Clinical judgment versus geriatric assessment for frailty in older patients with cancer.

BACKGROUND: Geriatric assessment (GA) is an appropriate method for identifying frailty in older patients with cancer, but a shorter instrument may be easier to use in clinical practice. Clinical judgment is always available and requires no investments in time or resources. The purpose of this study was to assess correlations between clinical judgment for frailty of the cancer specialist, the general practitioner and patient's self-assessment, and the correlation between clinical judgment and GA. METHODS: This was a dual-center inception cohort study of patients with cancer aged ≥70 years starting curative or first-line palliative chemotherapy. GA included the following domains: (instrumental) activities of daily living, nutrition, mobility, cognition, mood, and polypharmacy. Clinical judgment for frailty was rated on a scale from 0 to 10 (0 = not frail, 10 = frail). Correlation was tested using Kendall's tau-b correlation coefficient. RESULTS: Of all 55 patients, 76% had ≥2 geriatric impairments. Median clinical judgment frailty score was 3 (range 1-10 for cancer specialist and patient and range 0-10 for general practitioner) and did not vary much according to the number of impaired geriatric domains (ranging from 2 for 0-1 impaired domains to 4 for ≥3 impaired domains). Correlations between mutual clinical judgment scores and between clinical judgment and GA were negligible or low. CONCLUSION: Correlations between clinical judgment scores and between clinical judgment and GA were poor. Most patients with multiple geriatric impairments had low 'subjective' frailty scores. Other frailty assessments, such as frailty screening tools or GA, should be considered in addition to clinical judgment when selecting older patients for potential treatment with chemotherapy.