Neuromuscular electrical stimulation (NMES) may reduce intensive care unit – acquired weakness (ICU-aw). However, current protocols and patient’s responses remain heterogeneous. Identifying the determinants of muscle responsiveness could help optimize NMES strategies in patients at risk of developing ICU-aw.

This study is a secondary analysis of DETECT study (NCT04722445). On the first day deemed feasible, each critically ill patient receiving mechanical ventilation underwent quadriceps electrical stimulation aimed at eliciting a knee extension with heel lifted, concomitantly with ultrasound assessment of the quadriceps muscle. US quadriceps images were collected to analyse pennation angle, thickness and echogenicity of the rectus femoris (RF) muscle and skin to RF distance. We also analysed all anthropometric variables and ICU pathway determinants that could impact quadriceps responsiveness.

Among the 75 patients included in the analysis, 29 (39%) were classified as NMES non-responders (maximum stimulation intensity reached without achieving knee extension, with heel lifted: 120 mA). The 46 NMES responders (61%) required a median stimulation intensity of 72.8 [61.5-91] mA. NMES non-responders had greater RF thickness (1.5 (0.6) vs. 1.2 (0.4) cm; p = 0.01), higher BMI (30 [25.5-32.1] vs. 24.8 [21.9-29.5] kg/m²; p = 0.01), higher fluid accumulation (6.5 [2–8] vs. 1 [-1.5-2.5] kg; p < 0.001) and SOFA scores at admission (7 [6–10] vs. 6 [4–7]; p = 0.01). In the NMES non-responders, a larger proportion were diabetic than in the NMES responders (20.7% vs. 4.4%; p = 0.03) and the sepsis rate was also higher (34.5% vs 6.7%; p = 0.002). Aminoglycoside exposure (41.4% vs. 4.4%; p < 0.001) and vasopressor dose (0.2 [0-0.4] vs. 0 [0-0.2] µg/kg/min; p = 0.05) were also higher in non-responders. After multivariable analysis, only fluid accumulation remained associated with NMES non-responsiveness (OR 1.30, 95% CI 1.04–1.63; p = 0.02).

Fluid accumulation could decrease quadriceps excitability during NMES in critically ill patients. RF thickness, BMI, diabetes, illness severity, sepsis, exposure to aminoglycosides, and vasopressor doses could also decrease quadriceps responsiveness but further research is needed to determine the actual impact of these factors.