NParCov3: A SAS/IML Macro for Nonparametric Randomization-Based Analysis of Covariance

Analysis of covariance serves two important purposes in a randomized clinical trial. First, there is a reduction of variance for the treatment effect which provides more powerful statistical tests and more precise confidence intervals. Second, it provides estimates of the treatment effect which are adjusted for random imbalances of covariates between the treatment groups. The nonparametric analysis of covariance method of Koch, Tangen, Jung, and Amara (1998) defines a very general methodology using weighted least-squares to generate covariate-adjusted treatment effects with minimal assumptions. This methodology is general in its applicability to a variety of outcomes, whether continuous, binary, ordinal, incidence density or time-to-event. Further, its use has been illustrated in many clinical trial settings, such as multi-center, dose-response and non-inferiority trials. NParCov3 is a SAS/IML macro written to conduct the nonparametric randomization-based covariance analyses of Koch et al. (1998). The software can analyze a variety of outcomes and can account for stratification. Data from multiple clinical trials will be used for illustration.

[1]  A. Agresti,et al.  Categorical Data Analysis , 1991, International Encyclopedia of Statistical Science.

[2]  J. Torner,et al.  A randomized controlled trial of high-dose intravenous nicardipine in aneurysmal subarachnoid hemorrhage. A report of the Cooperative Aneurysm Study. , 1993, Journal of neurosurgery.

[3]  Gary G. Koch,et al.  Categorical Data Analysis Using The SAS1 System , 1995 .

[4]  G G Koch,et al.  Issues for covariance analysis of dichotomous and ordered categorical data from randomized clinical trials and non-parametric strategies for addressing them. , 1998, Statistics in medicine.

[5]  Gary G. Koch,et al.  Nonparametric Analysis of Covariance and Its Role in Noninferiority Clinical Trials , 1999 .

[6]  G. Koch,et al.  Nonparametric analysis of covariance for hypothesis testing with logrank and Wilcoxon scores and survival-rate estimation in a randomized clinical trial. , 1999, Journal of biopharmaceutical statistics.

[7]  Gary G. Koch,et al.  Categorical data analysis using the sas® system, 2nd edition , 2000 .

[8]  G. Koch,et al.  Non-parametric covariance methods for incidence density analyses of time-to-event data from a randomized clinical trial and their complementary roles to proportional hazards regression. , 2000, Statistics in medicine.

[9]  G G Koch,et al.  Non‐parametric analysis of covariance for confirmatory randomized clinical trials to evaluate dose–response relationships , 2001, Statistics in medicine.

[10]  Lisa M LaVange,et al.  Randomization-based nonparametric methods for the analysis of multicentre trials , 2005, Statistical methods in medical research.

[11]  Ralph B. D'Agostino,et al.  Wiley encyclopedia of clinical trials , 2008 .

[12]  A. Herring,et al.  A robust method for comparing two treatments in a confirmatory clinical trial via multivariate time‐to‐event methods that jointly incorporate information from longitudinal and time‐to‐event data , 2009, Statistics in medicine.

[13]  G. Koch,et al.  Estimating Covariate-Adjusted Incidence Density Ratios for Multiple Time Intervals in Clinical Trials Using Nonparametric Randomization-Based ANCOVA , 2011 .

[14]  G. Koch,et al.  Estimating Covariate-Adjusted Log Hazard Ratios for Multiple Time Intervals in Clinical Trials Using Nonparametric Randomization Based ANCOVA , 2011 .