Skip to main content
. 2017 May 29;7(5):e015546. doi: 10.1136/bmjopen-2016-015546

Table 3.

Sensitivity, specificity, PPV and NPV for each defined cut-point derived from the application of the attribution algorithm (using ‘a priori’ coefficients) to the first NP event observed in the international cohort

Cut-point Sensitivity (%) Specificity (%) Correctly classified (%) LR+ LR− PPV (%) NPV (%)
(≥0) 100.0 0.0 40.7 1 40.7
(≥1) 100.0 1.4 41.6 1.0141 0 41.1 100.0
(≥2) 99.0 6.2 44.0 1.0559 0.1616 42.1 90.0
(≥3) 99.0 16.7 50.2 1.1879 0.0606 45.0 96.0
(≥4) 96.0 31.9 58.0 1.41 0.1265 49.2 92.0
(≥5) 92.9 48.6 66.7 1.8084 0.1455 55.4 90.9
(≥6) 91.9 71.5 79.8 3.2284 0.113 68.9 92.8
(≥7) 87.9 82.6 84.8 5.0618 0.1467 77.7 90.8
(≥8) 69.7 91.0 82.3 7.7203 0.3331 84.1 81.4
(≥9) 47.5 97.2 76.9 17.0909 0.5403 92.1 72.9
(≥10) 19.2 99.3 66.7 27.6364 0.8137 95.0 64.1
(>10) 0.0 100.0 59.3 1 59.3

LR, likelihood ratio; NPV, negative predictive value; NP, neuropsychiatric; PPV, positive predictive value.