A Prospective Assessment of Practice Variation in the Treatment of Pediatric Gastroenteritis

Abstract

Background: Patient level data exploring the impact of practice variation and resource utilization on outcomes in children with gastroenteritis is unknown. Objective: The primary objective was to determine if significant variation exists in the frequency of intravenous rehydration administration. Secondary objectives included analyzing the variation in ondansetron administration, hospitalization, and revisit rates. Design/Methods: We conducted a prospective cohort study of consecutive children who presented to 11 Canadian emergency departments (EDs). Eligible children were aged 3-48 months and had gastroenteritis (3 watery stools in a 24-hour period within 72 hours prior to the ED visit). Clinical and historical features were collected via a survey conducted following triage; information regarding investigations, treatments, and disposition were abstracted from the medical record; a follow-up call two weeks later provided information on future health care visits. Results: 647 of 694 eligible children were enrolled. All children had a chart review completed and 455 (69%) participated in the ED interview. Of these 455 patients, 398 (89%) had telephone follow-up. The mean age of all participants was 2112 months. The number of children enrolled per site ranged from 24-139. Overall, 24% of children were treated with intravenous rehydration (range per hospital was 7-80%; P<0.001) and 14% were administered ondansetron (range 0-38%; P<0.001). No other antiemetics were administered. Logistic regression analysis revealed that the greatest predictor of intravenous rehydration was a history of bilious emesis (OR 6.9; 95% CI 1.6-29.4). Other predictors included the presence of vomiting in the 48 hours prior to ED arrival and hospital where care was provided. The hospitalization rate was 5% (range 0-12%) and varied between institutions (P=0.02). Children who received intravenous rehydration at the index visit were more likely to see a health care provider in the subsequent 2 weeks (29% vs. 19%; P=0.04) and to revisit an ED (20% vs. 9%; P=0.003). Conclusions: In this cohort, the use of intravenous rehydration and ondansetron varies dramatically. Use of intravenous rehydration at the index visit was associated with the color and presence of vomiting in addition to the site where care was provided. Knowledge translation efforts should focus on improving the care provided to children with gastroenteritis.