ISSN:1054-3406    

DOI:10.1081/BIP-100101194

出版商:Taylor & Francis Group    

年代:1999

数据来源: Taylor

摘要:

When the classical approach of testing the null hypothesis of equality is used and the results are not statistically significant, formal conclusions regarding the “closeness” of the treatments cannot be drawn. When the purpose of the investigation is to exhibit “closeness,” misinterpretations may result in inappropriate, or even incorrect, conclusions. Here methodology for use when the goal is to exhibit the equivalence (noninferiority) of the treatments is discussed. The techniques allow direct conclusions to be drawn regarding the equivalence of the treatments. A review is presented.

ISSN:1054-3406    

DOI:10.1081/BIP-100101195

出版商:Taylor & Francis Group    

年代:1999

数据来源: Taylor

摘要:

We compare published methods for placing statistical confidence limits around incremental cost-effectiveness ratio statistics and show that only a nonparametric, bootstrap approach using polar angles gives completely reasonable results when neither treatment has significant advantages in cost or effectiveness. We also discuss alternative ways to report analytical results using plots, confidence or tolerance limits, and quadrant acceptability fractions. Finally, we use simulation to study the multiplicity bias that can be introduced into ICER confidence limits when only the most favorable results are reported over several possible choices of numerator cost measure and denominator effectiveness measure.

ISSN:1054-3406    

DOI:10.1081/BIP-100101196

出版商:Taylor & Francis Group    

年代:1999

数据来源: Taylor

摘要:

Phase I first-in-man studies in normal, healthy volunteers are performed to define a maximum safe dose and to identify a range of acceptable doses for later drug development studies in patients. Analysis of pharmacokinetic and pharmacodynamic data using mixed-effects modeling can be used to fit an overall dose-response relationship. By expressing prior information as pseudodata, the same methodology can be used to perform a Bayesian analysis and to determine posterior modal estimates for the model parameters. Decision theory can then be applied to maximize a chosen gain function, utilizing real-time data capture for choosing safe doses in a way that will provide more informative responses, thus accelerating study completion. The methodology is introduced elsewhere [1]. The purpose of this paper is to describe software currently in development and to illustrate the method using an example from a recent study.

ISSN:1054-3406    

DOI:10.1081/BIP-100101197

出版商:Taylor & Francis Group    

年代:1999

数据来源: Taylor

摘要:

Clinical trials are monitored to determine whether a treatment is safe and effective. If it becomes clear that treatment is superior to control, ethical considerations compel us to stop the study and make the treatment available to control patients. On the other hand, if it becomes clear that the treatment will not be shown superior to control, we would like to stop the study and save valuable resources for more promising agents. But how much evidence is enough to declare benefit, and what criteria do we use to stop for lack of benefit? This article reviews monitoring procedures designed to answer these two questions. TheB-value approach of Lan and Wittes [1] and Lan and Zucker [2] is used to unify the monitoring of many different kinds of trials, including those with continuous, dichotomous, or survival outcomes.

ISSN:1054-3406    

DOI:10.1081/BIP-100101198

出版商:Taylor & Francis Group    

年代:1999

数据来源: Taylor

摘要:

Hierarchical random-effects models can be used to estimate treatment or other covariate effects in single-study analyses coordinated over multiple clinical units and can also be extended to a wide variety of cross-study applications. After reviewing the single-study case, we use data from five trial protocols to look for units that tend to have treatment effects consistently above or below the study-specific grand mean across several studies. As a first step, we summarize the patient-level data as study-specific and unit-specific estimated treatment effects and standard errors using independent Cox regression models. We then compare the results of a hierarchical model using these data summaries as input to those produced by a more fully Bayesian method that uses the actual patient-level survival data. We also compare various different models using a deviance information criterion, a recent extension of the Akaike information criterion designed for hierarchical models. Our procedure appears to be effective at answering the question whether certain clinical units of the Terry Beirn Community Programs for Clinical Research on AIDS are better than others at identifying treatment effects where they exist.

ISSN:1054-3406    

DOI:10.1081/BIP-100101199

出版商:Taylor & Francis Group    

年代:1999

数据来源: Taylor

摘要:

This article deals with a unifying approach to approximate sample size determination for different types of hypotheses formulated in terms of two means of normally distributed data. A simple approximation is given to the sample size required for testing hypotheses about the ratio of the means. The formula includes the situations of testing noninferiority, superiority, or equivalence. We present a more general formula that also covers hypotheses formulated in terms of the difference of means. We show that over a wide range of parameter values the approximation provides reliable sample sizes.

ISSN:1054-3406    

DOI:10.1081/BIP-100101200

出版商:Taylor & Francis Group    

年代:1999

数据来源: Taylor

摘要:

We propose a procedure for establishing equivalence that determines whether a specified percentage of the treatment effect of a known active agent over placebo is maintained. This procedure accounts for the error in the estimates from the historical studies of the known active agent and placebo as well as the error in the estimates from the equivalence study of the new test treatment versus the active control. After the procedure is presented, it is compared analytically to a procedure in which the equivalence boundary is estimated from historical data and then used with a one-sided test. We address sample size requirements for the proposed equivalence procedure. We also illustrate the use of the proposed procedure with an example from the clinical area of thrombolytic therapy in acute myocardial infarction.

ISSN:1054-3406    

DOI:10.1081/BIP-100101201

出版商:Taylor & Francis Group    

年代:1999

数据来源: Taylor

摘要:

Assessment of quality of life (QOL) in clinical trials becomes a challenging task from the viewpoint of clinical biostatistics. The responses of the items for measuring QOL indices usually vary widely from patient to patient and from time to time. Measurement errors might be present in the responses of the items, and they might be correlated. Hence, in analyzing QOL data, the usual assumption that there are no measurement errors in responses is too liberal. Because the QOL indices are likely to be correlated, separate analysis of each index might not be efficient from the point of view of statistical methodology. We apply linear structural equation modeling (LISREL) in assessing the QOL data obtained from a clinical trial. A basic premise of the LISREL approach is that the abstract concepts (latent constructs) that are not directly measurable can be studied. LISREL is a statistical procedure for conceiving and testing structural hypotheses that cannot be tested adequately with other statistical procedures. It allows us to specify relations between unobserved and observed variables while controlling for measurement errors and correlations among both the measurement errors and the latent constructs.

ISSN:1054-3406    

DOI:10.1081/BIP-100101202

出版商:Taylor & Francis Group    

年代:1999

数据来源: Taylor

摘要:

We discuss the problem of estimating the median time and comparison of survival curves when data are nonrandomly censored in analgesic trials. In these trials patients experience post-surgical pain at the time of randomization. Time to onset of analgesia is measured by patient-administered stopwatches. An effective analgesic is one for which the median time to onset is “short.” The study design allows patients to remedicate if their pain persists, and this remedication prior to pain relief censors the time-to-onset measures. The time to onset for patients who remedicate is nonrandomly censored. Assuming noninformative censoring yields misleading results with the Kaplan-Meier method (for estimation of median time) and the log-rank test (for comparison of survival curves). This assumption can also obscure the superior effect of an effective analgesic over an ineffective one. We propose a simple and intuitive way to handle the nonrandomly censored data in analgesic trials in order to (a) estimate the median time to pain relief and (b) compare the survival distributions between treatments. The method proposed is applied to data collected from an acute pain clinical trial, and the results are discussed.

ISSN:1054-3406    

DOI:10.1081/BIP-100101203

出版商:Taylor & Francis Group    

年代:1999

数据来源: Taylor