trials, unreported trials also directly waste taxpayers’ money, as invisible research benefits neither science nor patients.

National Institutes of Health policies have long required trial results to be made public, but the study found that the results of 40% of NIH-funded trials remained completely unpublished.

NIH recently started checking whether trials it funds actually make their results public. However, NIH still do not appear to be tackling the large mountain of research waste that their past inaction allowed to pile up over the course of many years.

Slow reporting speed

According to World Health Organisation best practices, the results of all clinical trials should be made public on trial registries within 12 months of trial completion, and then in academic journals within 24 months.

The study found that that the median reporting time across all U.S. trials was 22.1 months. The median time from the primary completion date of a trial to the availability of results was 18.8 months for ClinicalTrials.gov and 25.6 months for PubMed.

Across funders, the study team found that of the results available, 24.5% were exclusively found on ClinicalTrials.gov, 52.2% were exclusively found in scientific journals (via PubMed), and just 23.3% were available on both platforms. Publicly funded trials had a higher proportion of results available on ClinicalTrials.gov, whereas trials funded by industry and by other sources had more results available on PubMed.

Were any laws broken?

US disclosure laws mandate results publication on registries within 12 months for some types of trials.

The study did not distinguish between clinical trials that have to report results under US law, and those that are not subject to legal reporting requirements. Compliance with the law has increased in recent years, despite a lack of effective enforcement by the FDA, but many trials fall outside the law’s scope.

In contrast, in the UK, the ethics regulator requires the results of all clinical trials to be made public within a year of trial completion, without exceptions. The UK government is expected to introduce supporting legislation soon.

Methodology and limitations

The study team extracted 9,102 trials registered on ClinicalTrials.gov that involved at least one US facility and were initiated between January 2015 and August 2018. Of these, 400 trials were randomly sampled, with 100 trials representing each of the following funder types: National Institutes of Health, US federal agency (excluding NIH), industry, and ‘other’ (e.g., foundation). Minimum follow-up was 36 months following the primary completion date.

The study’s methodology is strong.

Four researchers independently searched for results for each trial (searches by just two people are standard in the field). The study team did not contact investigators or sponsors to ensure that no relevant results had been overlooked. The authors note that the exclusion of longer-running trials may limit the generalizability of the findings. Disappointingly, the study was published behind a paywall and the underlying data are only available upon request.

The blog above was written by Sophie Guzikowski, who is completing an MSc in International Health Policy at the London School of Economics and Political Science. She previously worked in clinical research, and is currently doing an internship with TranspariMED’s nonprofit partner organization Concilium Scientific.