Introduction Manual muscle testing, in the form of the Medical Research Council sumscore (MRC-SS) is a widely accepted clinical tool for diagnosing intensive care unit-acquired weakness (ICU-AW). Although MRC-SS is a simple bedside test, the nature of the test is volitional limiting its ability to distinguish poor motivation and impaired cognition from actual loss of muscle function. The clinical predictive value of the MRC-SS therefore needs to be assessed.

Method Unselected adult ICU patients (=18 years) ventilated for =48 h were eligible. The conscious level of the patients was determined using the Richmond Agitation Sedation Scale; a score −1 to +1 was indicative of awakening. Testing comprised of a two-stage process. Stage 1: Patients at awakening were required to follow 4 simple, one-stage commands. Stage 2: If all 4 one-stage commands were successfully completed, MRC-SS testing was performed by a specialist ICU rehabilitation clinician. ICU-AW was defined as MRC-SS <48. ICU and hospital mortality and length of stay (LOS) were recorded in all the patients.

Results 94 sequential awakening patients were recruited; 68.1% males (n=64), with a mean age for the whole cohort of 64.5±15.3 years. 29 patients were unable to successfully complete the 4 one-stage commands as a result of cognitive impairment. 65 patients completed the MRC-SS of whom 73.9% demonstrated ICU-AW at awakening. Results are shown in Abstract P67 table 1.

Conclusion Almost a third of critically ill patients, from a sequential cohort, were unable to complete 4 one-stage commands and thus could not perform the MRC-SS. Although inability to successfully complete the one-stage commands conferred limited predictive value, those patients that could perform this task were more likely to survive ICU. Similarly an MRC-SS <48 at awakening, presumed indicative of ICU-AW, conferred limited predictive value. However an MRC-SS =48 predicted ICU and hospital survival as well as an ICU LOS <2 weeks. These data highlight the limitations of volitional tests in critically ill patients. It clearly challenges the current view that ICU-AW, as measured by volitional tests, is a predictor of poor outcome. These data confirm that preserved peripheral strength predicts a good outcome.