Avoiding redundant health research

If you have NOK 25 million you want to spend on advancing medical science, how do you share this money out?

Text and image: Published in Norwegian in Aftenposten, Oct. 7, 2017.


Iain Chalmers and Paul Glasziou documented in 2009 that an estimated 85 per cent of health research is redundant and that such wastage could have been avoided. 85 per cent of health research is redundant and that such wastage could have been avoided.

They made this claim in an article on “Avoidable waste in the production and reporting of research evidence” which appeared in the prestigious journal The Lancet.

Several factors account for the high proportion of squandered spending, but Chalmers and Glasziou found that new projects make surprising little use of earlier work in the same field.

A 2011 study took a closer look at how often published clinical studies referred to previous research on the same issue. In roughly half the cases, none or only one of the earlier studies were cited, regardless of the number available.

Nor does the quality of a study have much influence on the frequency of future citations. On the other hand, work which supports the scientists’ own hypothesis is significantly more likely to be cited.

These tendencies increase the risk that research is duplicated unnecessarily. A 2009 publication took a closer look at studies on the use of aprotinin to limit bleeding in heart surgery.

The authors concluded that 52 projects out of 63 were superfluous, and that more than 5 600 patients were subject to gratuitous placebo trials or other controls.


This problem is attracting increasing attention, and The Lancet published a special issue on it in 2014. Chalmers contributed an article which addressed how unnecessary research could be avoided by changing priorities.

One conclusion was that “research funders and regulators should demand that proposals for additional primary research are justified by systematic reviews of what is already known.“

This refers to the use of surveys which summarise existing knowledge by searching systematically for all scientific work done in an area (including unpublished projects).

That would make it possible to acquire an overall view – and one which was as impartial as possible – of what the available findings say.

This approach has become the gold standard for research on the effect of health measures. That is precisely because individual studies are vulnerable to errors, so that a limited sample often provides a distorted picture of the knowledge base.

Sharing out

The Kavli Trust is due to share out NOK 25 million this year for health research in the Kavli countries of Norway, Sweden, Finland and the UK.

That represents a modest sum in the wider scheme of things, but the desire to avoid waste is just as great. In cooperation kavli/feature 2with Norway’s ExtraStiftelsen, the trust has accordingly made big changes to its allocation process for 2017.

Inspired by the Chalmers article, this kicked off with the trust board choosing an overall area of commitment – the mental health of children and young people – for its grants over the next three years.

A strategic scientific committee then reviewed updated systematic overviews to identify key knowledge gaps. Twenty-two of these were identified.

However, not all such gaps have to be closed, and priorities therefore need to be set between them. Those who are hopefully going to benefit from the results – patients, relatives or health personnel – should have a voice in deciding which research questions are important to answer.

As a result, the knowledge gaps were submitted to six interest organisations for assessment and prioritisation. Finally, the 10 gaps with the highest ranking were included in the call for research proposals. Thirty-one well-founded applications to share in the research funding were received by the deadline of 27 September 2017. These all addressed established and relevant knowledge gaps.

The Kavli Trust has thereby enhanced its awareness of research financing and established a structured approach with the primary aim of supporting useful work and reducing waste. That probably puts it a step closer to ensuring that the studies it funds are entrenched in genuine knowledge gaps and are relevant to users.

The assessment of the submitted proposals which is now under way will show whether this process really does meet the trust’s objectives. Scientific research is a resource under great pressure.

The trust receives far more good project ideas than it can fund. As the studies cited above show, good arguments exist for taking greater control of the priorities set. The trust would accordingly urge all institutions and funders to give greater emphasis to securing systematic overviews and user priorities when determining what health research to pursue.