A global measure of physical functioning: Psychometric properties

Objective. To examine the psychometric properties of a global physical functioning scale (GPFS) developed as a self-report measure and constructed to scale physical functioning from very poor (1) to excellent (100). Data Sources. Data collection took place between January 1997 and September 1999. It consisted of self-ratings of surgical patients and the ratings of clinicians. The setting was the surgical department at a university hospital. Study Design. Test-retest reliability and the convergence of the scores of patients and clinicians were examined in 106 patients before elective coronary or gastrointestinal surgery. Inter-rater reliability was tested in 36 hospitalized patients with cardiologic or vascular surgical diseases who were rated by random selection from a pool of 91 clinicians. The patients also rated their physical functioning. Discriminative validity, sensitivity to change, ceiling and floor effects, and influence of emotional state upon the scores were tested in 127 patients in six diagnostic groups who scored the GPFS before and subsequent to surgery, The concurrent validity was examined in 101 patients who scored the GPFS and the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36) before elective coronary surgery. Principal Findings. The test-retest correlation (.90), correlation of the scores of the clinicians and patients (.87), and rater intraclass correlation coefficient (.82) were high. The GPFS discriminated among patients with different levels of physical functioning, and it was sensitive to change following coronary surgery. There were moderate ceiling and no floor effects. The correlation with the physical functioning scale of the SF-36 (PF-10) was .67. The GPFS differentiated patients with middle levels of physical functioning better than did the PF-10. Conclusions. The psychometric properties of the GPFS appeared adequate as a measure of general physical functioning. The scale is easy to use and also appears suitable for outcome studies following substantial changes in physical functioning as after coronary surgery.