Skip to main content

Table 5 Simulation results in a domain with ICC = 5%, Pearson correlation X1 and random effect 0.0, number of patients 100, number of centers 5

From: Prediction models for clustered data: comparison of a random intercept and standard regression model

  Apparent performance Test performance
  Standard model Marginal risk calculation Conditional risk calculation Standard model Marginal risk calculation
Harrell’s C-index † 0.82 (0.742; 0.889) 0.82 (0.743; 0.888) 0.84 (0.752; 0.913) 0.77 (0.718; 0.779) 0.77 (0.719; 0.779)
C-index within clusters ‡ 0.82 (0.100) 0.82 (0.096) 0.82 (0.096) 0.77 (0.032) 0.77 (0.032)
Calibration intercept* -0.00 0.00 0.01 0.02 0.06
Calibration intercept* within clusters ‡ -0.00 (0.295) -0.00 (0.319) 0.01 (0.000) -0.01 (0.537) 0.03 (0.545)
Calibration slope (overall) 1.00 0.99 1.05 0.71 0.69
Calibration slope within clusters ‡ 1.06 (0.100) 1.01 (0.108) 1.07 (0.000) 0.74 (0.017) 0.72 (0.018)
  1. * With calibration slope equal to 1 (i.e. calibration in the large).
  2. † overall performance (2.5 and 97.5 percentiles).
  3. ‡ median of overall performance from 100 simulations (median of within cluster performances).