Addiction research can produce results that governments and funders are not keen to share. Credit: Christopher Furlong/Getty

Kypros Kypri was pleased to receive funding from a government agency in the Australian state of New South Wales to study problem drinking. But when the contract arrived in 2012, he was surprised to find a demand that the agency could review and sign off on any reports before they were published. Other language allowed the agency to terminate funding without notice or explanation.

Kypri, who now studies the epidemiology of alcohol-related injuries at Australia’s University of Newcastle, saw this as a threat to academic freedom and so fought for months to have the fine print removed. Eventually, it was. But he has since realized that his experience is not unusual. In March, Kypri and his colleagues published the results of a survey indicating that many researchers who study addiction think that funders have interfered with their work — most commonly by censoring it (P. Miller et al. Addict. Behav. 72, 100–105; 2017).

The survey was completed by 322 authors who had published in the journal Addiction, and a little more than one-third of them reported interference at least once in their careers. That proportion must be taken with a pinch of salt — it is possible that researchers who had experienced interference were more motivated to respond to the survey than those who had not, for example. And some of the reports go back almost a decade. But the survey nevertheless captures more than 100 experiences of research interference, spread across Europe, Australasia and North America.

There is a long and well chronicled history of private companies striving to keep tight reins on the results of research that they fund, particularly when it comes to studies of tobacco or pharmaceuticals. The survey showed that this remains a problem despite public attention, which is disappointing. Indeed, respondents reported their perception that such interference is on the rise.

But there has been less attention paid to censorship by government agencies, which is perhaps motivated by fears that politically sensitive results will highlight flaws in public programmes and so generate bad publicity. Some researchers and academic institutions accept clauses such as those that Kypri encountered as standard contract language. More should object, as he did.

Survey respondents highlighted a fear that standing up to funders could jeopardize their future funding opportunities — particularly given that emerging for-profit research organizations might be more willing to accept limitations on their publications and study designs. Other researchers may believe the clauses to be harmless and unlikely to be brought to bear on their work.

To accept such limits, however, runs counter to the public interest. And the addiction-research survey shows that such clauses are not harmless. One European respondent said a epidemiology publication had been blocked because it was not in the interest of the sponsoring government department; another, from North America, said the government had enforced a request from an industry representative to remove recommendations in an epidemiology report. Researchers from Australasia looking at fatal drug overdoses said that after they published data that were embarrassing to the government department, they were denied access to that department’s data. Interference can also come in other forms. Researchers must be wary of limits that public or private funders may attempt to place on study design or data sharing. For example, one senior researcher in North America said that his team was allowed to access a particular data set only if it agreed not to ask a politically sensitive question about the effectiveness of a government policy. Journals and journalists should make it a habit to inquire about the conditions, if any, imposed on researchers by their funders, so that those conditions can be disclosed when results are disseminated to the wider public.

Trends in some countries are encouraging. Kypri has encountered many researchers in the United States who say their institutions would not let them accept research contracts with clauses that allowed funder interference. In 2016, the UK government was forced to exempt scientific research contracts from new rules that would have banned government-funded organizations from lobbying for change.

Since his experience in 2012, Kypri has begun to systematically collect examples of clauses in government contracts that could enable interference in research. He worries that in some areas, particularly his own Australia, the clauses have become so common that they are viewed as normal. But his experience shows that it is possible to push back and perhaps even find compromises that satisfy both funder and researcher — without compromising research integrity.