Although conceived in different ways, unnecessary treatment has generally been understood in two ways: first, the likelihood that the treatment will benefit the patient is unacceptably low (quantitative futility); Second, the resulting quality of patient benefit is unacceptably low (qualitative futility).9,10 Some have suggested that physicians have authority over the former (as medical decision-makers) and patients and families over the latter (based on their values).11 Although this dichotomy has been challenged because medical decisions necessarily involve value judgments, 12,13 It is generally accepted that medical judgment, This treatment is probably unnecessary, is a necessary starting point. to discuss the value of continuing treatment. Another challenge is that advances in medicine make futility a moving target; New devices, procedures, and drugs can extend shelf life before evidence of efficacy has been provided.5,14 There are several approaches to assessing futility, including stochastic shortening, asymmetric stopping limits, and predictive power. 2. Use of asymmetric stopping limits. The uselessness limit may be based on how quickly you want to stop the study if the treatment is ineffective. However, the sooner you stop for ineffective treatment, the larger the sample size needs to be to detect a difference, if present. [2] More than one-third of the studies did not specify or when their interim analyses would be performed. Of those who pre-specified this trait, about half planned an interim analysis while about 50% of the total target information would be available. Others were already planning an interim analysis at 25% or only 75%. Two studies planned interim analyses at specific points in time, for example annually or after a certain number of years. However, there have been a few study reports in which all the details have been reported: an example is the study by Powers et al.
[S37], which shows its formal basis for a possible pause in an interim analysis, the alternative assumptions adopted in the calculation of the conditional power, the actual conditional performance result on which the judgment was based, and the fact that a predefined (and described) line of insignificance had been crossed. Predetermined plans for interim futility analyses should be reported; Key details include the number and timing of interim analyses, as well as criteria to explain futility and abandonment of the study. Guyatt GH, Briel M, Glasziou P, Bassler D, Montori VM. Br Med J. 2012;344:E3863. Various reasons were given for stopping a study early (Table 3). The most common reasons were that the desired effect of treatment was not observed in about 40% of studies, often with a verbal appeal to futility. Many researchers stated their general belief that the results would not change if the study continued, or that the final results would not be statistically significant, but without specifically mentioning the term “futility.” About 20% of studies indicated that a predetermined detention limit had been exceeded or that previous guidelines had been used. Other studies provided a variety of verbal summaries of outcomes, for example noting that `outcomes did not differ` between treatment groups, that the data showed `insufficient efficacy` or that `the intervention was inferior`. Some studies have verbally described its conditional force as being “close to zero” or involving low conditional power, noting that the study had “almost no chance” or was “unlikely to reach statistical significance.” One RCT claimed there was an “inability to prove efficacy”, while another described the results as “presumed” futility. The error in judgment or expected distortion of the treatment effect associated with futility stop rules cannot be easily calculated, as these amounts are a function of the actual (and therefore unknown) treatment effect itself.
Nevertheless, general theoretical and simulation work can shed light on the general patterns of misjudgment and bias that can be expected for various typical scenarios [1]. Awareness of these trends should help researchers interpret the study`s findings. At the very least, proper reporting of interim analyses and stopping rule protocols should alert users to these findings that studies that are discontinued prematurely due to futility will result in an underestimation of treatment effects, perhaps quite severe.
