Search PubMed⌕ Search

Biomedical subjects

D R Bristol

Publications and source records attributed to D R Bristol.

13 recordsLinked to original sources
PubMed1996

Determining equivalence and the impact of sample size in anti-infective studies: a point to consider.

The problem of establishing the equivalence of an experimental treatment to a control with respect to a binary ("success" or "failure") response variable may be solved using an approximate (1-alpha) 100% confidence interval for the difference in the response rates (i.e., success probabilities). If the goal is to show that the experimental treatment is not sufficiently worse than the control, then a decision rule based on the magnitude of one confidence limit can be used. A procedure suggested by the Food and Drug Administration allows the value to which the confidence limit is to be compared to depend on the data. The consequences of determining the sample size assuming that the aforementioned value is fixed are examined. The probability of declaring equivalence and exact sample sizes are also presented for the procedure.

Anti-Infective Agents↗
PubMed1993

Planning survival studies to compare a treatment to an active control.

Rubinstein et al. presented a procedure for determining the required duration of accrual for a clinical trial comparing the survival distributions of two treatments using a classical hypothesis testing formulation. Here their testing procedure is modified in two ways. First, the asymptotic variances used in computation of the probabilities of type I and type II errors are based on the values of the parameters specified by the null and alternative hypotheses, respectively. Second, the null hypothesis is modified for situations where it is desired to show that the experimental treatment is better or not much worse than the control.

Clinical Trials as Topic↗
PubMed1993

Sample size determination using an interim analysis.

Interim analyses are often employed to terminate comparative clinical trials for ethical or economic reasons when the evidence indicates that one treatment is superior to the other. Here an interim analysis is proposed for the situation where a one-sided test is to be performed. The proposed interim analysis consists of a one-sided test to terminate the clinical trial if it appears that the null hypothesis of interest is true. By noting that incorporation of a single interim analysis is similar to the two-stage procedure used for constructing a test procedure with power independent of the unknown variance, it also includes estimation of the variance, which can be used to control the power of the test if the trial is not terminated. Various properties of this two-stage procedure and derivation of the constants needed for its implementation are presented.

Clinical Trials as Topic↗
PubMed1991

Testing equality of treatment variances in a two-by-two crossover study.

Comparative clinical trials are usually conducted to compare the means associated with the treatments. However, it is often also of interest to compare the variances. Here the problem of testing the equality of the variances associated with two treatments in a two-by-two crossover design is examined.

Analysis of Variance↗
PubMed1991

A confidence interval for the ratio of treatment variances in a two-by-two crossover study.

The distance between two variances is usually measured using the ratio, and a confidence interval for this ratio can be used to measure its magnitude. For a two-by-two crossover study, special considerations must be made because the observations for any subject are correlated. Here a confidence interval for the ratio of the treatment variances in a two-by-two crossover study is presented.

Analysis of Variance↗
PubMed1990

Survival analysis techniques in angina pectoris trials.

The variable 'walking time to moderate angina' on an exercise stress test is the primary means to judge the efficacy of new treatments for angina pectoris. Unfortunately, 'walking time to moderate angina' is often censored by fatigue or other reasons for premature termination of the exercise stress test. If time to fatigue is not treatment-dependent, we propose use of survival analysis techniques in such trials. We present an example from a placebo-controlled multicentre clinical trial and results of simulations that compare various methods of analysis.

Analysis of Variance↗
PubMed1990

A Markovian model for comparing incidences of side effects.

For clinical trials that entail observations at successive visits for the occurrence of a side effect, this paper considers a likelihood-based method to compare side effect incidence rates. The method, which employs the assumption of a Markov chain of order one for the vectors of binary responses, handles missing data due to premature withdrawals. An actual numerical example and a simulated example illustrate the technique.

Clinical Trials as Topic↗
PubMed1989

Sample sizes for constructing confidence intervals and testing hypotheses.

Although estimation and confidence intervals have become popular alternatives to hypothesis testing and p-values, statisticians usually determine sample sizes for randomized clinical trials by controlling the power of a statistical test at an appropriate alternative, even those statisticians who recommend the use of confidence intervals for inference. There is merit in achieving consistency in the techniques for data analysis and sample size determination. To that end, this paper compares sample size determination with use of the length of the confidence interval with that obtained by control of power.

Angina Pectoris↗
PubMed1989

Designing clinical trials for two-sided multiple comparisons with a control.

When a clinical trial is to be conducted to compare more than one experimental treatment to a control treatment, Dunnett's two-sided multiple comparison procedure may be proposed to perform the analysis. During the planning stage, the problem of determining the appropriate sample size must be resolved. Here a solution to this problem is derived by controlling the power of the corresponding testing procedure that assumes that the common variance is known.

Arthritis, Rheumatoid↗
PubMed1988

A one-sided interim analysis with binary outcomes.

Wieand and Therneau (Controlled Clin Trials 8:20-28, 1987) proposed a technique for conducting a clinical trial to test the equality of response rates for two treatments using a one-sided test with an option of terminating the trial at the single interim analysis if it appears that the test treatment will offer no improvement over the control treatment. Chi, Bristol, and Castellana (Stat Med 5:387-392, 1986) considered a clinical trial with the same option when the treatments are compared with respect to the means of variables with normal distributions with a common known variance. Here a decision rule similar to the one proposed in the latter is employed for the problem examined in the former. This decision rule, a generalization of the one proposed by Wieand and Therneau, consists of a one-sided test at the final analysis with a one-sided test at the interim analysis, which is performed in the direction opposite to the one at the final analysis. It is shown that the proposed generalization may result in desirable properties regarding the probability of stopping the trial at the interim analysis in some situations.

Analysis of Variance↗
PubMedDate not supplied

A clinical trial with an interim analysis.

We consider a fixed-sample parallel-group clinical trial with an interim analysis that tests H0:mu x = mu y against H1:mu x less than mu y. If we do not reject at the interim analysis, then the probability of making a type I error by rejecting H0 in favour of H2:mu x greater than mu y and the power at the final analysis are not appreciably affected by performing the interim analysis for certain relevant critical regions.

Biometry↗