Biomedical subjects

H Saranadasa

Publications and source records attributed to H Saranadasa.

3 recordsLinked to original sources

PubMed2000

Setting specifications for non-normally distributed data.

In this paper, we focus on estimating an upper specification for non-normally distributed data. Cornish-Fisher expansions were exploited for derivations of the method. The method was applied for upper specification calculations of two particle-size data sets, and simulation studies were also performed to demonstrate the accuracy of the approximations. The method was also used to calculate the critical values for delta(min) for the Anderson-Hauck test for equivalence using bootstrap samplings. The main advantage of the method is in calculating specifications with limited data.

Chemistry, Pharmaceutical↗

PubMed2000

RSD requirement for different sample size for blend sampling.

A formula for calculating the equivalent relative standard deviation (RSD) for different sample sizes other than the Office of Generic Drugs (OGD) draft guidance was derived. These critical values vary with the population RSD and the sample size and guarantee the same confidence of passing a batch similar to the OGD blend uniformity test. Simulation results showed that the normal approximations give very close results to the simulation results. An assessment of the proposed OGD blend uniformity criteria against the first-stage USP content uniformity (CU) criteria for tablets was also made.

Drugs, Generic↗

PubMed1998

The analysis of small-sample multivariate data.

Clinical pharmacology studies typically consist of few subjects per treatment group, but with many, possibly highly correlated, measurements taken per subject. Permutational methods for testing equality of the multivariate treatment means when the number of variables exceeds the number of independent subjects have been developed, but are highly computationally intensive. In this paper, a parametric test was derived, using Edgeworth expansions, for the case of two groups, and compared with competing test statistics proposed by Mercante and Johnson (7), Dempster (5), Chung and Fraser (3), Mantel and Valand (6). The proposed test compares favorably with the others in terms of type I error rate, as well as power, and has the advantages of not requiring computationally intensive resources and being easily extendable to more than two groups.

Clinical Trials as Topic↗