Clinical trials with negative results have traditionally been viewed as ‘failures’. However, in the current age of transparency, there is a strong movement to ensure that all data, positive or negative is published. Additionally, there is an increasing understanding that negative trials can generate critical scientific observations and learnings. Therefore, assessing the scientific impact of both positive and negative trials is of value and may potentially guide future clinical trials investment.

In this month’s issue of JAMA Oncology, researchers from the Southwest Oncology Group (SWOG), part of the National Cancer Institute’s National Clinical Trials Network, assessed the comparative scientific impact of positive versus negative clinical trials. They analysed 94 phase III studies conducted by SWOG during a 30-year period involving 46 424 patients.

The authors found that primary publications from positive trials (with statistically significant results in favour of the new, experimental therapy for a designated primary end point) were more likely to be published in higher impact journals and were cited twice as often as negative trials. However, when the citation count from secondary publications (any article that relied on data from a given phase III study) were also considered, there was no significant difference between the positive and negative trials.  Interestingly, 28% of the trials analysed had a positive outcome and roughly an equal proportion of clinical trials closed early due to failure to accrue patients. There was no statistical evidence of publication bias between positive and negative trials.

While the scientific impact of primary articles from positive trials is greater than for negative trials, this study demonstrates the sizable impact of negative trials, particularly when the outcomes are used in secondary research. The authors suggest that in some cases negative trials may provide a more advantageous data resource for secondary science as the treatment affect can be discounted and full data set used. Overall, the data support the substantial impact of clinical trials, both positive and negative, emphasising their value in advancing cancer care.

——————————————————–

Summary by Louise Niven, DPhil from Aspire Scientific.