Table 2. Performance of risk scores in cross-sectional validation for the screening of undiagnosed diabetes.
| Risk score | Cutoff | Patients at high risk (%) | AROC (95% CI) | Sensitivity (%) | Specificity (%) | Youden Index | PPV (%) | NPV (%) | P value* |
|---|---|---|---|---|---|---|---|---|---|
| Korean Risk Score (KRS) [13] | ≥6 | 43 | 0.754 (0.740–0.769) | 81 | 58 | 40 | 5.8 | 99 | N/A |
| - | ≥5 | 61 | - | 91 | 40 | 32 | 4.6 | 99 | N/A |
| Australian score (AUSDRISK study) [16] | ≥13 | 29 | 0.782 (0.769–0.796) | 70 | 72 | 43 | 7.4 | 99 | <0.001 |
| - | ≥12 | 35 | - | 75 | 66 | 41 | 6.6 | 99 | - |
| Danish score [17] | ≥25 | 36 | 0.777 (0.763–0.792) | 77 | 65 | 43 | 6.6 | 99 | 0.001 |
| - | ≥31 | 20 | - | 55 | 81 | 36 | 8.4 | 98 | - |
| ADA questionnaire [18] | ≥3 | 36 | 0.776 (0.762–0.790) | 78 | 65 | 44 | 6.6 | 99 | 0.004 |
| - | ≥5 | 6 | - | 21 | 95 | 16 | 11.5 | 97 | - |
| Japanese score (TOPICS-10 study) [19] | ≥8† | 32 | 0.774 (0.760–0.788) | 71 | 70 | 41 | 6.9 | 99 | 0.002 |
| The Leicester Risk Assessment score [20] | ≥17 | 30 | 0.773 (0.759–0.788) | 71 | 71 | 42 | 7.2 | 99 | 0.007 |
| - | ≥16 | 38 | - | 78 | 64 | 42 | 6.3 | 99 | - |
| Thai score [21] | ≥8 | 34 | 0.763 (0.749–0.778) | 72 | 68 | 39 | 6.5 | 99 | 0.114 |
| - | ≥7 | 46 | - | 83 | 55 | 38 | 5.5 | 99 | - |
| Finnish score (DETECT-2 study) [22] | ≥5 | 38 | 0.759 (0.744–0.774) | 75 | 63 | 38 | 6.0 | 99 | 0.491 |
| - | ≥7 | 19 | - | 50 | 82 | 32 | 8.2 | 98 | - |
| Brazilian score [23] | ≥11 | 46 | 0.751 (0.737–0.766) | 85 | 55 | 40 | 5.7 | 99 | 0.723 |
| - | ≥18 | 16 | - | 42 | 85 | 27 | 8.1 | 98 | - |
| Indian score [24] | ≥18 | 35 | 0.751 (0.736–0.766) | 74 | 66 | 40 | 6.5 | 99 | 0.617 |
| - | ≥17 | 39 | - | 77 | 63 | 39 | 6.1 | 99 | - |
| Japanese score (Doi et al.) [25] | ≥12 | 32 | 0.750 (0.734–0.767) | 70 | 69 | 39 | 6.6 | 99 | 0.552 |
| - | ≥14 | 23 | - | 56 | 79 | 35 | 7.7 | 98 | - |
| Chinese score [26] | ≥16 | 40 | 0.733 (0.717–0.748) | 72 | 61 | 33 | 5.5 | 99 | 0.004 |
| - | ≥14 | 62 | - | 92 | 39 | 32 | 4.6 | 99 | - |
| British score [12] | ≥4 | 39 | 0.730 (0.714–0.747) | 74 | 62 | 36 | 5.8 | 99 | 0.002 |
| - | ≥6 | 15 | - | 41 | 86 | 27 | 8.5 | 98 | - |
| Rotterdam model [27] | ≥33.9 | 35 | 0.727 (0.710–0.743) | 68 | 67 | 35 | 6.0 | 99 | 0.001 |
| - | ≥37 | 23 | - | 52 | 78 | 30 | 6.9 | 98 | - |
| Oman score [28] | ≥11 | 33 | 0.726 (0.710–0.742) | 68 | 68 | 35 | 6.2 | 99 | <0.001 |
| - | ≥10 | 36 | - | 70 | 65 | 35 | 5.9 | 99 | - |
| French score (DESIR study) [29] | ≥2† | 59 | 0.705 (0.688–0.721) | 88 | 42 | 29 | 4.5 | 99 | <0.001 |
| Kuwait score [30] | ≥19 | 37 | 0.697 (0.681–0.714) | 69 | 64 | 32 | 5.6 | 98 | <0.001 |
| - | ≥32 | 7 | - | 23 | 93 | 16 | 9.8 | 97 | - |
Abbreviations: AROC, area under the curve of receiver operating characteristic curve; CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value; N/A, not applicable
Risk scores other than the KRS were arranged in the order of higher AROC values. The performance of different risk scores for screening for undiagnosed diabetes was evaluated with each score’s original cutoff value and the new cutoff value showing the highest Youden index. The results in the upper row of each score are based on the new cutoff value with the highest Youden index. The results in the bottom row of each score are based on the original cutoff value.
*P values for the comparison of ROC curves between the KRS and other scores were calculated using DeLong’s method [31].
†The new cutoff and original cutoff were the same for Japanese scores and French scores.