Quality-of-life outcomes in the evaluation of head and neck cancer treatments.

OBJECTIVES To review the published literature to evaluate the design, use of terminology, and interpretation of results in studies using quality-of-life (QOL) instruments to measure differences between head and neck cancer treatments at a point in time or to report changes over time in one or more treatment groups. DATA SOURCE MEDLINE search for subject headings "head and neck neoplasms" (as a main topic) and "quality of life" or "health status" restricted to English-language sources and a 10-year period from 1989 to 1999. STUDY SELECTION Four hundred forty-five abstracts were reviewed to find articles using an instrument to compare head and neck cancer therapy groups with a QOL outcome (13.7% included). DATA EXTRACTION Two readers reviewed each article to determine how terminology was used, if a scientific study design was used, and if differences or changes in scores were clinically interpreted. RESULTS Sixty-one articles were reviewed. Forty different instruments were used. Terminology was used inconsistently in 21 (34.4%) of the 61 articles. A scientific study design was used in only 11 (18.0%) of the 61 articles (P<.001). A clinical interpretation of results was given in 16 (26.2%) of the 61 articles (P<.001). CONCLUSIONS While QOL outcomes show promise for assisting with treatment decisions in head and neck cancer therapy, few studies using instruments to measure QOL outcomes are hypothesis driven and clinical interpretations of results are not commonly provided. We recommend that future studies identify the construct to be measured, specify comparator groups and hypotheses a priori, and provide clinical interpretations of results.