Abstracts

Rationale: The Personal Impact of Epilepsy Scale (PIES) is a validated QOL questionnaire with 25 items in 3 domains: seizures, medication side effects, and psychosocial factors. There are scores for each domain, and a total score. This study establishes the minimum clinically important difference (MCID), representing the smallest score change corresponding to meaningful changes in QOL.

Methods: Participants were age >16 seen in epilepsy clinic by GM >2 times between 1/1/15 and 12/31/17. All patients were offered PIES at all clinic visits. Each PIES subscale was scored 0-40, with total PIES score range 0-120, and lower scores indicating higher QOL. Clinic notes were reviewed for demographics, visit dates, diagnoses, seizure types and frequencies, medications and PIES. Clinical status was rated as “improved,” “unchanged” or “worse” compared to the prior visit, (overall, and for each of the three PIES domains) by a rater blinded to PIES scores. This analysis included patients with valid PIES scores within each respective domain for 2 consecutive visits. MCID was calculated using an anchor-based approach with clinical status change rating (improved, unchanged or worse) as the anchor. Significance of PIES score differences between clinical status groups was examined overall and within each domain using the Kruskal Wallis H test. MCID was determined by calculating mean between-visit change in PIES score, overall and within each domain, for each clinical status group.

Results: Of 400 patients reviewed, 120 met entry criteria. They were 56.6% female and reported mean 9.3±24.1 seizures/month. Figure 1 shows mean PIES changes by follow-up status: Overall mean PIES difference was -8.2±14.2 for the improved group, -1.3±14.4 for the unchanged group, and -0.57±19.2 for the worse group (p=0.056). Seizure impact mean PIES score change was -3.4±8.1 for the improved group, -2.2±8.4 for unchanged group, and 0.89±10.2 for the worse group (p=0.162). Medication side effects mean change was 0.68±6.0 for improved patients, -0.77±7.9 for the unchanged group, and 0.20±6.3 for worse group (p=0.705). Psychosocial status scores mean change was -4.2±7.5 for improved group, -0.44±6.8 for unchanged group, and 0.41±8.2 for worse group (p=0.025). Clinical improvement is represented by a MCID of -8, -3, and -4 for overall, seizure, and psychosocial status, respectively. PIES MCID for medication impact and for clinical worsening could not be determined.

Conclusions: PIES appeared most sensitive to clinical improvement. The psychosocial domain appeared to have the largest influence on the overall PIES score. These data suggest that PIES total score can be a reliable indicator of clinical improvement in QOL. Future research should confirm our findings using additional validation methods that further evaluate use of PIES for determining worsened clinical status.

Funding: Please list any funding that was received in support of this abstract.: Supported by an unrestricted grant from Eisai Co., Ltd.