Confounder-adjusted estimates of the risk difference using propensity score-based weighting

Authors

  • Obioha C. Ukoumunne,

    Corresponding author
    1. Clinical Epidemiology and Biostatistics Unit, Murdoch Childrens Research Institute, Flemington Road, Parkville, VIC 3052, Australia
    2. Department of Paediatrics, University of Melbourne, VIC 3010, Australia
    • Clinical Epidemiology and Biostatistics Unit, Murdoch Childrens Research Institute, Flemington Road, Parkville, VIC 3052, Australia
    Search for more papers by this author
  • Elizabeth Williamson,

    1. Clinical Epidemiology and Biostatistics Unit, Murdoch Childrens Research Institute, Flemington Road, Parkville, VIC 3052, Australia
    2. Centre for Molecular, Environmental, Genetic and Analytic (MEGA) Epidemiology, University of Melbourne, VIC 3010, Australia
    3. Department of Epidemiology and Preventive Medicine, Monash University, VIC 3800, Australia
    Search for more papers by this author
  • Andrew B. Forbes,

    1. Department of Epidemiology and Preventive Medicine, Monash University, VIC 3800, Australia
    Search for more papers by this author
  • Martin C. Gulliford,

    1. Department of Public Health Sciences, King's College London, London, U.K.
    Search for more papers by this author
  • John B. Carlin

    1. Clinical Epidemiology and Biostatistics Unit, Murdoch Childrens Research Institute, Flemington Road, Parkville, VIC 3052, Australia
    2. Department of Paediatrics, University of Melbourne, VIC 3010, Australia
    Search for more papers by this author

Abstract

Confounder-adjusted estimates of the risk difference are often difficult to obtain by direct regression adjustment. Estimates can be obtained from a propensity score-based method using inverse probability-of-exposure weights to balance groups defined by exposure status with respect to confounders. Simulation was used to evaluate the performance of this method. The simulation model incorporated a binary confounder and a normally distributed confounder into logistic models of exposure status, and disease status conditional on exposure status. Data were generated for combinations of values of several design parameters, including the odds ratio relating each of the confounders to exposure status, the odds ratio relating each of the confounders to disease status and the total sample size. For most design parameter combinations (474 of 486), the absolute bias in the estimated risk difference was less than 1 percentage point, and it was never greater than 3 percentage points. The confidence interval generally had close to nominal 95 per cent coverage, but was prone to poor coverage levels (as low as 78.5 per cent) when both the confounder-to-exposure and confounder-to-outcome odds ratios were 5, consistent with strong confounding. The simulation results showed that the conditions that are favourable for good performance of the weighting method are: reasonable overlap in the propensity score distributions of the exposed and non-exposed groups and a large sample size. Copyright © 2010 John Wiley & Sons, Ltd.

Ancillary