Monitoring Neurological Injury by qEEG - Quantitative EEG Analysis Methods and Clinical Applications - page 184

Begin
conventional
hypothermia

Begin
immediate
hypothermia

1.2

Hypothermia immediately

1.0

0.8

0.6

0.4

0.2

Hypothermia
after 60 minutes

0

0 1 2 3 4

Time (hours)

Figure 7.9 Subband IQ of two representative animals receiving 9-minute asphyxial CA, followed by
conventional hypothermia after 60 minutes and immediate hypothermia, and subsequently moni-
tored for the first 4 hours. Immediate hypothermia results in rapid and early recovery of IQ levels.

values at 72 hours (Pearson correlation, 0.408; two-tailed significance, 0.025). In
conclusion, early initiation of hypothermia improves not only the electrical response
but also functional brain recovery in rats after CA.

7.5.3 qEEG Markers Predict Survival and Functional Outcome
The final question considered is whether hypothermia improves eventual survival.
The actual question posed is this: Are the IQ values contrasted for the survivors and
the animals that died prematurely? IQ values for animals that died were 0.48 ± 0.04,

whereas values for the survivor group were 0.66 ± 0.02). The differences are statisti-

cally significant ( p 0.001) [Figure 7.11(a)]. IQ values are determined every 30 min-

utes starting from 30 minutes post-ROSC up until 4 hours. The animals that died
prematurely showed significantly lower IQ values during each 30-minute interval
studied compared with an average of the first 4 hours for the survivor group ( p

<

<

0.05) [Figure 7.11(b)].
Finally, IQ levels of rats with good and bad functional outcomes are compared.
Rats with good outcomes, defined as NDS = 60, had a higher IQ level (0.56 ± 0.02)

than those with poor outcomes, defined as NDS < 60 (0.56 ± 0.02). This difference is

statistically significant ( p 0.001). These differences are consistent throughout the

recovery periods from 30 minutes to 48 hours [Figure 7.11(c)]. To evaluate the over-
all performance, receiver operating characteristics (ROC) were calculated to deter-
mine IQ cutoff points. For an IQ value >0.523, the sensitivity was 81.8% sensitivity
and specificity was 100% for predicting good outcomes, giving an area under the
ROC curve of 0.864 [Figure 7.11(d)].

<