Table 4.
Performance of the Kunzmann, Wang and CanPredict risk algorithms in men in the QResearch validation cohort overall and by ethnic group.9,10
| Statistic | Age 50–84 years |
Age 40–84 years |
||
|---|---|---|---|---|
| 5-year risk estimate Kunzmann (OAC) |
5-year risk estimate CanPredict (OAC) |
10-year risk estimate Wang (OSCC) |
10-year risk estimate CanPredict (OSCC) |
|
| Mean (95% CI) | Mean (95% CI) | Mean (95% CI) | Mean (95% CI) | |
| Harrell's C | ||||
| Overall | 0.665 (0.655–0.675) | 0.699 (0.686–0.713) | 0.649 (0.64–0.657) | 0.771 (0.762–0.779) |
| White | 0.663 (0.658–0.667) | 0.702 (0.696–0.708) | 0.642 (0.638–0.646) | 0.769 (0.765–0.772) |
| Indian | 0.711 (0.676–0.747) | 0.693 (0.653–0.732) | 0.688 (0.656–0.72) | 0.781 (0.748–0.815) |
| Pakistani | 0.689 (0.617–0.761) | 0.684 (0.614–0.753) | 0.6 (0.535–0.666) | 0.713 (0.65–0.777) |
| Bangladeshi | 0.465 (0.337–0.592) | 0.595 (0.443–0.746) | 0.596 (0.533–0.659) | 0.546 (0.454–0.637) |
| Other Asian | 0.572 (0.511–0.633) | 0.57 (0.5120–0.629) | 0.665 (0.623–0.708) | 0.734 (0.693–0.776) |
| Caribbean | 0.592 (0.548–0.635) | 0.638 (0.59–0.687) | 0.652 (0.61–0.694) | 0.739 (0.703–0.776) |
| Black African | 0.632 (0.574–0.689) | 0.598 (0.526–0.67) | 0.674 (0.627–0.721) | 0.753 (0.711–0.796) |
| Chinese | 0.709 (0.62–0.797) | 0.707 (0.572–0.842) | 0.687 (0.61–0.763) | 0.849 (0.783–0.915) |
| Other | 0.612 (0.569–0.655) | 0.604 (0.554–0.654) | 0.641 (0.595–0.686) | 0.753 (0.71–0.797) |
| Royston'sD statistic | ||||
| Overall | 0.905 (0.846–0.965) | 1.19 (1.11–1.27) | 0.832 (0.778–0.886) | 1.63 (1.56–1.69) |
| White | 0.89 (0.826–0.954) | 1.2 (1.11–1.28) | 0.795 (0.739–0.851) | 1.58 (1.52–1.65) |
| Indian | 1.22 (.718–1.72) | 1.41 (.614–2.21) | 1.06 (.453–1.67) | 1.87 (1.21–2.52) |
| Pakistani | a | a | a | 1.05 (.0451–2.05) |
| Bangladeshi | a | a | a | a |
| Other Asian | a | a | 0.802 (0.052–1.55) | 1.41 (.377–2.44) |
| Caribbean | 0.766 (0.0452–1.49) | 0.926 (0.025–1.83) | 1.15 (.617–1.69) | 1.39 (.744–2.04) |
| Black African | a | a | a | 1.77 (.46–3.07) |
| Chinese | a | a | a | a |
| Other | a | 0.897 (0.085–1.71) | 0.846 (0.252–1.44) | 1.63 (.918–2.34) |
| R2 (%) | ||||
| Overall | 16.4 (14.6–18.2) | 25.3 (22.6–27.9) | 14.2 (12.6–15.8) | 38.7 (36.9–40.6) |
| White | 15.9 (14–17.8) | 25.5 (22.8–28.2) | 13.1 (11.5–14.7) | 37.5 (35.5–39.5) |
| Indian | 26.3 (10.3–42.2) | 32.1 (7.78–56.5) | 21.3 (2.26–40.4) | 45.3 (28–62.6) |
| Pakistani | a | a | a | a |
| Bangladeshi | a | a | a | a |
| Other Asian | a | a | a | 32 (.292–63.6) |
| Caribbean | a | a | 24.1 (7.12–41.2) | 31.7 (11.6–51.9) |
| Black African | a | a | a | 42.1 (4.64–79.6) |
| Chinese | a | a | a | a |
| Other | a | a | a | 38.6 (18.1–59.2) |
a
Cells with too few observationsto be estimated reliably.