Development of a validated patient-reported symptom metric for pediatric eosinophilic esophagitis: qualitative methods

Abstract

Background: Previous attempts to measure symptoms in pediatric Eosinophilic Esophagitis (EoE) have not fully included patients and parents in the item development process. We sought to identify and validate key patient self-reported and parent proxy-reported outcomes (PROs) specific to EoE.

Methods: We developed methodology for focus and cognitive interviews based on the Food and Drug Administration (FDA) guidelines for PROs, the validated generic PedsQL™ guidelines, and the consolidated criteria for reporting qualitative research (COREQ). Both child (ages 8-12 and 13-18) and parent-proxy (ages 2-4, 5-7, 8-12, and 13-18) interviews were conducted.

Results: We conducted 75 interviews to construct the new instrument. Items were identified and developed from individual focus interviews, followed by cognitive interviews for face and content validation. Initial domains of symptom frequency and severity were developed, and open-ended questions were used to generate specific items during the focus interviews. Once developed, the instrument construct, instructions, timeframe, scoring, and specific items were systematically reviewed with a separate group of patients and their parents during the cognitive interviews.

Conclusions: To capture the full impact of pediatric EoE, both histologic findings and PROs need to be included as equally important outcome measures. We have developed the face and content validated Pediatric Eosinophilic Esophagitis Symptom Score (PEESS™ v2.0). The PEESS™ v2.0 metric is now undergoing multisite national field testing as the next iterative instrument development phase.

Figure 1

Summary of content validation methodology. Focus interview transcripts of pediatric patients with EoE and their parents were used to develop the items and domains for the PEESS™ v2.0. Cognitive interviewing was conducted with separate cohorts of pediatric patients and their parents in the 8-12 and 13-18 year old age groups, while parent proxy-reports were also obtained in the 2-4 year old and 5-7 year old age groups.

Figure 2

From the PEESS v1.0 to the validated PEESS™ v 2.0. The instrument layout and overall design went through several iterative drafts based on patient and parent-proxy feedback through focus interviews and separate cognitive interviews. For example, participants favored presenting frequency and severity of particular items side by side rather than on two separate pages.