Table 2

Performance of the IHBI risk assessment tools (as built in this study) and ADA assessment tool for diabetes on Kuwaiti natives and Asian expatriates

Risk assignment by the ADA tool (%)Risk assignment by the IHBI (k-NN) tool (%)Risk assignment by the IHBI_Aggregate (k-NN) tool (%)
Data setDiabetic patientsNon-diabetic patientsDiabetic patientsNon-diabetic patientsDiabetic patientsNon- diabetic patients (%)
All ethnicities (k=7,N=10632)
 ‘Low’ risk23.416.712.470.76.671.4
 ‘Borderline’ risk32.732.228.420.018.923.7
 ‘High’ risk43.951.159.29.374.54.9
Kuwaiti natives (k=8,N=5513)*
 ‘Low’ risk15.39.711.464.64.964.4
 ‘Borderline’ risk38.144.231.624.43025.5
 ‘High’ risk46.646.157.010.965.210.2
Asians expatriates (k=7, N=3036) *
 ‘Low’ risk23.545.99.673.62.068.8
 ‘Borderline’ risk16.811.314.916.69.619.1
 ‘High’ risk59.642.875.59.888.412.2
  • *Split schema used is (0–1)—‘low’ risk; (2–)—‘borderline’ risk; (4–8)—‘high’ risk.

  • ADA, American Diabetes Association; k-NN, k-nearest neighbours.