Esophageal acid and bile reflux in mechanically ventilated patients

Introduction

Previous studies using scintigraphy have documented a high incidence (61%) of gastroesophageal reflux in mechanically ventilated (MV) patients. The aim of this study was to assess, over a prolonged period of time, the incidence and type (bile or acid) of reflux, the effect of body position on reflux, and the potential relationship between pathological reflux and esopahagitis in mechanically ventilated patients.

Methods

24 critically ill, MV patients (mean APACHE II score: 21, mean age 64 ± 14 yrs, 15 men), hospitalized in a medical intensive care unit, were prospectively included for 24-h esophageal pH and duodeno-gastroesophageal reflux (DGER) studies (Digitrapper III pH-meter and fiber-optic sensor for the presence of bilirubin, Bilitec 2000, Synectics, Sweden) with single sensors placed 5 cm proximal to the lower esophageal sphincter (LES) and instrumented with a nasogastric tube. Exclusion criteria were: MV <3 or >8 days, previous initiation of enteral feeding, history of reflux esophagitis or gastric surgery, acute gastrointestinal bleeding or exophageal varices, or medication with prokinetic agents. All patients received stress ulcer prophylaxis with ranitidine 50 mg iv, tid. On the day before the study patients underwent esophageal endoscopy to determine the presence of esophagitis. Patients were placed in supine semi-recumbent position and turned on either left or right lateral sides at intervals as deemed appropriate by the attending nurse. The following variables were analyzed depending on body position (total time, time in supine, right, or left position): % time pH <4 (normal <3.4 %) or % time bilirubin absorption >0.14 (normal: <3 %), total duration of reflux, and number reflux episodes. Data are medians with interquartile ranges (IQR), significance was tested with the Mann-Whitney U test. Fisher's exact test was used to analyze relationships between the presence or absence of pathological reflux and esophagitis.

Results

After a median of 5 days of MV (IQR 3.8) 12 of 24 patients (50%) patients had pathological DGER (median % reflux time over the whole recording time independent of body position: 8%, IQR 0.62%). In contrast, only one patient had significant acid reflux (4.8% of the recording time) and the median pH in the lower esophagus for all 24 patients was 6.7 (IQR 6.7). The median duration where bile salts were present in the lower esophagus was 112 min (IQR 0.865 min), in 8 of 24 patients this time period exceeded 5 h. The median number of DGER episodes was 12 (IQR 0.37). The relative time of pathological reflux was significantly higher in left lateral and supine position as compared to right lateral position (7.1% IQR 0.64% and 8%, IQR 0.65% vs. 4.7% IQR 0.70%, P <0.01. 12 of 24 patients (50%) had esophagitis. There was a significant positive relationship between the presence of pathological DGER and the presence of esophagitis (P = 0.04).

Conclusions

Under standard stress ulcer prophylaxis with ranitidine, critically ill patients with MV have a high incidence (50%) of DGER but not of acidic esophageal reflux. Reflux is highest in the left lateral and in the semirecumbent position. The presence of bile salts in the esophagus for prolonged periods of time suggest that: 1) the barrier function of the LES and the clearance function of the esophagus are deranged, 2) esophagitis in these patients may not only be a result of mechanical irritation due to the nasogastric tube, but may also be a chemical esophagitis, 3) intestinal fluids refluxing into the esophagus may be conducive to ventilator-associated pneumonia.