(1219) Patient, Practice, and Organizational Factors Associated With ICU-Based Mobility Performance

Introduction: Numerous barriers and facilitators to implementing mobility interventions have been identified, yet the impact of these factors on mobility performance across multiple ICUs is unknown. In this study, we identified patient, practice, and organizational factors associated with mobility performance and evaluated the impact these factors have in predicting next day early mobility performance.

Methods: Secondary analysis of adult ICU patients from 68 ICUs participating in the ICU Liberation Collaborative between August 2015 and April 2017. Mixed-effects logistic regression models were used to examine the association of patient, ICU practice, and organizational factors with next day mobility performance. Stepwise receiver operating characteristics analysis assessed the accuracy of significant factors on predicting next day mobility performance.

Results: A total of 12,489 adults with at least one 24-hour ICU day were included (37,393 ICU days). Next day mobility occurred in 33.3% of ICU days. Potentially modifiable factors associated with higher odds of next day mobility included significant pain (AOR 1.16; 95% CI, 1.09-1.23), documented sedation target level (AOR 1.09; 95% CI, 1.01-1.18), spontaneous awakening trial performance (AOR 1.77; 95% CI, 1.59-1.96), spontaneous breathing trial performance (AOR 2.35; 95% CI, 2.14-2.58), mobility safety screen performance (AOR 2.26; 95% CI, 2.04-2.49), and prior day physical or occupational therapy (AOR 1.44, 95% CI, 1.30-1.59). Factors associated with lower odds of next day mobility performance included deep sedation (AOR 0.44; 95% CI, 0.39-0.49), delirium (AOR 0.63, 95% CI, 0.59-0.69), benzodiazepines (AOR 0.85; 95%CI, 0.79-0.92), physical restraint use (AOR 0.74; 95% CI, 0.68-0.80), prior day mechanical ventilation (AOR 0.73; 95% CI, 0.68-0.78), and 12 or more pain assessments in the previous day (AOR 0.90; 95% CI, 0.82-0.98). None of the ICU organizational factors were associated with next day mobility performance. Models incorporating patient, practice, and unit variations displayed high discriminant accuracy (AUC 0.85) in predicting next day mobility performance.

Conclusions: Several non-modifiable and modifiable factors influence mobility performance. Collectively, these factors provide excellent prediction of next day mobility performance.