Noninferiority (NI) assessment is often performed to evaluate whether a test treatment can preserve certain proportion of the treatment effect of an active control compared to placebo. Two methods are usually applied ...
详细信息
Noninferiority (NI) assessment is often performed to evaluate whether a test treatment can preserve certain proportion of the treatment effect of an active control compared to placebo. Two methods are usually applied for the NI assessment. One is the so-called synthesis method and the other is the two-confidence-interval method. To fully use available information, historical data of the active control may be borrowed not just for the NI margin specification but also for the actual NI assessment. In that case, the historical data will be used twice which may create some concern. Moreover, as the historical data have already been observed and fixed when the current NI trial is designed, the NI assessment with historical borrowing could be treated as a conditional analysis. In this research, we discuss the methods and the associated considerations including sample size calculation and Type I error probability control for historical data borrowing for the NI assessment. An example and simulations are used to illustrate the applications and compare the characteristics of the methods.
Multiplicity is a challenging statistical issue in drug discovery, and a particular example is microarray study. The traditional approach of controlling of the family-wise error rate (FWER) is conservative when the nu...
详细信息
Multiplicity is a challenging statistical issue in drug discovery, and a particular example is microarray study. The traditional approach of controlling of the family-wise error rate (FWER) is conservative when the number of tests is large. A more appropriate approach is to control the false discovery rate (FDR). Since the development of the Benjamini and Hochberg (BH) FDR procedure in 1995, many modifications have been proposed aimed at relaxing the requirement for independent test statistics or improving the power of the BH FDR procedure. Comparisons of these procedures in the current literature are not comprehensive and the conclusions on performances are inconsistent. The objectives of this article are three-fold: (a) to perform a more comprehensive comparison of extant multiple testing procedures using two real microarray datasets and various simulated data sets;(b) to explore potential reasons for the inconsistencies in published simulation results;and (c) to identify suitable FDR procedures under different scenarios according to covariance structure, percent of true null hypotheses among multiple tests, and sample size.
Missing data exist in all clinical trials and missing data issue is a very serious issue in terms of the interpretability of the trial results. There is no universally applicable solution for all missing data problems...
详细信息
Missing data exist in all clinical trials and missing data issue is a very serious issue in terms of the interpretability of the trial results. There is no universally applicable solution for all missing data problems. Methods used for handling missing data issue depend on the circumstances particularly the assumptions on missing data mechanisms. In recent years, if the missing at random mechanism cannot be assumed, conservative approaches such as the control-based and returning to baseline multiple imputation approaches are applied for dealing with the missing data issues. In this paper, we focus on the variability in data analysis of these approaches. As demonstrated by examples, the choice of the variability can impact the conclusion of the analysis. Besides the methods for continuous endpoints, we also discuss methods for binary and time to event endpoints as well as consideration for non-inferiority assessment.
This article rigorously proves superiority of the proportion chi(2) test to the logistic regression Wald test in terms of power when comparing two rates, despite their asymptotic equivalence under the null hypothesis ...
详细信息
This article rigorously proves superiority of the proportion chi(2) test to the logistic regression Wald test in terms of power when comparing two rates, despite their asymptotic equivalence under the null hypothesis that the two rates are equal.
Decision making is a critical component of a new drug development process. Based on results from an early clinical trial such as a proof of concept trial, the sponsor can decide whether to continue, stop, or defer the...
详细信息
Decision making is a critical component of a new drug development process. Based on results from an early clinical trial such as a proof of concept trial, the sponsor can decide whether to continue, stop, or defer the development of the drug. To simplify and harmonize the decision-making process, decision criteria have been proposed in the literature. One of them is to exam the location of a confidence bar relative to the target value and lower reference value of the treatment effect. In this research, we modify an existing approach by moving some of the "stop" decision to "consider" decision so that the chance of directly terminating the development of a potentially valuable drug can be reduced. As Bayesian analysis has certain flexibilities and can borrow historical information through an inferential prior, we apply the Bayesian analysis to the trial planning and decision making. Via a design prior, we can also calculate the probabilities of various decision outcomes in relationship with the sample size and the other parameters to help the study design. An example and a series of computations are used to illustrate the applications, assess the operating characteristics, and compare the performances of different approaches.
The problem of power and sample size determination for distribution-free multiple comparison tests of K treatments versus a control group is addressed. We define the power as the probability of correctly rejecting one...
详细信息
The problem of power and sample size determination for distribution-free multiple comparison tests of K treatments versus a control group is addressed. We define the power as the probability of correctly rejecting one specified or all K hypotheses, corresponding to the per-pair and all-pairs power, respectively. The power formulas are derived for both joint ranking and pairwise ranking mechanism for general multiple comparison problems, followed by explicit form of these formulas when the single-step, step-down, or step-up adjustments are applied. The proposed power and sample size calculation methods apply to scenarios both when the underlying distributions are known and when they are unknown but a pilot study is available. Numerical methods via quasi-Monte Carlo integration and Monte Carlo integration are assessed. Our simulation studies show the accuracy of the power and sample size calculation formulas. We recommend the Monte Carlo integration as the calculation algorithm. An example from a mouse peritoneal cavity study is used to demonstrate the application of the methods.
Asymptotically, the Wald-type test for generalised estimating equations (GEE) models can control the type I error rate at the nominal level. However in small sample studies, it may lead to inflated type I error rates....
详细信息
Asymptotically, the Wald-type test for generalised estimating equations (GEE) models can control the type I error rate at the nominal level. However in small sample studies, it may lead to inflated type I error rates. Even with currently available small sample corrections for the GEE Wald-type test, the type I error rate inflation is still serious when the tested contrast is multidimensional. This paper extends the ANOVA-type test for heteroscedastic factorial designs to GEE and shows that the proposed ANOVA-type test can also control the type I error rate at the nominal level in small sample studies while still maintaining robustness with respect to mis-specification of the working correlation matrix. Differences of inference between the Wald-type test and the proposed test are observed in a two-way repeated measures ANOVA model for a diet-induced obesity study and a two-way repeated measures logistic regression for a collagen-induced arthritis study. Simulation studies confirm that the proposed test has better control of the type I error rate than the Wald-type test in small sample repeated measures models. Additional simulation studies further show that the proposed test can even achieve larger power than the Wald-type test in some cases of the large sample repeated measures ANOVA models that were investigated.
In medical and clinical studies, videos, images, or other form of information carriers may be centrally read to provide scores or measurements for medically or clinically meaningful endpoints. Since the simple one cen...
详细信息
In medical and clinical studies, videos, images, or other form of information carriers may be centrally read to provide scores or measurements for medically or clinically meaningful endpoints. Since the simple one central reading process may lead to biased final reported scores, different central reading processes that aim to reduce the potential bias are used including the double central reading with adjudication, the two central reading without adjudication, and the three central reading processes. The current article compares these four central reading processes and rigorously proves three propositions under several common scenarios of interest. Analytical calculations and simulation studies were used to numerically compare these four central reading processes. Theoretical and numerical assessments suggest the use of the three central reading process if cost allows. With a reduced cost, the double central reading with adjudication process with a relatively large cutoff C should be used, while C needs to be carefully selected using appropriate numerical assessments or clinical justifications.
In high-throughput screening (HTS), compounds can be tested in mixtures using the orthogonal pooling strategy. However, several experimental studies that used two-way orthogonal pooling have shown that the rate of fal...
详细信息
In high-throughput screening (HTS), compounds can be tested in mixtures using the orthogonal pooling strategy. However, several experimental studies that used two-way orthogonal pooling have shown that the rate of false negative testing errors is still not negligible. Because replicate screening is prohibited due to the large number of compounds that need to be tested, the false negative results eventually lead to loss of active compounds. This article generalizes the two-way orthogonal pooling strategy to a three-way pooling and develops statistical methods to evaluate the false negative rate (FNR) and the false positive rate (FPR) for the two-way and three-way pooling strategies. Calculations show that the three-way pooling strategy decreases the FNR but increases the FPR to approximately three times that of the two-way pooling strategy. In terms of strategy selection, three-way pooling requires 50% more tests than two-way pooling but gains benefit when the testing technique has low sensitivity and high specificity, while two-way pooling may be preferred when the technique has high sensitivity and low specificity.
Patient recruitment is challenging in rare disease clinical trials. To save time and resources, an inferential seamless phase II/III clinical trial design is considered for a clinical trial in a rare disease area. In ...
详细信息
Patient recruitment is challenging in rare disease clinical trials. To save time and resources, an inferential seamless phase II/III clinical trial design is considered for a clinical trial in a rare disease area. In particular, 2 doses compared to a placebo control are evaluated at phase II (ie, stage I). Based on the results of a phase II intermediate endpoint, additional patients may be enrolled into the 2 doses and control, 1 selected dose and control, or none of the 3 treatment arms at stage II. All patients including those of unselected dose (s) will be followed for the measurements of the phase III (ie, stage II) primary and secondary endpoints and incorporated in the final analysis. Under a reasonable condition, the type I error rate will be controlled at the nominal level if the same nominal level is applied for testing a dose effect based on either only the data of patients of stage I (in case the dose is not selected for stage II) or data of patients of stages I and II combined. A graphical testing procedure is also introduced for multiplicity adjustments to account for the 2 doses and 2 endpoints for the overall type I error rate control. Moreover, an imputation approach is proposed for conditional power calculation at stage I for a sample size adaptive design. Simulations are carried out under different parameter configurations to compare the performances of various approaches. The trial example is further used to illustrate the applications of the methods.
暂无评论