The results of many clinical studies are never reported. This can result in patients not receiving the best treatment.

Major shortcomings in the reporting of clinical studies

Svenska 2021-01-13

Swedish researchers are bad at reporting the results of clinical studies. Many results cannot be found either in databases or in articles, according to an article from the organisation Transparimed. The consequence may be that patients receive poorer care. However, a change now seems to be on the way.

A clinical study is an investigation carried out on human beings, healthy or ill. The purpose is to test the effects of a medicine or a treatment.

In 2018, the ethical review boards, as they were then called, received more than 2 400 applications to carry out clinical studies in Sweden. Of these, 260 were trials of medicines. Both academic researchers and companies are behind the studies.

What they discover can be difficult to find out, however. Far from all results are reported to databases or journals. But the poor reporting is not just a Swedish phenomenon.

“Globally, the best estimate is that half are reported or published in some form. This is a waste of taxpayer money, and can result in patients not receiving the best treatment. It is also important for those planning a new study to know what has already been done. This interest also exists among patients and patient organisations,” says Till Bruckner, who is a political scientist and does research on a half-time basis at the Berlin Institute of Health Quest Center.

His remaining working hours are spent at the organisation Transparimed, which he established three years ago. The aim of the organisation is to ensure more studies are reported openly and accessibly to all. Transparimed collaborates with around 30 non-profit organisations, among them Health Watch UK, Transparency International and Cochrane.

Must be open access

According to the Declaration of Helsinki on ethical principles for medical research, clinical studies shall be public. For clinical trials of medicines, the requirements are more precise. If a trial is conducted in Europe, it must be registered in the European clinical studies database. The results must be reported in the same database no later than twelve months after completion.

There are currently major shortcomings in the reporting, however. This applies particularly when the body behind the study, the ‘sponsor’, is at a university.

“Major pharmaceutical companies report nearly all studies. University researchers are generally less good at reporting, but it varies between different countries,” says Till Bruckner.

His organisation has carried out a number of mappings of how reporting is managed in different countries. They recently presented a report of how things are working in Sweden and in Finland.

Only two of the twenty largest sponsors included in the investigation have managed reporting well – the companies Orion and Sobi. Among the universities and university hospitals, there are big gaps in the reporting.

“The universities don’t know how it works, but, when they are made aware of the problem, they deal with it. It seems that this is happening in Sweden now.”

Improvements in progress

Karolinska Institutet (KI), Uppsala University and its university hospital, and the University of Gothenburg have now all stated that they will be improving their reporting. KI started this work last spring, by establishing a central support function for researchers who are setting up new clinical studies. This is a good example, according to Till Bruckner.

“Here, the responsibility for reporting has to date rested with the researchers in charge of the studies,” says Sofie Possmark, who is one of the coordinators of the work at KI.

The reporting will now be coordinated by KI. Studies registered in the US database are managed via a central administrator account. Trials registered in the European database cannot be managed via a central account, however. Here, the coordinators instead have to inform and follow up the reporting in an alternative way.

Sofie Possmark believes that university researchers are used to presenting their results in articles, and therefore forget that they also have to update the results in the databases.

“Studies often lead to some kind of publication, but this does not always have open access, and then the general public cannot access the results,” she says.

Selective reporting

The unpublished studies are often those that show negative results. According to Till Bruckner, many investigations show that negative results are not reported as often as positive ones.

This is something that worries Matteo Bruschettini, the head of Cochrane in Sweden. The organisation carries out systematic reviews of different scientific fields.

“Sometimes we find studies, but not their results. This means that we do not get the whole picture. For example, if a study has been discontinued because it discovered dangerous side effects, this information is important to us.

But the probability that negative results are published is much lower than for studies with positive results,” emphasises Matteo Bruschettini.

“This means that they show an excessively positive effect of the medicine or the therapy. We want to highlight this problem.”

Cochrane was involved in bringing to light one of the more well-known examples of what shortcomings in reporting can lead to. In 2014, the organisation and the British Medical Journal, with the help of non-published data from the pharmaceutical company Roche, could show that the anti-viral medicine Tamiflu did not have the promised effect on virus infection.

It neither reduced mortality, nor the number of hospital stays. By that stage, the Swedish government had already spent 300 million SEK on Tamiflu.

More must report

Awareness is now growing of the importance of reporting the results of clinical studies – and the change is happening fast. In the UK, the percentage of reports has increased sharply over the last two years.

Today, the University of Leeds, Nottingham University, Imperial College and others are reporting 100 per cent of all clinical studies. Universities in Ireland, Germany, Denmark and Austria are also reporting results more and more.

One reason for the advances in the UK may be that the governmental Medical Research Council, the Wellcome Trust and other major research funding bodies now require studies to be registered. If this is not done, despite notices to do so, the study will not receive any more funding, Till Bruckner tells us.

Ethical rules shall guide

In Sweden, the Swedish Research Council for example does not require specifically that studies funded by it shall produce a final report. On the other hand, laws and regulations that apply, such as the Declaration of Helsinki, must be complied with. So says Madeleine Durbeej-Hjalt, Secretary General for Medicine and Health at the Swedish Research Council.

“In this way, the fact that results shall be reported is included indirectly. We are seeing few reports, however, and we intend to being a review of what we can do to ensure more studies are reported. We think it is important, both to report and to publish.”

The Swedish Medical Products Agency does not follow up the reporting of clinical trials of medicines either. On the other hand, the trials have to be conducted according to ‘good clinical practice’ (GCP). If this is not done, it might lead to a notice of failure to comply with GCP.

Text: Siv Engelmark