Evaluating the quality of supportive oncology using patient-reported data.

Purpose: Patient-reported data may provide valuable information for oncology quality measurement. We evaluated applying patient-reported data from the Cancer Care Outcomes Research and Surveillance Consortium national study to quality indicators addressing supportive oncology domains.

Methods: We analyzed surveys including validated instruments for symptoms and care planning from 4,174 patients newly diagnosed with lung or colorectal cancer. We adapted existing supportive oncology indicators for use with patient-reported data from Cancer Quality-ASSIST and the American Society of Clinical Oncology Quality Oncology Practice Initiative. Where feasible, we examined indicators using medical record data for comparison purposes.

Results: Using patient-reported data, the percentage of patients meeting indicator criteria for screening ranged from 74% (significant depressive symptoms) to 93% (pain, nausea/vomiting). The percentage meeting indicator criteria for symptom treatment ranged from 73% (significant depressive symptoms) to 99% (nausea/vomiting). Symptom severity did not affect the results. Using medical record data, the percentage meeting indicator criteria varied between 4% (significant depressive symptoms) and 23% (pain). For information and care planning, 44% met criteria for discussion about resuscitation and 32% for hospice discussion using survey data. Using medical record data, 11% met criteria for goals-of-care discussion in intensive care and 46% for hospice/pain management/palliative care referral before death.

Conclusion: Quality of care for symptoms measured using patient self-report was higher than when including medical record data. Use of patient-reported and medical record data allowed measurement of different perspectives on care planning indicators. Patient-reported data provide a complementary, patient-centered perspective on the quality of supportive oncology care.