Introduction The secondary physical effects of critical illness, for example, muscle atrophy, can be detrimental to functional outcome and quality of life in critical care survivors. To minimise these problems early physiotherapy in the Intensive Care Unit (ICU) is advocated. However, research to identify the optimal rehabilitation strategy is hindered by lack of an ICU specific objective measure of physical recovery. Current measures are either impairment specific, thus not capturing the full picture, or their validity is unproven in this population.

Aim To develop a measurement tool to assess physical recovery in the heterogeneous general adult critical care population.

Face/content validity The initial tool was developed in a clinical setting by specialist physiotherapists and is called the Chelsea Critical Care Physical Assessment tool (CPAx). It is a pictorial composite of 10 numerical evaluations of pertinent functions and impairments. It was tested for face and content validity with a content validity index questionnaire (CVIQ), a subsequent focus group and a repeat CVIQ.

Inter-rater reliability Five physiotherapists using the CPAx regularly all assessed the same two patients. Data were analysed for intra-class correlation co-efficient (ICC) and internal consistency.

Construct validity A dual centred cohort study of 32 ICU patients was completed. Treating physiotherapists assessed patients using the CPAx thrice weekly. Additionally, biceps and quadriceps strength, cough peak flow (CPF) and components of the AusTOMs scores (a physiotherapy outcome measure) were tested by a researcher blinded to the CPAx scores. Correlation co-efficients were then used to analyse the data.

Face/content validity Pre and post focus group CVIs were 0.67 (p>0.05) and 0.83 (p<0.05), respectively.

Reliability Reliability was strong with internal consistency of 0.98 and ICC 0.902 (95% CI 0.799 to 0.969; p=<0.001).

Construct validity Pilot data showed strong positive correlations (0.744–0.922, p<0.01), between the CPAx, muscle strength and AusTOMs scores. Data were insufficient to analyse CPF results.

Conclusion Preliminary data show that the CPAx is a reliable and valid measure of physical function in ICU patients. It can now be used as an outcome measure for physiotherapy research and has potential to act as a predictor of functional outcome in the ICU population.