In a takedown of trials testing whether cell transplants could help ailing hearts, a team at Imperial College London suggests that the more discrepancies a study had, the more powerful the benefit it reported for patients with heart problems. The new paper  appeared today in BMJ and its implications are nicely explored in this story  in Forbes.
The cardiac cell therapy field has experienced highs and lows in the last decade, received hundreds of millions of dollars in research money, and most recently been drawn into a university investigation . There’s been much confusion over whether and how well this cell therapy really works—with some studies reporting a robust effect, and others none at all.
The BMJ authors dissected 49 clinical trials. And cardiology professor Darrel Francis and his colleagues say they have nailed down one big commonality among the upbeat studies: They typically had numerous “discrepancies,” sometimes 20 or more. Among them: inconsistencies in the number or type of patients (“Women present in early reports seem to have become men by later reports,” the authors write); statistically insignificant differences reported as significant; and contradictory information in figures and written results. The five clinical trials without any discrepancies reported no benefit from the therapy.
Today also marked the release of a meta-analysis of these same types of trials published in The Cochrane Library. And that one, not surprisingly, found a modest benefit, as touted in this University of Oxford press release . It included many of the same trials that Francis and his team examined.