Quality of Life Studies from the Perspective of an FDA Reviewing Statistician *

Abstract

Statisticians in the Statistical Evaluation and Review Branch of the Division of Biometrics of the Food and Drug Administration (FDA) are observing an increase in the employment of various quality of life measures in both new drug applications (NDAs) and in study protocols. They vary from well conceived studies to studies that appear to have been carried out, and even analyzed, on someone's whim. The emphasis which a statistical review places on quality of life studies as a measure of drug efficacy will depend on several elements: the importance given the results in the sponsor's presentation (primary or secondary efficacy variable), the feelings of the specific medical officer and of his/her division, and guidelines which have been established for certain disease groups. In many NDAs it is not clear how the improvement of quality of life differs from the absence of measurable adverse side effects, which are always carefully scrutinized as a part of the safety of any new drug. Among primary concerns are the reproducibility of the study's results, the obvious multiple comparisons and repeated measures problems, the noise inherent in most of this data, the relevance of the specific questionnaire used, the validity of summary variables, the amount of change (or difference) in scores which would be considered clinically important, and the analysis submitted by the sponsor. The acceptability of quality of life results for use in drug advertising is also a major concern to the agency. Careful design and analysis of quality of life studies using validated measures are necessary for the FDA to consider them valuable in measuring drug efficacy.