OBJECTIVES: Interpretation of patient-reported outcome (PRO) measures for oncology efficacy trials requires an understanding of what constitutes meaningful change (minimal clinically important difference or MCID) for patients. We assessed the evidence base for MCIDs for the EQ-5D visual analogue scale (VAS) and the EORTC QLQ-C30 physical functioning (PF) and role functioning (RF) scales for use in gastrointestinal stromal tumor (GIST) efficacy trials. The VAS was selected because it is a patient-centered assessment of overall health. The PF and RF scales were chosen because they characterize GIST patients’ daily functioning.

METHODS: A targeted literature review was initially conducted within PubMed for methodologically-sound MCIDs for the VAS and the PF and RF for gastrointestinal (GI)-related cancers (e.g., GIST, colorectal, colon, rectal, stomach, gastrointestinal, gastric, or intestinal sarcomas or carcinomas). The search around MCID for VAS was subsequently expanded to include all cancers.

RESULTS: No articles were identified that provided MCIDs for the VAS in GI-related cancers. When the review was expanded to all cancers, five articles were identified. One article provided an empirically derived MCID for the VAS (Pickard 2007) based upon cross-sectional anchors. These MCID estimates ranged from 7-8 points for anchor-based and 9-11 for distribution-based estimates. Of the 29 articles meeting search criteria, none provided de novo, empirically-derived MCIDs for the PF and RF scales in GI-related cancers. Where no MCID exists, heuristics may be helpful; the MCID for health-related quality-of-life is often found to be in the range of 0.3 – 0.5 times standard deviation of the baseline value (Norman et al., 2003).

CONCLUSIONS: Even in the most used cancer-specific function scales, the MCID has rarely been established. Where no MCID exists, heuristics can be helpful.