Background Improving the quality of cancer care throughout the disease trajectory is an international priority. Population-based quality indicators (QIs) are key to this process yet remain predominantly used for evaluating care during the early, often curative, stages of disease.

Aim To identify all existing QIs for cancer patients with advanced disease and/or at the end-of-life, and to evaluate each indicator’s measurement properties and appropriateness for use.

Methods We searched five electronic databases from inception to February 2019 for studies describing the development, review and/or testing of QIs for adults with advanced cancer and/or at the end-of-life. For each QI identified we extracted descriptive information (numerator; denominator; benchmarking data; care domain) and assessed six measurement properties (acceptability; evidence-base; definition; feasibility; reliability; validity). Assessments were based on previously established criteria with four possible ratings: positive; intermediate; negative; unknown. Ratings were collated, and each QI classified as either: appropriate for use; inappropriate for use; or, of limited testing. Among the QIs determined appropriate for use we generated a further shortlist by excluding those that were specific to certain patient sub-groups and/or care settings. The shortlist was further reduced by identifying similar/related QIs and retaining the indicator with the highest rating.

Results Our search yielded 7,231 references. Following screening, 35 references met our eligibility criteria and were included in the review. We identified 288 QIs for patients with advanced cancer and/or at the end-of-life. Most evaluated physical aspects of care (n=103, 35.8%) or structure and processes of care (n=109, 37.8%). There was a limited number of QIs relevant to psychosocial (n=18) or spiritual/cultural (n=3) care domains. 27.8% (n=80) of QIs were assessed as appropriate for use; 40.3% (n=116) inappropriate for use, and 31.9% (n=92) had limited testing. Acceptability and validity were the measurement properties with the fewest positive assessments (13.2% and 21.9% respectively). Only 16 QIs (5.6%) reported any benchmarking data. Our shortlist comprised 36 QIs after those specific to patient sub-groups or care settings were excluded. This was further reduced to 15 once duplicate and/or related QIs were removed.

Conclusion Only a small proportion of QIs developed for patients with advanced cancer and/or at the end-of-life have received adequate testing and/or are appropriate for use. Further testing is recommended, particularly with regards to acceptability and validity, as well as research to establish benchmarking data and to expand QIs relevant to psychosocial, cultural and spiritual care domains. To support cancer services in conducting comprehensive and meaningful assessments of quality, we propose 15 QIs, identified from our review as being scientifically sound, applicable across care settings and which collectively evaluate quality across multiple domains of care.