Prolonged Mechanical Ventilation Weaning at Long Term Acute Care Hospitals: Mobilization and Outcomes

Rationale: Numbers of patients who require prolonged mechanical ventilation (PMV) is growing. Few studies examine the effectiveness of active mobilization therapies on the health outcomes of patients on PMV in the long-term acute care setting. The effect of physical therapy assisted mobility of patients on PMV at Long Term Acute Care Hospitals (LTACHs) and the effect on clinically significant outcomes is unknown. Furthermore, distinct subgroups of PMV patients at LTACH’s have not been described; it is unknown how subgroups of PMV patients at LTACH’s differ in clinical outcomes, including mobilization. The purpose of this doctoral dissertation research is threefold. First, examine the relationship between three specific physical therapy assisted mobilization activities of bedside dangling, stand-turn-pivot to an out-of-bed chair, and ambulation, on ventilator liberation, mortality, discharge disposition of patients who require PMV at a Midwestern LTACH. Second, identify and describe distinct subgroups of patients on PMV at a Midwestern LTACH as identified by grouped clinical indicators present at the time of LTACH admission. Finally, analyze subgroup differences in mortality, ventilator liberation, discharge disposition, and mobilization throughout the LTACH hospitalization. Methods: Retrospective, medical record review of patients on PMV from a Midwestern LTACH admitted between January 1, 2006 through December 31, 2015. Measures included: baseline demographic data; The Charlson Co-Morbidity Index Score; ventilator wean status; mortality status; and LTACH length of stay (LOS). Physical therapy assisted mobilization was defined as: edge of bed dangle; stand-turn-pivot to a bedside chair; ambulation with, or without, an assist device such as a walker. Logistic regression models were used to assess the relationship between the weekly average number of physical therapy assisted active mobilization activities (count/LTACH LOS*7) on ventilator liberation and mortality. Latent class analysis was performed to analyze the data for subgroup analysis. Complete data were available for analysis of 249 PMV patients with an average age of 68.6 (± 14.0); 122 (49%) were male, and 127 were female (51%). ANOVA and chi-square were used in post hoc analysis of the subgroups. Results: The overall rate of successful liberation from mechanical ventilation was 172 (69.1%), and 62(24.9%) of the patients died during hospitalization. Controlling for age, Charlson Co-Morbidity total score, and LTACH LOS, the odds of dying decreased as the amount of weekly chair sitting and ambulation increased. Dangling did not have a statistically significant relationship with mortality. Controlling for age, Charlson Co-Morbidity total score, and LTACH LOS, the odds of successful ventilator liberation increased as the amount of weekly mobilization increased, which was independent of activity type. A three-class subgroup solution was identified, differentiated by physiologic and comorbid burden, containing the youngest, obese patients with low levels of physiologic and co-morbid burden (Class 1, n=73); the oldest patients with low levels of physiologic burden but multiple co-morbid conditions (Class 2, n=105); and older patients with high levels of multiple physiologic and co-morbid burden (Class 3, n=71)). Analysis of variance showed no difference amongst the subgroups in LTACH length of stay [F(2,246)=2.243, p= 0.108] and number of ventilator days [F(2,246)=0.641, p= 0.528]. Patients in Class 3 were less likely to wean from mechanical ventilation [χ2(2, N= 249) = 25.48, p< 0.001] and more likely to die [χ2(2, N= 249) = 23.68, p< 0.001] than those patients in Class 1or Class 2. Conclusion: There is an association between physical therapy assisted active mobilization therapies on the probability of ventilator liberation and survival for patients on PMV at LTACHs. Infrequent mobilization is a risk factor related to poor clinical outcomes. Additionally, PMV patients can be described by distinct profiles which are independent of clinical outcome. This analysis has implications for the screening and monitoring of Class 3 patients, those with high levels of physiologic burden and multiple co-morbid conditions. This patient subtype is at high risk for low mobilization, poor clinical outcome, and death.