Findings of clinical research cannot be predicted in advance: an empirical study of cancer trials

Article type
Authors
Soares H, Kumar A, Serdarevic F, Cantor A, Clarke M, Hozo I, Djulbegovic B
Abstract
Background: How often innovative, experimental treatments are superior to standard, established treatments is not known. To describe accurately the outcomes of innovations that are tested in randomized controlled trials (RCTs) three factors have to be taken into consideration: publication rate, quality of trials and the choice of the comparator intervention.

Objective: To determinate how often innovative interventions are better than standard ones in the care of people with cancer.

Methods: All RCTs that were conducted and completed to date by four National Cancer Institute cooperative groups (COG, NSABP, NCCTG and RTOG) were eligible for the analysis. Out of 283 trials, data on 92,712 patients from 263 published trials and 2 unpublished trials were available for the analysis. The quality of the trials was assessed for possible effects of bias and random error on the outcomes of the trials. The possible impact of the choice of a comparator intervention was also assessed. To evaluate the outcomes, we extracted data on survival, disease free survival and treatment-related mortality.

Results: In terms of overall survival, innovative treatments were as likely as standard treatments to be successful- Peto odds ratio [OR] 0.98 (99%CI 0.95-1.01)]. Disease-free survival was slightly improved (2.1%) by administration of innovative treatments [Peto OR=0.92 (99%CI 0.89-0.94)]. Treatment-related mortality, however, was somewhat worse with innovative treatments (0.2%) [OR=1.64 (99% CI 1.29-2.10)]. In general, the quality of trials was high and did not affect the results of the treatment. Similarly, we did not find evidence that the choice of comparator influenced the outcomes of the treatments.

Conclusions: In aggregate, there is about an equal chance for innovative and standard treatments to result in successful outcomes or that the outcomes may not differ between two types of the treatments. We found no evidence that the outcomes of RCTs consistently and predictably favor innovative over control interventions, in spite of the trialists' hopes that new interventions are better. Our findings of equal odds of success for innovative versus standard treatments highlights the existence of prospective uncertainty about the outcomes of RCTs.

Acknowledgement: This study was funded by the US NIH (the grant number # R01 NS044417-01)