Understanding Equivalence and Non Inferiority Testing

Question: Write an essay onUnderstanding Equivalence and Non Inferiority Testing. Answer: Introduction The research article deals with the effective standards developed in clinical settings that determines whether the current efficacies in use are similar or different from each other. The equivalence as well as non-inferiority with higher efficacies is either to find the difference between the two or to analyze the superiority than the tradition methods. However, the basic aim of the article states the concepts and the statistical methods implicated in testing the equivalence/ non-inferiority efficacies and the critical evaluation of the current usage of these efficacies. Procedure The simplest method that can be used to test equivalence is one-sided test known by TOST procedure which can be extended to testing equivalency in other parameters like odds ratios, means, etc. It is generally carry forward after checking its relevance at the (12) 100% confidence interval for the difference in efficacies because this method is identical to performing two one-sided tests. However, efficacy is measured by two ways firstly, through success rates, when higher is better and through, failure rates which when lower is better. However, after applying the TOST procedure of hypothesis testing using confidence intervals, the strength of evidence is carried out using p value. However, the test can be distinguished as: Traditional testing p-value is less than alpha such that hypothesis (equivalence) is established. Non inferiority p-value is readily available for the TOST. Equivalence Margin The equivalence margin is determined by such that a narrow difference in the value will cause difficulty in establishing equivalence. The value of the study helps in giving result as well as the credibility of the study through relevant evidence and sound clinical considerations. On the other hand, it is primarily designed to minimize the possibility of new theory that stands out to be non-superior to the sample. However, the margin of non-superiority can be evaluated through the previous studies. The displacement from the equivalence with a small fraction can be given by . Conversely, different like Kaul and Diamond, is governed by the maximum efficacy loss of the new theory. Before, the data is recorded, it is important to know the equivalence margin to maintain the Type 1 error at the anticipated level. On the other hand, if there is no difference it does not imply that there would be no equivalence. However, equivalence often lead to incorrect conclusions through wrong hypothesis, non-significant result and no considering of marginal equivalence. Nevertheless, the inconsistencies are measured through the graphs. In TOST and traditional testing, the difference lies between the confidence interval such that in the former, the null hypothesis is rejected and does not cover zero whereas in the latter, the confidence interval is included between the intervals (, ) (Walker and Nowacki 2011). Sample Size and Measure of Effect The sample size is the main dimension in determining number of observations through desired power such that the sample size depends on the equivalence margin. However, the sample size was collected between a new dug and an active control for testing the equivalence using the software PASS. The measure of effect is measured using the proportions in absolute or the relative way. The former explains the difference between the proportions called the absolute risk difference (ARD) whereas when the study is made on the ratios of proportions then the study measures relative risk (RR) as well as odds ratio (OR). The difference lies between the two is that absolute measures are independent of the baseline rate whereas relative measure is dependent to the denominator of equivalence margin (Walker and Nowacki 2011). Analysis of Data The analysis of data is made regarding the decision to be performed on per protocol (PP) or the intention to treat (ITT) analysis. In data analysis, in ITT, the patients either receive the treatment or follow the protocol whereas in PP, the patients receive the treatment, followed by the protocol. The difference in efficacies is smaller in ITT because it makes the equivalence/ non-inferiority hard to reject the null hypothesis and is based on the conservative approach. Therefore, discrepancies experienced in the results the possibility of exclusion bias and the reasons for the non-inclusion of patients in the PP analysis. Conclusion To conclude, the research article emphasizes on the increase in the equivalence due to the new therapies with the same effectiveness and better properties such as lower costs and convenience. Nevertheless, as depicted the confusion in the literature is regarding the equivalence/ non inferiority. The confusion is primarily due to the lack of transparency and uniformity. This applicable when the new methods are introduced with a state of affair that the guidelines for reporting equivalence/ non inferiority studies has improved. However, the use of equivalence/ non inferiority studies has increased if the clinician judges its value in the current times. Reference Walker, E. and Nowacki, A. (2011). Understanding Equivalence and Noninferiority Testing.J GEN INTERN MED, 26(2), pp.192-196.