PCN457 THE SISAQOL INITIATIVE: ESTABLISHING INTERNATIONAL STANDARDS AND RECOMMENDATIONS FOR THE ANALYSIS OF PATIENT-REPORTED OUTCOMES AND QUALITY OF LIFE DATA IN ONCOLOGY RANDOMIZED CLINICAL TRIALS

Abstract

Patient-reported outcome (PRO) data, such as symptoms and health-related quality of life, are increasingly being captured in cancer randomized clinical trials (RCTs) to provide valuable information on treatment risks, benefits and tolerability. Our literature review showed heterogeneity in analysis, interpretation and reporting of these data in various cancer fields, hindering comparability in results across trials. The Setting International Standards in Analyzing Patient-Reported Outcomes and Quality of Life Endpoints Data (SISAQOL) Consortium was convened to set recommendations for PRO analysis in cancer RCTs. The Consortium comprises 40 international experts including PRO researchers and statisticians, representatives from regulatory bodies, academic societies, pharmaceutical industry, cancer institutes and patient organizations to work on four priorities: (a) specification of well-defined PRO research objectives, (b) recommendations for appropriate statistical PRO analysis methods, (c) standardization of statistical terminology and (d) development of guidelines for analyzing missing data. Recommendations were developed through literature review, surveys, meeting discussions and ratified through voting in a final meeting. A taxonomy of research objectives was developed. Appropriate statistical methods, with the exception of summary measures, were proposed. Consensus was reached on the taxonomy of research objectives and statistical methods, along with a missing data definition and two rates to report missing data occurrence. While some statements concerning handling missing data or statistical analyses are still to be discussed, many statements were ratified for each of the priorities. A robust first set of PRO analyses recommendations were established in a joint process with diverse international stakeholders. Addressing the needs and requirements of these stakeholders provides a strong foundation for widespread endorsement of these recommendations. Ultimately, we expect enhanced interpretability and better impact of PRO data in cancer RCTs from harmonization of current research practices. Disclaimer: This publication reflects solely the views of the individual authors