BMC Medical Research Methodology (Aug 2025)

A scoping review of statistical methods used to report EORTC QLQ-C30 quality of life scores measured longitudinally

  • Rosie A. Harris,
  • Chris A. Rogers,
  • Jessica Harris,
  • Eric Lim

DOI
https://doi.org/10.1186/s12874-025-02622-1
Journal volume & issue
Vol. 25, no. 1
pp. 1 – 15

Abstract

Read online

Abstract Background The European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (QLQ-C30) is often used in cancer studies to assess patient reported quality of life over time. A number of challenges are faced when analysing scores derived from this questionnaire, and a multitude of statistical methods can be used. Whilst methods exist to overcome issues such as non-independence of repeated measurements and informative dropout, it is unclear how often these are implemented in practice. The aim of this scoping review was to comprehensively describe the statistical methods used to analyse longitudinal quality of life scores derived from the QLQ-C30 questionnaire. Methods Two databases, MEDLINE and Embase, were searched for randomised controlled trials and prospective observational studies presenting statistical analyses of QLQ-C30 scores collected over time. Only studies published in English between 1 January 2021 and 31 December 2022 were included. Studies not reporting the statistical methods used were excluded. A REDCap database was developed to store extracted information and results are presented as descriptive summaries. Results Two-hundred and seventy-one eligible studies were identified including 161 parallel group randomised controlled trials and 84 cohort studies. A linear mixed effects model was the most utilized analysis method, applied in 121/271 (45%) studies, followed by time to event analyses (54/271, 20%) and t-tests (44/271, 16%). Nearly one third of studies did not apply any longitudinal analysis method (82/271, 30%). Missing data due to death was accounted for in 23/271 (8%) studies. Statistical approaches did not differ greatly depending on the design of the study. Conclusions There is no current consensus on the statistical method to analyse repeated QLQ-C30 scores. Many studies continue to ignore the longitudinal structure of repeated measures data, which could impact the interpretation of the data and conclusions drawn.

Keywords