Core outcome sets in cancer and their approaches to identifying and selecting patient-reported outcome measures: a systematic review

Abstract

Abstract Objectives Issues arising from a lack of outcome standardisation in health research may be addressed by the use of core outcome sets (COS), which represent agreed-upon recommendations regarding what outcomes should be measured as a minimum in studies of a health condition. This review investigated the scope, outcomes, and development methods of consensus-based COS for cancer, and their approaches and criteria for selecting instruments to assess core patient-reported outcomes (PROs). Methods Studies that used a consensus-driven approach to develop a COS containing PROs, for use in research with cancer populations, were sought via MEDLINE, CINAHL, Embase, Cochrane Library, and grey literature. Results Seventeen studies met the inclusion criteria. Most COS (82%) were specific to a cancer type (prostate, esophageal, head and neck, pancreatic, breast, ovarian, lung, or colorectal) and not specific to an intervention or treatment (76%). Conducting a systematic review was the most common approach to identifying outcomes (88%) and administering a Delphi survey was the most common approach to prioritising outcomes (71%). The included COS contained 90 PROs, of which the most common were physical function, sexual (dys) function, pain, fatigue, and emotional function. Most studies (59%) did not address how to assess the core PROs included in a set, while 7 studies (41%) recommended specific instruments. Their approaches to instrument appraisal and selection varied. Conclusion Efforts to standardise outcome assessment via the development of COS may be undermined by a lack of recommendations on how to measure core PROs. To optimise COS usefulness and adoption, valid and reliable instruments for the assessment of core PROs should be recommended with the aid of resources designed to facilitate this process.