OBJECTIVES: Health-related quality of life (HRQOL) is increasingly incorporated into cancer clinical trials as an important endpoint. The notion of minimally important difference (MID) is relevant for interpreting differences and changes in HRQOL scores in terms of clinical meaningfulness. MIDs are useful for assessing the benefit of an intervention (e.g. a new treatment) or designing future clinical trials (e.g., determining sample sizes). We aimed to estimate MIDs for interpreting group-level change of QLQ-C30 scores in melanoma (Mel), breast cancer (BC) and head and neck cancer (HNC) patients.

METHODS: Data are derived from six published Phase III EORTC trials comprising 3,595 Mel, 616 HNC and 448 BC patients. An anchor-based approach was used where anchors were constructed using clinical data e.g. physician examinations and common terminology criteria for adverse events (CTCAE). A panel of clinicians and methodologist provided independent guidance on anchor selection. Changes in QLQ-C30 scores and anchors were evaluated and pooled across all pairwise time-points. MIDs were estimated via mean change method and linear regression.

RESULTS: Performance status (PS) and CTCAEs were sufficiently correlated (r ≥ 0.3) with many of the QLQ-C30 scales while laboratory results yielded no suitable anchors. MIDs varied according to scale, direction (improvement vs deterioration), selected anchor and cancer type. For example, MIDs for the global health status (QL) and social functioning (SF) scales for improvement (deterioration) based on PS as anchor were QL: 8(-9) and SF: 7.7(-7.6) in Mel, QL: 8.2(-4.5) and SF: 6.1(-7.3) in HNC and QL: 12.6 (-7.3) and SF: 6.4(-4.0) in BC.

CONCLUSIONS: Our findings provide further evidence that MIDs differ by direction, across QLQ-C30 scales and by cancer type. This reinforces the evidence that there is no single global standard for clinically meaningful change, and highlights the need to update and diversify the current MID standards.