Skip to main content
. 2015 Aug;136(2):e351–e360. doi: 10.1542/peds.2014-3550

FIGURE 2.

FIGURE 2

Recursive partitioning model to identify children at low risk of clinically relevant intracranial abnormalities. Prediction rule sensitivity: 25/40 (62.5%; 95% CI: 45.5–76.9%). Prediction rule specificity: 201/314 (64.0%; 95% CI: 58.4–69.2%). Prediction rule negative predictive value: 201/216 (93.1%; 95% CI: 88.6–95.9%). Likelihood ratio (−): 0.51 (95% CI: 0.39–0.88). Likelihood ratio (+): 1.74 (95% CI: 1.31–2.30).