Many medical publication professionals are aware of unpublished negative data from clinical trials and many believe obligations to make these data available should go beyond current legal requirements in the US and the EU, a new survey has found.

In the survey conducted during August by Ryan Woodrow of Woodrow Medical Communications and colleagues from Dianthus Medical, Network Pharma, Informa Healthcare Communications and Eastmond Medicomm, 58% of respondents were aware of negative data from any clinical trial remaining unpublished, while 32% were aware of negative data from a trial in the past three years that had not been published.

Of those professionals who were aware of selective data publication, 40% said an overriding reason for withholding data was that the compound in question had been discontinued; 36% cited journal rejection of the data; 31% poor trial design/methodology; 27% lack of resources, budget or key staff; 27% damage to the product profile; 22% lack of thought or discussion about making the data public; 20% investigator unwillingness to publish the data; and 14% said the data had been superseded.

The researchers found that 44% of respondents felt data should be made available prior to Phase II clinical trials, which would be earlier than under current US and EU legal requirements. Among these respondents, 10% believed the data should be available from preclinical trials onwards and 11% from the discovery phase.

As far as the type of data for publication went, 33% of respondents said all analysed data, including all protocol-specified/exploratory endpoints (but not raw, patient-level data) should be made available; 22% felt all analysed data should be published, but including only protocol-specified endpoints (not exploratory or raw data); 20% wanted all data published, including raw data; and 18% said just a brief report with summary tables should be available.

 Barriers to publication

Asked about the main barriers to publishing all clinical trial data in the future, 47% of respondents cited fear of data misinterpretation; 41% a negative impact on trial protocols; 38% lack of resources/budget; 38% reservations about enabling others to analyse raw data; and 24% protection of intellectual property rights.

The authors concluded that the views of medical publication professionals should be taken into account in the drafting of guidelines and legislation relating to publications.