Prediction of refractory gastroesophageal reflux disease in young children

Abstract

BACKGROUND: Many children respond to medical treatment for gastroesophageal reflux disease (GERD). However, some may require invasive intervention for refractory disease. Due to the lack of prognostic tools in children, this study aimed to develop a predictive model for refractory GERD. METHODS: A retrospective review was performed in children with symptoms of GERD at a university hospital. Refractory GERD was defined as an unresponsive disease after optimal treatment with medication for >8 weeks. The predictive model was constructed based on clinical features and 24-h multichannel intraluminal impedance-pH (MII-pH) monitoring results. RESULTS: A total of 205 children were included with a median (IQR) age of 0.6 (0.3, 2.0) years. Over half of the patients (59.5%) had motor disabilities. Forty-four children (21.5%) were diagnosed with refractory GERD and subsequently underwent fundoplication. Multivariable analysis suggested that the refractory disease was associated with motor disabilities (OR: 5.35; 95% CI: 2.06-13.91), recurrent aspiration pneumonia (OR: 2.78; 95% CI: 1.24-6.26), prematurity with an onset of GERD at a post-conceptual age <40 weeks (OR: 6.76; 95% CI: 1.96-23.33), and abnormal total reflux episodes according to age (OR: 2.78; 95% CI: 1.24-6.19), but not the acid exposure time or symptom association analysis. The predictive model for refractory GERD based on associated factors revealed an area under the ROC curve of 76.8% (95% CI: 69.2%-84.3%) with a sensitivity of 77.3% and a specificity of 64% when applying a cutoff score of ≥2.5. CONCLUSIONS: The predictive model, using clinical features and MII-pH, may be an additional tool to predict refractory GERD in young children.