At least 190 randomized clinical breast cancer trials conducted between 2001 and 2009 measured patients' quality-of-life, representing an outcome that, although growing in importance, still only influences clinical decision making in a minority of cases, according to a study published in the forthcoming February issue of the Journal of the National Cancer Institute.
According to the World Health Organization, quality-of-life (QOL), "is a broad-ranging concept affected in a complex way by the person's physical health, psychological state, level of independence..." and other patient-reported outcomes. "[F]or patients and clinicians, the effect of an intervention on QOL is of major interest...[as data] derived from QOL measured in clinical trials can be used to select the optimal intervention, describe a patient's experience, or provide prognostic information," wrote Julie Lemieux, MD, of the University of Quebec in Quebec City, and co-authors.
Despite the growing consideration of patient QOL, Lemieux and colleagues reported that in "breast cancer, patient-reported outcomes are rarely used for approval of a medical product. ... The incremental contribution of QOL measurement to clinical decision making beyond these traditional medical outcomes is unclear." Lemieux and colleagues sought to update their previous QOL review by examining the extent of QOL reporting in clinical trials and to evaluate the effect of QOL on clinical decision making.
The researchers found 190 randomized clinical breast cancer intervention trials on MEDLINE that measured patient-reported, psychosocial QOL indicators. Of these trials, 103 evaluated biomedical interventions and 87 studied nonbiomedical interventions (such as group therapy). Pairs of reviewers studied all articles for specific study variables.
Studies indicating that QOL metrics that influenced clinical decision making tallied 30.1 percent of biomedical trials and 63.2 percent of nonbiomedical trials. These numbers differed from the authors' previous review of clinical trials conducted between 1981 and 2001, in which the authors found 66 clinical trials that evaluated QOL. In this earlier review, 15.2 percent of biomedical trials and 95.0 percent of nonbiomedical trials evaluated QOL.
The authors also discovered a greater variety of interventions in the more recent review, including studies of physical activity. The most common QOL questionnaires administered to participants were the European Organization for Research and Treatment of Cancer QOL Questionnaire and the Functional Assessment of Cancer Therapy/Functional Assessment of Chronic Illness Therapy.
In assessing the quality of the studies, the authors observed no change in the proportion of studies that reported the sample sizes of the trials, the response-rates, the timing of the questionnaires' administrations and the statistical methods that were utilized, which in both reviews exceeded 80 percent.
"Three criteria were not evaluated in the previous review but were assessed in this update: whether power and sample size calculations for QOL outcomes were reported, a description of how missing data were handled, and whether the clinical significance of expected or obtained results was reported," Lemieux and colleagues wrote. "Adherence to all of these criteria was poor, and less than one-third of articles reported on them. It is surprising that sample size calculations were not always reported, even in trials in which QOL was a primary endpoint."
"The large increase in the number of published studies identified in this high-quality systematic review reflects the increased acceptance of the patient's voice in assessing the outcome of trials as well as the participation of expert QOL investigators in the design, conduct, and analysis of clinical trials," commented Patricia A. Ganz, MD, of the Jonsson Comprehensive Cancer Center at the University of California, Los Angeles.
Ganz argued that symptom-focused patient-reported outcomes should play a larger role in the assessment of QOL and its influence on clinical decision making. Moreover, Ganz contended that because many of the studies identified in Lemeiux and co-authors' review were "designed in the mid to late 1990s, they may not reflect the many improvements in study design and data collection that are now standard in clinical trials that incorporate patient-reported outcomes."
Lemieux and co-authors themselves acknowledged that their review probably fell short of