Bias and Trials Stopped Early for Benefit

Abstract

However, the notable findings in the study by Talcott et al 1 and the earlier report by Zietman et al 2 were that (1) at 5 years, a significantly higher proportion of men remained free of biochemical failure in the higher-dose radiation group (80.4%) vs the conventional-dose treatment group (61.4%) and that (2) at a median of 9.4 years following treatment, there were no significant differences in key functional parameters between conventional-dose and high-dose treatment, such as urinary obstruction, urinary incontinence, bowel problems, and decreased sexual function. Furthermore, at a median follow-up time of 8.9 years, long-term cancer control was still significantly higher in the highdose vs conventional-dose treatment groups. 3 We suspect that these findings about symptoms probably were not necessarily only "because eventually the patients will adapt and stop complaining" (as Vikram suggests), but instead may represent important and meaningful outcomes for patients. Similarly, Talcott et al also found that patients who had received conventional doses were significantly more concerned about control of their cancer than patients who had received high doses, and also tended to express more regret.The quality-of-life consequences for men treated for prostate cancer and the psychological toll experienced by men concerned with disease recurrence should not be underestimated. The results from the study by Talcott et al add to the evidence that physicians can use when discussing treatment options for men with early prostate cancer by showing that high-dose radiation treatment is not associated with worse long-term adverse effects than conventional-dose treatment.

When to Stop a Clinical Trial Early for Benefit: Lessons Learned and Future Approaches

“…There is a lack of consensus among clinical trial experts about how statistical techniques and/or prespecified stopping guidelines can overcome this bias. [2][3][4] Ensuring that stopping guidelines require very strong evidence (proof beyond a reasonable doubt) and allow for accrual of an adequate number of events are important considerations.…”

“…[1][2][3][4][5][6][7][8][9][10][11][12][13] This practice has important implications from many viewpoints: clinicians who practice evidence-based medicine; future patients to whom the results of research studies apply; patients who voluntarily agree to participate in clinical trials; and scientists, investigators, and regulators who strive to balance conducting scientifically rigorous studies with disseminating data that support therapeutic advances as quickly as is reasonable.…”

Infusion of statistical science in comparative effectiveness research

“…Some have argued that well-done observational studies can provide effect size estimates similar to those obtained from (much more expensive) randomized trials [13,14]. Many authors disagree [4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] Ioannidis et al [20] conclude that differences can be and are common. D'Agostino and D'Agostino [21] argue that observational studies are quite appropriate for studying effectiveness in adding to the results of efficacy from RCTs.…”

Stopping a trial early – and then what?

“…Korn et al . [15] wrote, ‘Stopping a trial and releasing the information early allows current and future patients to benefit from new therapies as soon as possible’. Ellenberg et al .…”