Addressing Ceiling Effects in Health Status Measures: A Comparison of Techniques Applied to Measures for People with HIV Disease

Authors

  • I-Chan Huang,

    1. Department of Epidemiology and Health Policy Research, College of Medicine, University of Florida, Gainesville, FL,
    Search for more papers by this author
  • Constantine Frangakis,

    1. Department of Biostatistics, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD,
    Search for more papers by this author
  • Mark J. Atkinson,

    1. Health Outcomes Research Center, School of Medicine, University of California, San Diego, CA,
    Search for more papers by this author
  • Richard J. Willke,

    1. Pfizer Inc., Bridgewater, NJ
    Search for more papers by this author
  • Walter L. Leite,

    1. Department of Educational Psychology, College of Education, University of Florida, Gainesville, FL
    Search for more papers by this author
  • W. Bruce Vogel,

    1. Department of Epidemiology and Health Policy Research, College of Medicine, University of Florida, Gainesville, FL,
    Search for more papers by this author
  • Albert W. Wu

    1. Department of Health Policy and Management, Bloomberg School of Public Health, Johns Hopkins University, 624 North Broadway, Baltimore, MD 21205-1901,
    Search for more papers by this author
    • Address correspondence to Albert W. Wu, M.D., M.P.H., Department of Health Policy and Management, Bloomberg School of Public Health, Johns Hopkins University, 624 North Broadway, Baltimore, MD 21205-1901. I-Chan Huang, Ph.D., is with the Department of Epidemiology and Health Policy Research, College of Medicine, University of Florida, Gainesville, FL. Constantine Frangakis, Ph.D., is with the Department of Biostatistics, Bloomberg School of Public Health, Johns Hopkins University, Baltimore, MD. Mark J. Atkinson, Ph.D., is with the Health Outcomes Research Center, School of Medicine, University of California, San Diego, CA. Richard J. Willke, Ph.D., is with Pfizer Inc., Bridgewater, NJ. Walter L. Leite, Ph.D., is with the Department of Educational Psychology, College of Education, University of Florida, Gainesville, FL. W. Bruce Vogel, Ph.D., is with the Department of Epidemiology and Health Policy Research, College of Medicine, University of Florida, Gainesville, FL.


Abstract

Objectives. To compare different approaches to address ceiling effects when predicting EQ-5D index scores from the 10 subscales of the MOS-HIV Health Survey.

Study Design. Data were collected from an HIV treatment trial. Statistical methods included ordinary least squares (OLS) regression, the censored least absolute deviations (CLAD) approach, a standard two-part model (TPM), a TPM with a log-transformed EQ-5D index, and a latent class model (LCM). Predictive accuracy was evaluated using percentage of absolute error (R1) and squared error (R2) predicted by statistical methods.

Findings. A TPM with a log-transformed EQ-5D index performed best on R1; a LCM performed best on R2. In contrast, the CLAD was worst. Performance of the OLS and a standard TPM were intermediate. Values for R1 ranged from 0.33 (CLAD) to 0.42 (TPM-L); R2 ranged from 0.37 (CLAD) to 0.53 (LCM).

Conclusions. The LCM and TPM with a log-transformed dependent variable are superior to other approaches in handling data with ceiling effects.

Ancillary