Abstracts

Rationale: In this retrospective study we evaluated the usefulness of repeated serum prolactin measurements in patients admitted to the epilepsy monitoring unit to distinguish psychogenic non-epileptic seizures from epileptic seizures.Methods: The medical records of 607 consecutive patients admitted to the EMU between 2009 and 2012 were reviewed. All patients had a baseline serum prolactin level obtained upon admission and within 30 minutes following a clinical event. A significant elevation of serum prolactin was defined as twice the patient s baseline value. The diagnosis of psychogenic non epileptic seizures versus epileptic seizures was confirmed by video EEG.Results: Out of 607 patients admitted to the EMU, 389 were excluded because they did not have a clinical event or incomplete information. Therefore 218 patients were included in the analysis. There are 59 males and 159 females (age range 15-84 years). There are 146 patients diagnosed with PNES and 42 had elevated serum prolactin representing 28.8% of false positive indicators. Nine patients with GTC seizures and 13 with CPS and secondary generalization all had signification elevation of serum prolactin. Four out of five patients with simple partial seizures had elevated serum prolactin. There were 32 patients with CPS and 5 of them did not have elevated serum prolactin representing 15.6% of false negative levels. Conclusions: This observation demonstrates that serum prolactin has no added diagnostic value in the EMU and may confuse the clinical diagnosis by the addition of false positive and false negative results. Furthermore, measurements of serum prolactin add unnecessary blood draws and financial burden to the patients. Therefore, we recommend that serum prolactin levels should not be obtained routinely in the EMU.