Comparison of individual and composite radiographic markers of frailty in trauma

Background: Clinical frailty scores usually involve questionnaires or physical testing. Many trauma patients are not able to participate in these. Radiographic measurement of frailty may be a viable alternative. Individual radiographic markers of frailty have been investigated, such as sarcopenia or osteopenia. The ideal radiographic variable (or variables) to measure frailty in trauma is unknown. Study design: A retrospective review was performed of restrained drivers ages 40 and greater at a single institution from 2010-2015. Multiple markers of radiographic frailty were measured including: sarcopenia, osteopenia, vascular calcifications, sarcopenic obesity, emphysema, renal volume, cervical spine degeneration, and cerebral atrophy. Frailty was defined as the worst quartile for each radiographic variable, and these values were summed to create a composite marker of frailty. The primary outcome was discharge disposition. We hypothesized that a composite frailty score would be associated with discharge disposition while individual markers would not be associated with discharge disposition. Results: Overall 489 patients were included in this study. Cerebral atrophy (p = 0.05), renal volume (p = 0.004), sarcopenia (p = 0.05), vascular calcifications (p = 0.02) and sarcopenic obesity (p = 0.01) were associated with discharge disposition. Pearson's correlation coefficients between radiographic frailty markers were all less than 0.4. Youden's Index was 0.26 (p < 0.001) at a composite score of 3. In multivariable analysis, the composite score of 3 or greater was associated with poor discharge disposition (OR 2.39, 95% CI 1.10–5.18, p = 0.03). Conclusions: Individual radiographic frailty markers are inadequate markers of frailty, as they may miss patients who are frail. This study also suggests that a composite radiographic frailty score may better predict patient outcome than individual radiographic markers of frailty.