Development of a physical performance and mobility examination.

OBJECTIVE: To develop and validate the Physical Performance and Mobility Examination (PPME), an observer-administered, performance-based instrument assessing 6 domains of physical functioning and mobility for hospitalized elderly.

DESIGN: Development of a pass-fail and 3-level scoring system and training manuals for the PPME instrument for use in both clinical and research settings. Two patient samples were used to assess construct validity and interrater reliability of the PPME. A third sample was selected to assess the test-retest reliability of the instrument.

SETTING/PATIENTS: (1) 146 subjects > or = 65 years of age with impaired mobility admitted to Medical Units of Stanford University Hospital. (2) 352 subjects > or = 65 admitted to acute Medical and Surgical Services of the Palo Alto VA Medical Center. Patient samples were obtained during hospitalization and followed until 3 months post-discharge. To study test-retest reliability, 50 additional patients, whose clinical condition was stable, were selected from both settings.

METHODS: An expert panel selected 6 mobility tasks integral to daily life: bed mobility, transfer skills, multiple stands from chair, standing balance, step-up, and ambulation. Tasks were piloted with frail hospitalized subjects for appropriateness and safety. Test-retest and interrater reliability and construct validity were evaluated. Construct validity was tested using the Folstein Mini-Mental State Examination, Activities of Daily Living (ADL), Instrumental Activities of Daily Living (IADL), Geriatric Depression Scale, and modified Medical Outcomes Study Measure of Physical Functioning (MOS-PFR). Two scoring schema were developed for each task: (1) dichotomous pass-fail and (2) 3-level high pass, low pass, and fail. A summary scale was developed for each method of scoring.

MAIN RESULTS: High interrater reliability and intrarater reliability were demonstrated for individual tasks. The mean percent agreement (interrater) for each pass/fail task ranged from 96 to 100% and from 90 to 100% for the 3 pairs of raters for each task using the 3-level scoring. Kappas for individual pairs of raters ranged from .80 to 1.0 for pass-fail scoring and from .75 to 1.0 for 3-level scoring (all P < 0.01). Intraclass correlation coefficients for 3-level scoring by pairs of raters ranged from .66 to 1.0. For summary scales, the mean intraclass correlation was .99 for both scoring schema. Test-retest reliability for summary scales using kappa coefficients was .99 for both pass-fail and 3-level scoring, and .99 and .98, respectively, using Pearson Product Moment Correlation. Correlations of PPME with other instruments (construct validity) suggest that the PPME adds a unique dimension of mobility beyond that measured by self-reported ADLS and physical functioning, and it is not greatly influenced by mood or mental status (r = 0.70 (ADL), r = 0.43 (IADL), r = 0.36 (MMSE), r = 0.71 (MOS-PFR), r = 0.23 (GDS)). The 3-level summary scale was sensitive to the variability in the patient population and exhibited neither ceiling nor floor effects.

CONCLUSIONS: The PPME is a reliable and valid performance-based instrument measuring physical functioning and mobility in hospitalized and frail elderly.