D25-08 Relationship Between Pre Critical Illness Walking Speed and Post Critical Illness Disability

Abstract Rationale Pre-critical illness walking speed is associated with mortality in those with Acute Respiratory Distress Syndrome. Its relationship with post-critical illness disability in older survivors of critical illness is unknown. Methods We used data from the National Health and Aging Trends Study (NHATS) linked to Medicare claims data to identify participants with an intensive care unit (ICU) hospitalization. We excluded those without an ICU admission and those with ICU admissions in the initial year of participation who lacked preceding measures of walking speed. As the main exposure, we used assessments of walking speed from the year prior to ICU hospitalization. To define prevalence, we classified low walking speed as 0.8 m/s. We assessed post-critical illness disability in the year following the ICU hospitalization using a 13-item score assessing basic and instrumental activities of daily living (ADL), where higher scores indicate worse disability. We used linear regression to determine the association between pre-critical illness walking speed with post-critical illness disability, adjusting for prespecified covariates including sex, age, a modified frailty phenotype score, baseline ADL score, comorbidities, BMI, and whether participants lived alone. Results We identified 1225 ICU admissions in Rounds 2-10 of NHATS. Participants were a median (interquartile range IQR) age 81 (IQR 76-87) years old, 44% were female, and had a median Charlson comorbidity index score of 5 (IQR 3-7). The analytic cohort of complete data included 852 ICU admissions with similar demographics and baseline characteristics (median age 81 IQR 75-86, 41% female, Charlson comorbidity index score 5 IQR 2-7). In the total cohort, the median pre-critical illness walking speed was 0.66 m/s (IQR 0.49-0.82) and participants had a median of 2 (IQR 1-4) disabilities. The analytic cohort had the same pre-critical illness walking speed (0.66 m/s IQR 0.49-0.83) and disability burden (2 IQR 1-4). The prevalence of low walking speed was 55% (n = 679) in the total cohort and 70% (n = 642) in the analytic cohort. In adjusted linear regression, slower pre-critical illness walking speed was associated with greater disability after critical illness; each 0.1 m/s decrease in walking speed corresponded to a 0.21-unit increase in disability (β = -2.1, 95% CI -2.76 - -1.48, p 0.001). Conclusion In this nationally representative cohort of older Americans, slower pre-critical illness walking speed was associated with greater disability in ADLs among survivors of critical illness. Further research is needed to prospectively identify this at-risk population and develop targeted interventions to mitigate post-critical illness disability. This abstract is funded by: 1T32HS029590-01