Validity of the Karnofsky Performance Status in an HIV-infected sample

The Karnofsky Performance Status (KPS) is the most widely used health status measure in human immunodeficiency virus (HIV) medicine and research. Because there are limited data on KPS metric properties in this setting, we present evidence of the construct validity of the KPS in a sample of HIV-infected persons using data from the AIDS Time-Oriented Health Outcome Study (ATHOS). The sample consisted of 160 primarily white, homosexual/bisexual men with a mean age of 45 years and a mean KPS score of 82 (range, 40-100). Ninety percent were classified in Centers for Disease Control (CDC) clinical category B or C. Pearson's product-moment correlations were strong between the KPS and measures of global health status, physical disability, numbers of symptoms, CDC clinical category, social function, days off work, and energy/fatigue (r = 0.39-0.52, p < 0.0001). Correlations with measures of mental health and cognition were less impressive but statistically significant. Analysis of variance followed by Student Newman-Keuls test showed significant differences among three KPS groupings for global health status but not for physical disability. Regression analysis indicated three significant variables accounting for KPS variance: visual analogue global health status (27%), days off work (10%), and energy/fatigue (1.7%). We conclude that the upper range of the KPS reflects global health status better than physical performance and much better than psychosocial constructs, in persons with HIV infection. Further research examining the ability of the KPS to detect clinically significant change over time is needed.