0634 Factors Influencing Best Practices in Narcolepsy: Results from a Predictive Modeling Analysis of Educational Outcomes Data

Abstract

A continuing medical education (CME) activity was developed to improve knowledge and practice behaviors in healthcare providers (HCPs) treating patients with narcolepsy. The goal of this study was to utilize predictive modeling on educational outcomes data to determine factors influencing best practice behaviors so that any barriers may be addressed in future educational activities. Educational outcomes data were obtained from an activity on narcolepsy, which involved a 3-part webcast. Surveys assessing knowledge, confidence, and behavior were administered before and immediately following the activity. Two analyses using PredictCME (based on a method known as chi-square automatic interaction detection) were conducted on data from the pre-activity survey, which included questions related to practice behaviors; one question asked how often HCPs utilized a screening tool and one asked how often HCPs utilized diagnostic recommendations from the American Academy of Sleep Medicine (AASM). Data from each behavior question were used as the response variable in the 2 separate analyses. Demographics, knowledge, confidence, and evaluation data were entered as predictors. Data from 601 pre-survey participants were available for the analysis. Findings revealed specialty to be the strongest predictor of screening and diagnostic behaviors (X2(1) = 139.15, p < .001 and X2(1) = 40.42, p < .001, respectively), with sleep specialists more likely to perform these behaviors compared to non-sleep specialists. Results from the PredictCME analysis are not surprising in that sleep specialists are often directly involved in addressing sleep issues in patients. However, one could argue for the importance of non-sleep specialists utilizing screening tools, as that would help determine whether or not a referral to a sleep specialist is needed. Further, although a greater percentage of sleep specialists versus non-sleep specialists utilized diagnostic recommendations, only half of them did so. Therefore future educational activities should emphasize the importance of screening tools for non-sleep specialists, as well as diagnostic recommendations for both sleep and non-sleep specialists. The educational activity described in this presentation was supported by an educational grant from Jazz Pharmaceuticals.