Skip to main content

Table 3 Comparing Predictors of Death using Inverse Probability Weighting and Imputation in adjusted models

From: Treating loss-to-follow-up as a missing data problem: a case study using a longitudinal cohort of HIV-infected patients in Haiti

  Complete Case
OR (95% CI)
IPW
OR (95% CI)
MICE
OR (95% CI)
MICE with deletion
OR (95% CI)
N = 660 N = 698 N = 910 N = 799
Female 0.74 (0.51, 1.08) 0.69 (0.46, 1.03) 0.62 (0.44, 0.87) 0.63 (0.45, 0.88)
Age (per 10 yrs) 1.34 (1.11, 1.60) 1.44 (1.18, 1.76) 1.37 (1.17, 1.61) 1.37 (1.16, 1.61)
Residence 1.08 (0.74, 1.56) 1.38 (0.93, 2.05) 1.18 (0.85, 1.65) 1.18 (0.84, 1.64)
Severe poverty 1.63 (1.12, 2.36) 1.64 (1.10, 2.44) 1.80 (1.28, 2.52) 1.76 (1.25, 2.47)
CD4 (per 100 cells) 0.95 (0.82, 1.09) 0.92 (0.81,1.05) 0.91 (0.80, 1.05) 0.92 (0.80, 1.06)
Baseline Weight (per 10 kg) 0.70 (0.58, 0.85) 0.69 (0.54, 0.87) 0.66 (0.55, 0.78) 0.67 (0.56, 0.80)
WHO stage 1.50 (1.04, 2.17) 1.62 (1.10, 2.40) 1.76 (1.25, 2.47) 1.77 (1.28, 2.47)
Baseline Tuberculosis 1.83 (1.05, 3.20) 1.78 (0.97, 3.26) 1.61 (0.97, 2.67) 1.61 (0.97, 2.68)
  1. p-value≤ 0.05