DASC-21: a novel geriatric assessment for discriminating best supportive care in older patients with inoperable...

This study investigated whether the Dementia Assessment Sheet for the Community-based Integrated Care System is useful for decision-making or problem detection in the treatment and care of older patients with inoperable advanced non-small cell lung cancer compared with the current standard model using performance status.

This study retrospectively examined 1595 cases admitted to the Department of Respiratory Medicine at the Tokyo Metropolitan Geriatric Hospital between 26 July 2016 and 28 January 2020. Among these, 29 and 31 patients who received pharmacotherapies and best supportive care were extracted, respectively. The performance in identifying best supportive care using the Dementia Assessment Sheet for the Community-based Integrated Care System was evaluated in comparison with performance status. The ability to detect impairments in each Dementia Assessment Sheet for the Community-based Integrated Care System domain was also assessed.

The Dementia Assessment Sheet for the Community-based Integrated Care System total score had an area under the curve of 0.831 (95% confidence interval, 0.694-0.914), which was statistically equivalent to performance status. The discriminatory cut-off value for identification of best supportive care was set at 29 with a sensitivity and specificity of 0.742 and 0.897, respectively. Dementia Assessment Sheet for the Community-based Integrated Care System total score showed good concordance with performance status especially when reported by family members or caregivers. Deficits other than activities of daily living were recognized (2.8-19.4%) in patients with good performance status. Impairments were more frequently detected when reported by family members or caregivers.

The Dementia Assessment Sheet for the Community-based Integrated Care System discriminates the best supportive care for older patients with inoperable advanced non-small cell lung cancer. Moreover, it can identify vulnerabilities especially when reported by family members or caregivers that cannot be detected by performance status.