Rationale:
We have noted an increase in length of stay (LOS) defined as greater than 10 days in patients who undergo stereotactic electroencephalogram (SEEG) procedure mostly due to lack of or delayed seizure occurrence during their stay. Our proposed theories for this prolonged LOS include opiate effects in epilepsy, insertional effect of depth electrodes, and prolonged half-life of antiseizure medications (ASMs).
Methods:
A total of 31 patient charts were reviewed from July 2020 to April 2025. Exclusion criteria included pediatric patients (those who underwent SEEG prior to the age of 18), inaccessible charts dated prior to July 2020 due to shift in electronic medical records, and patients who had a prolonged LOS due to post-operative complications. Charts were reviewed for LOS between implantation and explantation, dosages and types of opiates administered during hospitalization for post-operative pain, home ASMs, rate of ASM cessation, pre-operative seizure frequency, duration between implantation and first seizure, seizure type, seizure occurrence during mapping, type of epilepsy surgery, and Engel Class Outcome. Opiates were converted to oral morphine milligram equivalents (OMME) and directly compared to patient’s LOS. ASMs were also compared to LOS. Patient’s pre-operative seizure frequency were stratified into categories (monthly, weekly, daily, unknown) which were also compared to their LOS.
Results:
Of the 31 patients, all but one had seizures after during their admission. The single patient who did not have a seizure did have one during cortical mapping. Results found a modest correlating trend between OMME and LOS as displayed in the attached graph. The more opiates a patient received, the longer their LOS were. Subgroup analyses did not find that a specific opiate (i.e. hydromorphone versus hydrocodone) contributed more to a prolonged LOS than another. Patients on cenobamate had the longest LOS (on average 13.6 days). Cenobamate was also noted to be held on post-operative day 0 in these patients. Patients who reported daily seizures averaged a 9 day LOS compared to those reporting monthly seizures with an average 13 day LOS. Patients who had a LOS at 10 days or longer were found to have their first seizure occur at 6.4 days on average compared to 2.6 days on average for those who stayed less than 10 days.
Conclusions:
Data analysis indicates that prolonged LOS may be influenced by multiple factors, including quantity of opiates and ASMs with prolonged half-lives. Initially we thought that specific mechanisms of action of opiates may be an influencing factor, however our results have not contributed to that conclusion at this time. Our results also clearly show that a prolonged length of stay is indeed secondary to delay in seizure occurrence which may be a result of the aforementioned factors. A confounding variable in this study is those with more frequent seizures had a shorter LOS than those without. Limitations include small sample size and retrospective chart review. Phase 2 of this project will evaluate if an insertional effect and various activation measures (i.e. sleep deprivation, etc.) contribute to a prolonged LOS in spite of the confounding variable present.
Funding: None