Ida Løchting, Margreth Grotle, Kjersti Storheim, Erik L. Werner, Andrew M. Garratt
Communication and Research Unit for Musculoskeletal Disorders (FORMI), , Oslo University Hospital, P.O. Box 4950 Nydalen, 0424 Oslo, Norway. E-mail: ida.lochting@ous-hf.no
DOI: 10.2340/16501977-1826

Objective: To evaluate the reliability and validity of the improved version of the Patient Generated Index (PGI) in patients with low back pain.
Methods: The PGI was administered to 90 patients attending care in 1 of 6 institutions in Norway and evaluated for reliability and validity. The questionnaire was given out to 61 patients for re-test purposes.
Results: The PGI was completed correctly by 80 (88. 9%) patients and, of the 61 patients responding to the re-test, 50 (82. 0%) completed both surveys correctly. PGI scores were approximately normally distributed, with a median of 40 (range 80), where 100 is the best possible quality of life. There were no floor or ceiling effects. The 5 most frequently listed areas affecting quality of life were pain, sleep, stiffness, socializing and housework. The test-retest intraclass correlation coefficient was 0. 73. The smallest detectable changes for individual and group purposes were 32. 8 and 4. 6, respectively. The correlations between PGI scores and other instrument scores followed a priori hypotheses of low to moderate correlations.
Discussion: The PGI has evidence for reliability and validity in Norwegian patients with low back pain at the group level and may be considered for application in intervention studies when a comprehensive evaluation of quality of life is important. However, the smallest detectable change, of approximately 30 points, may be considered too large for individual purposes in clinical applications.