Are we reporting well enough? A systematic survey of missing data in patient-reported outcomes from non-small cell lung cancer randomized trials.

Abstract

BACKGROUND: Missing patient-reported outcome (PRO) data represent a frequent and significant challenge in clinical trials. Inadequate handling or reporting of such data may introduce bias, undermine the validity of findings, and limit the interpretability of treatment effects, particularly in oncology research where patient experience is a key outcome. PURPOSE: This systematic survey evaluated how missing PRO data were handled in randomized controlled trials (RCTs) involving patients with non-small cell lung cancer (NSCLC). METHODS: A comprehensive search was conducted in PubMed, Embase, Web of Science, Scopus, and ClinicalTrials.gov for NSCLC RCTs reporting PROs (up to June 6, 2024). Two independent reviewers assessed study eligibility and extracted predefined data on the extent of missing PROs and the statistical methods used to address them. Descriptive statistics and logistic regression analyses were performed. RESULTS: Of 13,019 records screened, 252 RCTs met the inclusion criteria. Most trials (70.2%) reported missing PRO data; however, only 45.2% specified statistical handling methods, most commonly complete case analysis, mixed-effects models, or mixed-effects model for repeated measures. Among 52 trials with PROs as primary outcomes, only 12 reported sensitivity analyses. Trials with larger sample sizes or those with PROs as primary endpoints were significantly more likely to report missing data. Clinical events (e.g., death or disease progression) were the predominant reasons for missingness, while procedural or methodological causes were less frequently reported. CONCLUSION: The reporting and statistical management of missing PRO data in NSCLC RCTs remain suboptimal. Most trials relied on conventional methods with limited use of advanced techniques or sensitivity analyses, raising concerns about bias and interpretability. Future trials should prioritize prespecified strategies, robust statistical modeling, and transparent reporting to enhance the validity and clinical relevance of PRO evidence.