Prediction of seizure control in non-ketotic hyperglycemic induced seizures.

Background: To study the factors predictive for seizure control in non-ketotic hyperglycemic induced seizures (NKHS).

Methods: We studied 21 patients who were clinically diagnosed as NKHS at Khon Kaen University hospital, Thailand. Multiple linear regression analysis was used to identify the factors predictive for seizure control.

Results: Most patients had no previous history of diabetes and presented with repetitive partial seizures. The mean number of seizure attacks was 45 times prior to admission. The average duration to terminate seizure was 36 hours and significantly predicted by frequency of seizures (estimate 0.9, p value 0.013).

Conclusion: Frequency of seizures is the only predictive factor for the success of seizure control in NKHS.