WASHINGTON — Most clinical trials for cancer, heart disease and mental health are too small to offer adequate medical evidence, said a review of the US database of such studies released on Tuesday.

The analysis in the Journal of the American Medical Association was led by experts at Duke University in collaboration with the US Food and Drug Administration.

That partnership, known as the Clinical Trials Transformation Initiative (CTTI), examined more than 96,000 studies registered as of September 2010 with the website ClinicalTrials.gov, a US registry that was started in 1997.

In 2007, the registry became mandatory for most mid- to late phase interventional drug and device trials and study authors are now required by law to record key data elements, report basic results and adverse events.

The studies chosen for review were in "intervention medicine," in which human subjects enrolled to test a certain treatment in either heart, cancer or mental health.

Ninety-six percent of those trials had fewer than 1,000 participants, and 62 percent had 100 or fewer, the analysis found.

"While there are many excellent small clinical trials, these studies will not be able to inform patients, doctors and consumers about the choices they must make to prevent and treat disease," said lead author Robert Califf of the Duke Translational Medicine Institute in North Carolina.

Other weaknesses included the finding that 66 percent of the trials were conducted at one single center, and 47 percent were funded by organizations other than industry or the US government's National Institutes of Health.

Furthermore, many studies had little in common in the way they were designed, randomized and blinded to avoid bias, and nearly 60 percent had no committee to monitor or review the data.

Some had up to 71 "primary outcomes," making comparisons across studies difficult.

International peer-reviewed science journals often insist that studies be registered with a database like ClinicalTrials.gov before they can be considered for publication.

The database was initially ordered by Congress to make sure seriously ill patients could gain access to ongoing clinical trials.

But more engagement is needed on the part of scientists, clinicians and patients, ethics committees and trial sponsors, said an accompanying editorial by Kay Dickersin of the Johns Hopkins Bloomberg School of Public Health and Drummond Rennie, deputy editor of JAMA.

"It appears that despite important progress, ClinicalTrials.gov is coming up short, in part because not enough information is being required and collected, and even when investigators are asked for information, it is not necessarily provided," they wrote.

"If all trial investigators and sponsors, as well as the regulators, were equally engaged, this might help further the universal adoption of trial registration," they added.

It might also "improve the completeness, validity, and accessibility of information in the registries; and restore the public's confidence in investigators, sponsors, and the clinical trial enterprise."