THE RUMINATION SEVERITY INDEX (RUMSI): DEVELOPMENT AND EVALUATION OF A SCORING TOOL FOR RUMINATION SYNDROME

Background: Rumination syndrome (RS) is increasingly recognized in the pediatric population, however, there is no existing tool to measure symptom severity. The objective of this study is therefore to develop and evaluate a new scoring tool, the Rumination Severity Index (RumSI), that can be used to measure rumination severity in children.
Methods: We developed a patient/family-reported questionnaire, the RumSI, with 7 main questions to evaluate RS symptoms and impact on daily functioning. Questions were generated based on the clinical experience of several pediatric gastroenterologists and psychologists who specialize in caring for children with RS. The RumSI questions are shown in Figure 1. Patients with RS who were evaluated at our institution or their parents/caregivers were invited to complete the RumSI and the Pediatric Quality of Life Inventory (PedsQL). An overall RumSI score was then calculated based on summing symptoms and statistically analyzed for correlation with PedsQL.
Results: We contacted 105 children with RS and 66 completed the questionnaires (74% female, median age of 15 years) and were included in our study for a 63% completion rate. Patient characteristics are shown in Table 1. Questions regarding vomiting and re-swallowing of regurgitant were combined and a maximum score of 6 was given to those who reported vomiting daily and who were unable to re-swallow. Questions on skipped meals, use of feeding tube or central line, and weight loss were given a maximum score of 4 each if patients skipped more than 1 meal a day, used a feeding tube or central line for extra nutrition, or had weight loss because of RS symptoms. A score for total days of missed school and sports/social activities was assigned, with a maximum score of 5 for patients who missed 40-60 days of school or sports/social activities. The maximum possible RumSI score was 28. RumSI scores for the current sample ranged from 0 to 26.5, with an overall mean of 10.58 (SD 7.16). The RumSI and PedsQL total scores were significantly correlated (r=-0.45, p<0.001), with every 1 point increase in RumSI score associated with a 1.33 decrease in PedsQL total score. The interrater reliability for the RumSI score was ICC1=0.84 (95% CI 0.69-0.92). Bland-Altman plots indicated follow-up measures were positively biased relative to initial measures (mean difference 2.36, 95% CI 1.16-3.58), with the difference between measures increasing with time.
Conclusions: The newly developed RumSI is a valid tool designed for use in children with RS to measure symptom severity. In this study, we demonstrate that the RumSI correlates well with a widely used pediatric quality of life measure. We believe that widespread use of the RumSI will improve the quality of RS research and ultimately provide clinical benefit.