Less than half of clinical trials published in leading nursing journals are officially registered, according to a study, whose authors say their findings raise serious questions about standards and “do not cast the profession in a positive light”.
The new research was led by Richard Gray, professor of clinical nursing practice at La Trobe University in Melbourne, Australia, who said trial registration was vital in ensuring that all the evidence about an intervention, drug or treatment was publically available and could be checked and scrutinised.
“As a profession we are failing to adhere to what is internationally regarded as best practice”
“It means researchers can’t hide inconvenient results and they can’t change how they report the findings of trials if they don’t get the results they were expecting. In short, registration keeps researchers honest,” he told Nursing Times.
“The failure to register a trial is bad science and there really is no excuse,” he said. “In my opinion, if a trial is not registered it should raise questions in the reader’s mind about the overall integrity of the work.”
His study focused on randomised controlled trials – said to be the gold standard of research – published in three top international nursing journals between August 2011 and September 2016.
The journals in question were the Journal of Advanced Nursing, the International Journal of Nursing Studies and the Journal of Psychiatric and Mental Health Nursing.
“It would be interesting to explore this further and talk with editors and nurse researchers about the issue”
Of the 135 trials included in the study, 58% were found not to be registered. Meanwhile, around a quarter were registered retrospectively – after the trial had begun – with the delay varying from 39 days to more than six years.
Only 24, equivalent to 18%, were registered prospectively – before anyone had been recruited to take part and acknowledged as best practice among the scientific and research community. Of those, only 11 met criteria for “timely registration”.
The study, published in the open access journal Research Integrity and Peer Review, contrasted the findings with research on trials featured in leading medical journals that suggested nearly all – 96% – are registered.
“There is an unacceptable difference in rates of trial registration between leading medical and nursing journals,” stated the paper. “Concerted effort is required by nurse researchers, reviewers and journal editors to ensure that all trials are registered in a timely way.”
The study paper suggested nurse researchers may not be registering trials because of administrative delays, lack of awareness of the importance of registration and vague guidelines from journals.
“I worry nursing uses its ‘emerging discipline’ status as an excuse for poor science”
However, it largely ruled out admin issues because none of the trials registered retrospectively were done so within a month of the trial starting – the kind of delay that might simply be down to paperwork.
Professor Gray said lack of registration was a global issue for the nursing profession.
“From talking to colleagues, I get a sense that they do not perceive registration as important and/or it somehow does not apply to them,” he said.
“Journal editors do not seem to take the issue particularly seriously,” he said. “It would be interesting to explore this further and talk with editors and nurse researchers about the issue.”
He maintained that nursing journal editors themselves were “pivotal to changing practice”.
“Editors can require authors to register their trials. If authors have to do this retrospectively, editors can make sure they explain why they failed to register prospectively. Medicine has done it, nursing needs to catch up,” he said.
Nurse research often ‘failing to adhere to trial best practice’
While many nursing journals encourage trial registration, his study paper argued that editors, researchers and authors have “not spoken with the same clarity and consistency as their medical colleagues”.
“Medicine has been well organised, identified a problem and got it sorted,” added Professor Gray. “Submit an unregistered trial to the British Medical Journal and it will bounce back with a rejected stamp without going to peer review.
“I worry nursing uses its ‘emerging discipline’ status as an excuse for poor science,” he said. “I do also think there is an issue with how nurses are taught research, a lack of focus on the basics of how to run a trial.”
Overall, he said the fact many nursing-related trials were not registered risked de-valuing nursing research as a whole.
“As a profession we are failing to adhere to what is internationally regarded as best practice,” he said. “It does not cast the profession in a positive light.”
Responding to the claims, Professor Ian Norman, editor-in-chief of the International Journal of Nursing Studies, questioned some of the criticism around trial registration but agreed that the profession needed to consider the issues raised by Professor Gray’s study.
“This study show that there is much more to be done”
He said: “Prospective registration of studies gives an opportunity for the scientific community to determine that what is published properly represents the study that was planned and potentially avoids sources of bias such as selective reporting of outcomes.
“I agree that journal editors have an important role in encouraging this and the results of this study show that there is much more to be done,” he said. “However, the reality is that it is not common practice to prospectively register trials in many countries and retrospective registration achieves very little.
“Mandating registration at this point would probably just encourage spurious registrations after the event. However, we do need to think about how we move on from here,” he told Nursing Times.
Wiley, the company that publishes both the Journal of Advance Nursing and the Journal of Psychiatric Mental Health Nursing, said it had recently tightened its guidelines on trials.
“The reality is that it is not common practice to prospectively register trials in many countries”
In a statement provided to Nursing Times, a spokeswoman said: “In 2016, the Journal of Advanced Nursing enhanced its approach towards trial reporting with manuscripts rejected if they do not properly meet the registration criterion.
“These guidelines state that ‘trials should be registered publicly and the registration number be provided in the manuscript, after the abstract’,” said the spokeswoman.
“It also notes that ‘we cannot consider manuscripts reporting trials which were registered retroactively’, and requires authors to complete a Consolidated Standards of Reporting Trials (CONSORT) checklist with their trial submission,” she said.
She added: “The Journal of Psychiatric and Mental Health Nursing also requires the completion of a CONSORT checklist as a minimum, and is currently in the process of revising its author guidelines to include a clear policy on clinical trial reporting and registration.”