Reliability of epigastric auscultation to detect gastric insufflation

We studied the reliability of epigastric auscultation to detect gastric insufflation in 30 anaesthetized, paralysed intubated patients.

A 16FG gastric tube was positioned with the tip in the mid-oesophagus with the proximal end attached to an injection port with a one-way valve. Four observers participated in the study. Observers were paired and each pair studied 15 patients. Each patient underwent four test sequences in random order, two by each observer. Each test sequence comprised one observer injecting different volumes of air (0.25 ml, 0.5 ml, 1 ml, 2 ml, 3 ml, 4 ml, 5 ml, 10 ml, 15 ml and 0 ml as a control) in random order whilst the second blinded observer listened with a stethoscope over the epigastrium. Each randomized volume was injected rapidly at 5 s intervals for 1 min. The number of injections required to detect air entering the stomach was recorded. The stomach was deflated between each test sequence.

To detect air entering the stomach with 95% confidence, 11 injections were required for 0.25 ml; 7 for 0.5 ml; 3 for 1 ml; 2 for 2 ml and 3 ml, and I for > or =4 ml. The mean (range) inter- and intraobserver reliability was 0.73 (0.71-0.75) and 0.76 (0.76-0.89), respectively. The incidence of false positives was 21% (25/120) and the incidence of false negatives was 10% (103/1080), making the specificity and sensitivity 79% and 91%, respectively.

We conclude that epigastric auscultation can detect gastric insufflation of 0.25 ml air after 11 breaths and > or = 4 ml air after one breath with 95% confidence. Inter- and intraobserver reliability is moderate to excellent. Epigastric auscultation should be repeated to reduce the risk of false positives.