Gastric emptying measurement of liquid nutrients using the 13C-octanoate breath test in critically ill patients: a comparison with scintigraphy

Abstract

PURPOSE: Scintigraphy is considered the most accurate technique for the measurement of gastric emptying (GE) but, for patients in the intensive care unit, it is technically demanding, involves radiation and can interfere with care. The 13Coctanoate breath test (13C-OBT) is a simple, non-invasive technique that does not involve radiation exposure. AIM: To evaluate the performance of the 13C-OBT in the assessment of GE in critically ill patients. METHODS: The GE was assessed in 33 mechanically ventilated patients (23 M; 54.3 ± 3.0 yrs; APACHE II: 22.0 ± 1.1). Following test meal administration (100 ml Ensure_), concurrent scintigraphic measurement and breath samples (13C-OBT) were collected over 4 h. Scintigraphic meal retention was determined and the gastric emptying coefficient(GEC) and half emptying time [t50(BT)] were calculated for the 13C-OBT. Delayed GE was defined as meal retention[13 % at 180 min. RESULTS: Delayed GE was identified in 27/33 patients. Meal retention correlated modestly with t50(BT)(r = 0.55–0.66; P\0.001) and well with GEC (r = -0.63 to -0.74; P\0.0001). The strength of agreement between the two techniques was highest between GEC and retention at 120 min. The best cut-off GEC for defining delayed GE was 3.25 (AUC = 0.75; 95 % CI = 0.52–0.99; P = 0.05), with 89 % sensitivity and 67 % specificity to detect delayed GE. The GE was delayed in all(23/23) patients with feed intolerance(GRV[250 ml) on scintigraphy and 91 % (21/23) patients on 13C-OBT. CONCLUSION: In critical illness, there was a correlation between 13C-OBT and gastric scintigraphy, with GEC performing as a better and more sensitive marker of detecting delayed GE than t50. However the relatively wide95 % confidence intervals suggest that 13C-OBT is more suitable as a technique to assess GE in a group setting for research studies rather than for individual patients in clinical practice.