RESUMO
OBJECTIVES: To compare the ultrasound examination with pH-metry in order to evaluate the diagnostic accuracy of sonography for diagnosis of gastro-oesophageal reflux (GOR) in preterm infants. METHODS: A prospective study was conducted on 31 premature infants <34 weeks with clinically diagnosed GOR. First, they underwent 24-hour pH-monitoring; successively, the sonographic assessment was performed within 12 h after pH-monitoring. The two operators who performed the pH-monitoring and sonography respectively, were unaware of each other's results. RESULTS: Twenty-one patients (67.7%) had significant GOR with a reflux index >5 (GOR group). The median (range) reflux index in this group was 9.19% (6.04-20.1). Ten newborns (32.3%) did not have significant GOR with a reflux index <5. Sonography was positive for GOR in 8 patients (25.8%); all 8 infants with sonographic diagnosis of GOR had a reflux index >5. Therefore, sonographic diagnosis did not produce false positives. Sonography was negative in 23 newborns (74.2%); 13 of these were positive to pH-monitoring and may therefore be considered as false negatives. Respect to continuous 24-hour pH-monitoring, sonography showed a specificity of 100% but a very low sensitivity of 38% with a positive prediction value of 100% and a negative prediction value of 43%. CONCLUSIONS: Sonography should not replace 24-hour pH monitoring for detecting GOR in preterm infants. However, sonography has a very high specificity and a positive predictive value of 100%. When clinicians suspect GOR in preterm infants, it could be useful for selection of cases to refer for pH-metry.
