Quality, interpretation and presentation of European Organisation for Research and Treatment of Cancer quality of life questionnaire core 30 data in randomised controlled trials

Abstract

Aim

To review reporting standard, presentation and interpretation for quality of life (QOL) outcomes in randomised controlled trials (RCTs) using the European Organisation for Research and Treatment of Cancer quality of life questionnaire core 30 (EORTC QLQ-C30).

Methods

Cancer RCTs reporting EORTC QLQ-C30 data were identified and reviewed against a reporting quality checklist. Interpretation/presentation methods for QOL data were also recorded.

Results

Eighty-two papers were reviewed. Seventy percent met criteria for high quality reporting; 94% reported mean scores; 84% presented results in tables/graphs; 80% reported p-values or statistical significance. Clinical significance was addressed in 38%. Where clinical significance was not addressed, reliance was usually on statistical significance to interpret the results.

Discussion

EORTC QLQ-C30 results are generally reported well, although it was common to rely on statistical significance alone for interpreting results. Whilst interpretation in terms of clinical significance has improved in recent years, there is still a lack of robust clinical interpretation of QOL results even in papers reported to a high standard.

Introduction

While a considerable body of evidence about health-related quality of life (QOL) is accruing from cancer clinical trials, the extent of its impact on clinical practice is unclear. One of the barriers is poor communication of the clinical relevance of the results.¹ Reviews of prostate cancer² trials, breast cancer trials³ and surgical oncology trials⁴ report that 11%, 33% and 67% were able to inform clinical decision making. Part of this variability in these estimates arose because of differences in how the ability to inform decision making was measured. Regardless of this, it is clear that in order to inform clinical decision making, a QOL study needs to be designed robustly, reported adequately and interpreted appropriately.

The CONSORT statement5, 6 provides a checklist for reporting randomised controlled trials (RCTs). Efficace et al.² proposed a checklist specifically for evaluating QOL outcomes, listing criteria for reporting QOL outcomes and identifying the essential issues to be addressed in order for a trial to have reliable QOL outcomes. The checklist comprises 11 items grouped into four categories: conceptual, measurement, methodology and interpretation. The authors define a paper with high quality QOL outcomes as one that meets at least eight of the 11 criteria and these have to include three high-priority concerns (‘baseline compliance reported’, ‘psychometric properties reported’ and ‘missing data documented’).

Osoba et al.⁷ and Guyatt and Schunemann¹ recommended that the presentation of QOL data include proportions of patients reporting a QOL benefit. They argued that this provides results meaningful to clinicians and therefore the results are more likely to influence clinical decision making. Osoba et al.⁷ recommended 10% of the scale as the cut-off point to define improvement, with a stipulation that this degree of change should persist for a reasonable period. As an additional guide to interpretation Guyatt et al.1, 8 also showed how to generate the number needed to treat for one patient to benefit from therapy.

A number of different approaches have been used to develop interpretation for QOL scores. Some are entirely data driven and some use clinical anchors to interpret differences (over time or between groups). However, there are a number of shortfalls of the current methods and no single method has emerged as a standard for interpretation. Some are not specific to the QOL instrument being used and the validity of these is rarely tested for the specific instrument prior to relying on them for interpretation. It is also common to rely on statistical significance in order to interpret whether differences in scores are clinically significant. However, statistical significance does not necessarily imply a meaningful difference in a clinical context, particularly if the minimum clinically important difference in QOL was not determined a priori and used to determine the sample size for the trial.

Several authors provide a comprehensive overview of existing interpretation strategies.9, 10, 11, 12 Various methods have been used to aid interpretation of the European Organisation for Research and Treatment of Cancer quality of life questionnaire core 30 (EORTC QLQ-C30) specifically, which are of interest for our review. King¹³ used 14 studies to estimate effect sizes (mean difference divided by standard deviation) using clinical anchors, and compared these to Cohen’s guidelines¹⁴ which propose small, medium and large effect sizes are 0.2, 0.5 and 0.8, respectively. King’s estimates were about the same as Cohen’s guidelines for physical, role and symptom scales, although for the global and psychosocial dimensions of the QLQ-C30, the estimates were smaller. Osoba et al.¹⁵ provided estimates for small, moderate and large changes in scores from the EORTC QLQ-C30 based on retrospective global ratings of change, an approach based on individual patients rating the importance of changes in QOL. These were found to be 5–10 for a small difference, 10–20 for a moderate difference and >20 for a large difference, similar to those yielded by King’s analysis.13, 16 This is the basis of Osoba et al.’s¹⁵ recommendation of 10% of the scale as the cut-off point to define a clinically important change.

This review summarises the quality of QOL reporting for cancer RCTs using the EORTC QLQ-C30 and looks at the methods used to present and interpret the QOL data, particularly in papers that score well on the quality checklist. The interpretation methods used across studies are reviewed to assess how widely used current methods are, the extent to which clinical significance is addressed and whether there is a need for additional interpretation guidelines for the EORTC QLQ-C30. A summary of the methods of presenting the data is used to assess whether it is reported clearly and in a way that may be utilised by clinicians.