Low Physical Activity and Reported Barriers Among Survivors of Critical Illness: A Mixed-Method Study

OBJECTIVES: To examine physical activity levels in survivors of critical illness across 6 months, and to understand the barriers and facilitators of physical activity. DESIGN: Prospective, longitudinal, convergent mixed-methods study. SETTING: Functional testing laboratory. PATIENTS: Adults admitted to the ICU for at least 72 hours with acute respiratory failure or sepsis. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Patients wore an accelerometer starting at hospital discharge, at 3- and 6-month follow-up to quantify daily step counts. At 6 months, physical function and patient-reported outcomes were collected in-person, with a subset of participants completing semi-structured interviews. Eighty-two patients were enrolled, with 67 (82%) having physical activity data with median age of 60 years (interquartile range [IQR], 52-66 yr), 54% male, and ICU stay 9.0 days [IQR, 5.8-17.0 d]. The median daily steps were 1358 (IQR, 742-2799) in 50 patients at baseline, 2305 (IQR, 1081-3898) in 54 patients at 3 months, and 2826 (IQR, 1609-4781) in 44 patients at 6 months. Higher comorbidity burden and earlier measurement associated with lower step counts. At 6 months, patients walked a mean of 344 ± 169 meters on the 6-minute walk test and reported being sedentary a median of 300 min/days (150-480 min/d). Eleven participants completed interviews, yielding three themes: 1) "a marathon, not a sprint": the extended ICU recovery trajectory; 2) engagement in physical activity is highly regulated by personal context; 3) value-driven and motivation-informed physical activity decisions. Mixed-methods integration demonstrated that qualitative themes expanded upon quantitative findings to explain nonlinear recovery patterns, contextualizing persistent sedentary behavior, and revealing that motivational and behavioral gaps are not captured by accelerometry. CONCLUSIONS: Survivors showed gradual improvement in daily steps over 6 months yet remained below recommended activity levels. Survivorship and attrition biases may lead to overestimation of step counts, further warranting larger studies. Integrating quantitative and qualitative data reveals modifiable contextual factors that may inform the design of pragmatic physical activity interventions.