Erythromycin or metoclopramide for feed intolerance in the critically ill
© BioMed Central Ltd 2006
Published: 21 March 2006
In critical illness, nasogastric (NG) feeding is frequently compromised by delayed gastric emptying. This is usually treated by administration of prokinetics, such as metoclopramide or erythromycin. Data on the relative and individual effects of these agents, in critically ill patients, are limited. In addition, erythromycin has been reported to cause hypotension, which could compromise splanchnic perfusion.
To compare metoclopramide (M) and erythromycin (E) in the initial treatment of feed intolerance in critical illness, to determine the success of combination treatment in patients who fail to respond to a single agent, and to examine the effect of E on blood pressure (BP).
Ninety feed-intolerant (6-hourly NG aspirate >250 ml) mechanically ventilated patients (mean APACHE II score = 21.4 ± 0.6) were randomly allocated to receive either M (n = 45; 10 mg i.v. four times daily) or E (n = 45; 200 mg i.v. twice daily) in a prospective, double-blind randomized fashion. Age, sex, APACHE II scores and initial gastric residual volumes were similar between the two groups. After the first dose of prokinetic therapy, NG feeding was restarted at 40 ml/hour and increased in a standardized fashion and 6-hourly NG aspirates were performed. If NG aspirate >250 ml recurred, combination therapy was commenced. Patients were studied for 7 days. The BP was recorded.
Inotrope, opioid and benzodiazepine usage, and blood glucose concentrations were similar in the two groups. After 24 hours of treatment, both M and E significantly reduced 24-hour gastric residual volumes (M: 830 ± 32 ml to 435 ± 30 ml, P < 0.0001; E: 798 ± 33 ml to 201 ± 19 ml, P < 0.0001) and improved the proportion of patients with successful feeding (no aspirate >250 ml; M: 62% and E: 87%). Treatment with E was more effective than with M (P < 0.05). The effectiveness of both treatments on subsequent days declined rapidly. In patients who failed single-agent therapy, combination therapy was highly effective (day 1, 92%) and its effectiveness was maintained for the duration of the study (day 6, 77%). Predictors of poor response to single or combined prokinetic therapy included increased age, APACHE II score, pretreatment gastric residual volume, trauma, and abnormal renal function. E caused a nonsignificant reduction in systolic BP (+1.0 ± 0.9 vs -3.6 ± 0.9 mmHg, P = NS) and diastolic BP (-0.1 ± 0.9 vs -1.7 ± 1.1 mmHg, P = NS).
Erythromycin is safe and more effective than metoclopramide in treating feed intolerance in critical illness; however, rapid tolerance occurs to both agents when used individually. In patients who develop feed intolerance on a single agent, combination therapy is effective with little evidence of tolerance. The role of combined therapy as initial treatment requires further evaluation.