The interobserver agreement of handheld dynamometry for muscle strength assessment in critically ill patients.

OBJECTIVE: Muscle weakness often complicates critical illness and is associated with increased risk of morbidity, mortality, and limiting functional outcome even years later. To assess the presence of muscle weakness and to examine the effects of interventions, objective and reliable muscle strength measurements are required. The first objective of this study is to determine interobserver reliability of handheld dynamometry. Secondary objectives are to quantify muscle weakness, to evaluate distribution of muscle weakness, and to evaluate gender-related differences in muscle strength.
DESIGN: Cross-sectional observational study.
SETTING: The surgical and medical intensive care units of a large, tertiary referral, university hospital. PATIENTS: A cross-sectional, randomly selected sample of awake and cooperative critically ill patients.
INTERVENTIONS: None.
MEASUREMENTS AND MAIN RESULTS: Handheld dynamometry was performed in critically ill patients who had at least a score of 3 (movement against gravity) on the Medical Research Council scale. Three upper limb and three lower limb muscle groups were tested at the right-hand side. Patients were tested twice daily by two independent raters. Fifty-one test-retests were performed in 39 critically ill patients. Handheld dynamometry demonstrated good interobserver agreement with intraclass correlation coefficients >0.90 in four of the muscle groups tested (range, 0.91-0.96) and somewhat less for hip flexion (intraclass correlation coefficient, 0.80) and ankle dorsiflexion (intraclass correlation coefficient, 0.76). Limb muscle strength was considerably reduced in all muscle groups as shown by the median z-score (range, -1.08 to -3.48 sd units). Elbow flexors, knee extensors, and ankle dorsiflexors were the most affected muscle groups. Loss of muscle strength was comparable between men and women.
CONCLUSIONS: Handheld dynamometry is a tool with a very good interobserver reliability to assess limb muscle strength in awake and cooperative critically ill patients. Future studies should focus on the sensitivity of handheld dynamometry in longitudinal studies to evaluate predictive values toward patients' functional outcome.