Predictive value of oesophageal pH variables in children with gastro-oesophageal reflux.

Sixty three 24 hour oesophageal pH studies were performed in children with known or suspected gastro-oesophageal reflux and their progress was followed for at least a year. Forty two studies were of children who were well at follow up who were not receiving medical treatment for gastro-oesophageal reflux, and had not required antireflux surgery. Twenty one studies were from children who were still on treatment for gastro-oesophageal reflux or had undergone antireflux surgery. The results were analysed using non-parametric statistics to assess the value of pH recordings in predicting which children were likely to have continuing problems. Sleep reflux (minutes/hour) and acid clearing time were the most highly significant variables (p less than 0.0005 and p less than 0.0001 respectively). The results suggest that either sleep gastro-oesophageal reflux of more than 6.4 minutes/hour or an acid clearing time of seven minutes or more are good predictors of the clinical outcome (specificity 88.1%, sensitivity 81.0%). Three of the five children with false positive results were under 10 months of age. Positive results in this age group should therefore be interpreted with caution.