Abstracts

Rationale: Epilepsy dietary therapies are well‐established, nonpharmacologic treatments used for children and adults who have failed two or more medications for their epilepsy. (Kossoff et al. 2018). It is well established that the diet has a high (80%) side effect rate and psychosocial burden (Lin et al 2017; Lightstone et al 2001). As clinicians we have found ourselves advising families that there is a “learning curve” or duration of time on diet that is required to feel subjectively confident in preparing the diet for their children. This quality improvement study set out to determine if there is a relationship between duration on diet and subjective confidence and if this subjective confidence was related to diet type or way the child is fed. Additionally, since we know that neither Medicaid insurance nor family functioning scores were significantly associated with successful dietary treatment, this QI project sought to see if income based on zip codes had a correlation to perceived confidence (McNamara et al., 2013). Methods: February 2018 - February 2019, clinicians gave paper surveys for families to fill out (voluntarily). This survey contained 1) duration on diet 2) type of diet (ketogenic [KD], modified atkins [MAD], low glycemic index [LGIT] or mixed) 3) way the child is fed (orally, enterally or mixed) 4) subjective confidence score 1-10 (10 most confident, 1 least confident) and a “free text” area to provide comments for future families new to diet. This survey was also translated into Spanish. At one point during the study, all families with our hospitals electronic portal access were provided the survey. We received 115 returned surveys with 85 unique patients and removed follow up surveys from families for this series of analyses. Exclusively oral eaters made up 57% of the group and those enterally fed in any way made up the remainder. We also compared confidence score to average income based on zip code from the International Revenue Service. Results: Though duration on diet did show a positive correlation to parental subjective confidence in meal preparation (confidence at a level of 0.345, p<0.001), at time zero, the mean regression line intercepts the y-axis at approximately 9.1. This indicates that families on average feel confident in meal preparation upon initiation of diet. We also separated groups for those on diet <60 days and those on diet >365 days and there was no significant difference (p=0.10). Additionally, neither analysis demonstrated a difference for keto vs. MAD (p=0.88) or when the group was separated into oral eaters vs enterally (p=0.45). When analyzing income by zip codes, there was also no significant difference in confidence level upon comparison of income class by zip code (Chi Square = 4.44, p = 0.99). Conclusions: Given the very high confidence reported at time zero based on regression, and no difference in confidence for groups on diet <60 days vs >365 days, this QI project did not feel it necessary to ascribe a duration of time on diet that is required to feel subjectively confident in preparing meals. While pleasantly surprised to find such high confidence from time zero, this QI project did not ultimately delineate the groups potentially in need of greater education or follow up appointments (if said groups exist). As we have found that duration on diet, type of diet, the feeding route and socioeconomic factors are not related to diet success or confidence, perhaps those choosing to do the diet are a group of patients/caregivers who are inherently determined to be successful with the diet. Funding: No funding