Mobilizing smarter, not harder: optimizing mobilization in critical illness

Intensive care unit (ICU)–acquired weakness affects almost 50% of mechanically ventilated patients and is associated with higher mortality.1,2 It can contribute to prolonged physical disability long after hospital discharge. Many ICU survivors experience persistent muscle weakness, reduced mobility, and difficulty returning to work or daily activities, with profound effects on independence and on the caregivers who support them.3,4 These long-term limitations also place substantial demands on the health care system through delayed recovery, increased rehabilitation needs, and recurrent hospitalizations. Given the scale of this problem and its far-reaching consequences, mobilization has emerged as a central strategic intervention to counter ICU-acquired weakness and support recovery in critical illness. The TEAM trial was an important milestone for ICU mobilization.5 Conducted in 750 patients across 49 ICUs in 6 countries, it compared an early high-intensity active mobilization strategy with usual care mobilization in adults receiving mechanical ventilation. The intervention aimed to achieve the highest level of activity possible for the longest safe duration each day through structured, physiotherapist-led sessions. On average, the intervention group received 21 minutes and the control group received 9 minutes of active mobilization. The trial did not show improvement in the primary outcome of days alive and out of hospital at day 180, and it raised concerns for higher adverse events in the intervention group. Regardless, the trial made one message clear. Mobilizing patients earlier and for longer does not guarantee better outcomes and may expose vulnerable patients to harm. In this issue of the Journal, Serpa Neto and colleagues6 extend the insights from the TEAM trial using a target trial emulation of the same dataset (Figure 1A). Target trial emulation is a causal inference approach that imagines the ideal randomized trial we would conduct if possible, then uses observed longitudinal data to estimate outcomes for each prespecified strategy.7 In plain terms, it allowed researchers to simulate how patients might fare under different mobilization prescriptions while accounting for daily changes in illness severity, sedation, and readiness for activity. It is particularly useful when the actual delivered mobilization dose varies widely, as in both arms of the TEAM trial. Overview of the target trial emulation and considerations for future intensive care unit (ICU) rehabilitation trials. (A) Schematic representation of the target trial emulation based on the TEAM dataset. The emulation compared active mobilization time and intensity, defined using the ICU Mobility Scale, across simulated mobilization strategies from ICU admission until ICU discharge or day 28. Baseline characteristics and daily time-varying clinical variables were incorporated to estimate the effect of each strategy on the primary outcome of 28-day mortality, with sensitivity analyses exploring alternative timing of mobilization initiation and selected subgroups. (B) Conceptual framework outlining key considerations for future ICU rehabilitation trials. In addition to traditional baseline variables, future trials could incorporate measures of pre-ICU physical function and physical activity; define mobilization dose through frequency, intensity, type, and timing of initiation; and extend rehabilitation efforts beyond the ICU to hospital discharge and follow-up. Primary outcomes emphasizing patient-reported or performance-based functional recovery, with longitudinal follow-up to capture the durability of treatment effects, will advance the field further. The authors evaluated 25 combinations of mobilization duration and intensity (defined by the ICU Mobility Scale [IMS]).8,9 The emulation showed that in patients with IMS