Measuring quality of life in rural Uganda: reliability and validity of summary scores from the Medical Outcomes Study HIV Health Survey (MOS-HIV)

Abstract

Summary scores derived from the medical outcomes study HIV health survey (MOS-HIV) are used to assess treatment impacts among HIV-infected patients in Western settings, but have yet to be validated in rural, African settings. We examined the reliability, validity and responsiveness of scores among a prospective cohort of 947 HIV-1-infected adults initiating antiretroviral therapy between May 2003 and May 2004 in rural Uganda. Physical (PHS) and mental health (MHS) summary scores were developed from baseline MOS-HIV sub-domains using exploratory factor analysis. Construct and discriminant validity were established by comparing mean summary scores across known groups of sociodemographic, clinical and health status characteristics. Effect sizes were calculated to assess responsiveness to therapy. Reliability of the PHS and MHS scores was 0.79 and 0.85, respectively. Mean baseline PHS and MHS scores varied significantly by CD4 cell count, HIV viral load, WHO stage of disease and Karnofsky performance status scores. By 12 months on antiretroviral therapy, PHS and MHS scores improved by 14.6 points (P < 0.001) and 13.9 points (P < 0.001), respectively. PHS and MHS scores can be derived from the MOS-HIV and used to assess health status among cohorts of patients taking antiretroviral therapy in rural Uganda.