U.S. law requires certain research results to be posted online within a year of a study's end date, but a new analysis found that only about half of cancer drug research results are made public after three years.
Researchers who looked at 646 studies examining the safety and effectiveness of cancer drugs found that 55 percent were published online or in a medical journal three years after the studies' end dates.
"That has great potential to bias what we know about these drugs and devices," said Dr. Joseph Ross, who has studied the availability of research results but wasn't involved in the new analysis.
For example, doctors may believe that an unsafe or ineffective drug works if its researchers choose to only publish positive results from clinical trials, which are considered the gold standard of medical research.
"This is why it is important to have trial results posted in a timely manner independently of publication in journals to have these results available to the people concerned," Dr. Agnes Dechartres, one of the study's authors from the Hôpital Hôtel-Dieu in Paris, wrote in an email.
"All of this information needs to be in that common collective understanding about drugs and devices," Ross, from the Yale School of Medicine in New Haven, Connecticut, said.
As a result of concerns over unpublished research, the U.S. Food and Drug Administration Amendments Act was signed into law by then-President George W. Bush in 2007.
The law requires results from large, rigorous trials - with some exceptions - to be published on ClinicalTrials.gov, a database of research trials, no later than one year after the last data are collected, write the researchers in the Journal of Clinical Oncology.
For their study, Dechartres and colleagues looked at 646 clinical trials conducted in the U.S. examining the effectiveness and safety of cancer drugs. Those trials all had an end date between December 2007 and May 2010.
One year after completion, only nine percent of the results were published on ClinicalTrials.gov, 12 percent were published in academic journals and 20 percent were published in at least one of those locations.
After another two years, those numbers grew to 31 percent of results published on ClinicalTrials.gov, 35 percent published in academic journals and 55 percent published either place.
The researchers also found studies that were sponsored by drug companies - not government or academic institutions - were more likely to comply with the law.
Ross said there could be a number of reasons for the delay in publishing results, including researchers not knowing the requirement applies to them or that they're responsible for publishing findings on ClinicalTrials.gov. He said researchers would often rather publish in medical journals.
"It's a little bit of everything," he said.
The analysis did have limitations, however, according to Dr. Sandip Prasad from the Medical University of South Carolina and Dr. Charles Bennett from the University of South Carolina, who wrote an editorial accompanying the new study.
For example, the researchers only used one database to find whether trial results were published in medical journals, and that database doesn't include about 1,800 journals.
Still, Prasad and Bennett recommend requiring researchers to post trial results before receiving additional research funding. They also say the Food and Drug Administration should monitor the compliance of research funded by industry and an independent organization should provide data analysis on what research has been made available.
Ross said that people taking part in clinical trials can ask doctors if and when the study's results will be published online.
"If they start asking the physicians these types of questions, the imperative to make this happen more quickly will change," he said, adding that the number of results published to ClinicalTrials.gov will probably increase in the future.
"It's growing very quickly," he said.