Table 4. ROC analysis of K10 score as a predictor of any past-month disorder: Prevalence-dependent indices (overall N = 62280).
| K10 Cut-off | Proportion at or above cut-off (95% CI) |
False negative proportion (95% CI) |
False positive proportion (95% CI) |
Correctly classified proportion (95% CI) |
Positive predictive value (95% CI) |
Negative predictive value (95% CI) |
|---|---|---|---|---|---|---|
| 4 | 0.636 (0.623–0.649) |
0.001 (0–0.002) |
0.567 (0.554–0.58) |
0.432 (0.419–0.445) |
0.109 (0.099–0.119) |
0.996 (0.994–0.998) |
| 5 | 0.535 (0.521–0.549) |
0.002 (0.001–0.003) |
0.467 (0.453–0.481) |
0.531 (0.517–0.545) |
0.128 (0.116–0.14) |
0.995 (0.992–0.998) |
| 6 | 0.451 (0.438–0.464) |
0.004 (0.002–0.006) |
0.384 (0.371–0.397) |
0.612 (0.599–0.625) |
0.149 (0.135–0.163) |
0.993 (0.99–0.996) |
| 7 | 0.378 (0.365–0.391) |
0.005 (0.003–0.007) |
0.312 (0.299–0.325) |
0.683 (0.67–0.696) |
0.173 (0.157–0.189) |
0.991 (0.988–0.994) |
| 8 | 0.322 (0.309–0.335) |
0.006 (0.004–0.008) |
0.257 (0.245–0.269) |
0.737 (0.725–0.749) |
0.2 (0.182–0.218) |
0.991 (0.988–0.994) |
| 9 | 0.27 (0.258–0.282) |
0.007 (0.005–0.009) |
0.206 (0.195–0.217) |
0.787 (0.776–0.798) |
0.237 (0.216–0.258) |
0.99 (0.987–0.993) |
| 10 | 0.225 (0.214–0.236) |
0.009 (0.006–0.012) |
0.163 (0.154–0.172) |
0.827 (0.817–0.837) |
0.274 (0.25–0.298) |
0.988 (0.985–0.991) |
| 11 | 0.191 (0.181–0.201) |
0.012 (0.009–0.015) |
0.132 (0.123–0.141) |
0.856 (0.847–0.865) |
0.307 (0.28–0.334) |
0.985 (0.981–0.989) |
| 12 | 0.165 (0.156–0.174) |
0.014 (0.011–0.017) |
0.107 (0.099–0.115) |
0.879 (0.871–0.887) |
0.347 (0.317–0.377) |
0.983 (0.979–0.987) |
| 13 | 0.137 (0.128–0.146) |
0.018 (0.015–0.021) |
0.084 (0.077–0.091) |
0.898 (0.89–0.906) |
0.388 (0.355–0.421) |
0.98 (0.976–0.984) |
| 14 | 0.115 (0.107–0.123) |
0.022 (0.018–0.026) |
0.066 (0.06–0.072) |
0.912 (0.905–0.919) |
0.423 (0.386–0.46) |
0.975 (0.971–0.979) |
| 15 | 0.098 (0.09–0.106) |
0.025 (0.021–0.029) |
0.052 (0.046–0.058) |
0.923 (0.916–0.93) |
0.469 (0.429–0.509) |
0.972 (0.967–0.977) |
| 16 | 0.083 (0.076–0.09) |
0.029 (0.024–0.034) |
0.041 (0.036–0.046) |
0.93 (0.923–0.937) |
0.506 (0.463–0.549) |
0.969 (0.964–0.974) |
| 17 |
0.071 (0.065–0.077) |
0.033 (0.028–0.038) |
0.033 (0.028–0.038) |
0.934 (0.927–0.941) |
0.532 (0.485–0.579) |
0.964 (0.959–0.969) |
| 18 | 0.059 (0.053–0.065) |
0.037 (0.032–0.042) |
0.025 (0.021–0.029) |
0.938 (0.932–0.944) |
0.573 (0.521–0.625) |
0.961 (0.956–0.966) |
| 19 | 0.052 (0.046–0.058) |
0.04 (0.035–0.045) |
0.02 (0.016–0.024) |
0.94 (0.934–0.946) |
0.602 (0.547–0.657) |
0.958 (0.953–0.963) |
| 20 | 0.045 (0.04–0.05) |
0.043 (0.038–0.048) |
0.016 (0.013–0.019) |
0.941 (0.935–0.947) |
0.633 (0.576–0.69) |
0.955 (0.95–0.96) |
| 21 | 0.037 (0.032–0.042) |
0.046 (0.041–0.051) |
0.012 (0.009–0.015) |
0.942 (0.936–0.948) |
0.675 (0.616–0.734) |
0.952 (0.946–0.958) |
| 22 | 0.032 (0.027–0.037) |
0.048 (0.042–0.054) |
0.009 (0.007–0.011) |
0.942 (0.936–0.948) |
0.707 (0.644–0.77) |
0.95 (0.944–0.956) |
| 23 | 0.025 (0.021–0.029) |
0.053 (0.047–0.059) |
0.007 (0.005–0.009) |
0.94 (0.934–0.946) |
0.718 (0.646–0.79) |
0.946 (0.94–0.952) |
| 24 | 0.019 (0.015–0.023) |
0.056 (0.05–0.062) |
0.005 (0.003–0.007)E |
0.939 (0.933–0.945) |
0.75 (0.671–0.829) |
0.942 (0.936–0.948) |
| 25 | 0.014 (0.011–0.017) |
0.059 (0.053–0.065) |
0.003 (0.002–0.004)E |
0.938 (0.932–0.944) |
0.822 (0.738–0.906) |
0.94 (0.934–0.946) |
| 26 | 0.013 (0.01–0.016) |
0.06 (0.054–0.066) |
0.002 (0.001–0.003)E |
0.937 (0.931–0.943) |
0.821 (0.735–0.907) |
0.939 (0.933–0.945) |
| 27 | 0.011 (0.008–0.014) |
0.062 (0.056–0.068) |
0.002 (0.001–0.003)E |
0.936 (0.93–0.942) |
0.848 (0.752–0.944) |
0.937 (0.931–0.943) |
| 28 | 0.009 (0.006–0.012) |
0.063 (0.057–0.069) |
F | 0.936 (0.93–0.942) |
0.862 (0.762–0.962) |
0.937 (0.931–0.943) |
| 29 | 0.006 (0.004–0.008)E |
0.065 (0.059–0.071) |
0.001 (0.001–0.001)E |
0.934 (0.928–0.94) |
0.947 (0.868–1.026) |
0.934 (0.928–0.94) |
Note: CI = confidence interval; optimal cut-off for prevalence estimation (where false positive proportion = false negative proportion) is show in bold. Area under the ROC curve (AUC) = 0.915; 95% CI, 0.900–0.929. Indices for cut-offs < 4 or >29 are suppressed due to Statistics Canada’s vetting requirements related to small cell sizes. Cases representing 3.3% of the population are missing due to missing K10 or any past-month disorder variables.
E Coefficient of variation > 16.5% and < = 33.3%; high level of sampling error.
F Values suppressed due to extreme level of sampling error (coefficient of variation > 33.3%).