. Author manuscript; available in PMC: 2017 Jun 20.

Published in final edited form as: Neurocrit Care. 2016 Jun;24(3):324–331. doi: 10.1007/s12028-016-0245-y

Table 2.

Predictors of PET Hypermetabolism

cEEG risk factor	Sensitivity (95 % CI)	Specificity (95 % CI)	Accuracy (95 % CI)	p value
Periodic discharges
PD^a	0.91 (0.57–0.99)	0.43 (0.12–0.80)	0.72 (0.46–0.89)	0.25
LPD^b	0.70 (0.35–0.92)	0.50 (0.10–0.91)	0.64 (0.35–0.86)	0.58
Duration, long^c	0.70 (0.35–0.92)	0.75 (0.22–0.99)	0.71 (0.41–0.90)	0.25
Amplitude, high^c	0.30 (0.08–0.65)	1.00 (0.40–1.00)	0.57 (0.30–0.81)	0.36
Superimposed fast activity^c	0.40 (0.14–0.73)	0.50 (0.10–0.91)	0.43 (0.19–0.70)	1.00
Frequency
>1 Hz	0.40 (0.14–0.73)	1.00 (0.40–1.00)	0.57 (0.30–0.81)	0.25
>2.5 Hz	0.18 (0.03–0.52)	1.00 (0.56–1.00)	0.56 (0.32–0.78)	0.49
Seizures	0.27 (0.07–0.60)	0.86 (0.42–0.99)	0.50 (0.27–0.73)	1.00
Status epilepticus
Electrographic SE or electroclinical SE (vs. IIC with no SE)	0.79 (0.53–0.93)	1.00 (0.51–1.00)	0.83 (0.61–0.94)	0.01
Electroclinical SE without electrographic SE (vs. IIC with no SE)	0.80 (0.44–0.94)	1.00 (0.51–1.00)	0.86 (0.60–0.96)	0.01

Significant p-values (<0.05) are in bold

PD periodic discharges, GPD generalized periodic discharges, LPD lateralized periodic discharges, RDA rhythmic delta activity

Versus RDA

Versus GPD

Evaluated only for subjects with PDs