Statistical Considerations for Performing Multiple Tests in a Single Experiment. 2. Comparisons Among Several Therapies

The aim of this study was to evaluate to what extent the quality of life (QOL) of young patients with chronic myelocytic leukemia (CML) was affected by treatment with interferon (IF) and intensive chemotherapy. In a main study performed by The Swedish CML Group, aiming at reduction of the malignant pH⁺ cell clone by treatment with hydroxyurea and IF followed by ABMT, QOL was evaluated with VAS scales and the Life Ingredient Profile in 44% of the patients. The intensive treatment did not lead to intolerable suffering or protracted reduction in QOL. However, 80% of the patients were on sick leave during the first year of treatment.

During a controlled clinical trial, data accumulate that contain information on the relative efficacy of the two treatments, yet these data often are not inspected or analyzed until the planned sample size has been reached and the trial has been terminated. This type of fixed-sample-size trial design, with a single terminal data analysis, has the ethical disadvantage that more patients than are necessary to obtain a reliable result may be randomized to the less efficacious treatment, whichever one that turns out to be. Thus, it often is desirable to schedule one or more analyses of the data, to be conducted before planned termination, to see if a reliable conclusion may be drawn from the data and the trial terminated early. Such analyses are called interim analyses. When interim analyses occur after each of several relatively large groups of patients, the trial is called a group-sequential trial. Interim data analyses must be planned in advance to avoid increasing the risk of committing a Type I error and to achieve adequate power. This article introduces the statistical issues involved in the planning of interim data analyses and the design of group-sequential clinical trials.

Microscopic colitis and collagenous colitis are similar conditions that are differentiated by the presence or absence of subepithelial collagen table thickening. To better understand the relationship between these two disorders and the role of collagen table thickening in the pathogenesis of diarrhea, colonic mucosal biopsy specimens from 24 patients with microscopic or collagenous colitis and 9 control subjects were analyzed using a computerassisted morphometric method to evaluate the average thickness of the subepithelial collagen table. The collagen table thickness in colitis patients taken together formed a multimodal rather than a unimodal distribution. There was no tendency for collagen table thickening to increase with age or with duration of symptoms. In general, the types and distribution of inflammatory cells were similar in patients with normal and thickened collagen tables. Stool weight correlated with lamina propria cellularity but not with collagen table thickening. The multimodal distribution of collagen table thickening and the lack of correlation with age, duration of symptoms, or inflammation suggest that microscopic colitis and collagenous colitis are discrete conditions, although the inflammatory changes in the two conditions are similar. Moreover, because stool weight correlates with lamina propria cellularity but not with collagen table thickening, diarrhea probably is caused by the inflammatory changes and not by collagen table thickening per se.

Although primarily used as a lipid lowering drug, probucol also possesses anti-oxidant activity and has been shown in animal models to inhibit or delay the progression of atherosclerosis. It has been suggested that this anti-atherosclerosic effect may occur through inhibition of free radical oxidation of low density lipoprotein. The aim of this study was to investigate the effects of probucol on free radical activity in hyperlipidaemic patients. Plasma lipid peroxides were measured before probucol treatment, at 4 and 12 weeks treatment and then 4 weeks after stopping probucol. Lipid peroxide concentrations were significantly reduced during and 4 weeks after stopping treatment with probucol, when compared with baseline values. There were no changes in plasma vitamin E concentrations. The results of this study indicate that probucol reduces lipid peroxidation in patients, an effect which may occur through a free radical scavenging action.

An understanding of statistical concepts and methods is essential for the clinician who wishes to interpret the results of clinical studies. In this article the concepts of descriptive statistics, classical hypothesis testing, P values, a priori information, and Type I and Type II errors are discussed with examples to illustrate their application to the interpretation of clinical trials. In addition, descriptions of Student's t test, the chi-squared test, Fisher's exact test, the rank sum test, and sequential methods are given.