Plasma erythromycin concentrations predict feeding outcomes in critically ill patients with feed intolerance

Objective: Motilin receptors are rapidly down-regulated by exposure to erythromycin, and its progressive loss of clinical prokinetic effect may relate to higher plasma drug concentrations. This study aimed to evaluate the relationship between plasma erythromycin concentrations and feeding outcomes in critically ill patients.

Design: Observational comparative study.

Setting: Tertiary critical care unit.

Patients: Twenty-nine feed-intolerant (gastric residual volume >250 mL) mechanically ventilated, medical critically ill patients.

Interventions: Patients received intravenous erythromycin 200 mg twice daily for feed intolerance.

Measurements: Plasma erythromycin concentrations were measured 1 and 7 hrs after drug administration on day 1. Success of enteral feeding, defined as 6-hourly gastric residual volume of ≤250 mL with a feeding rate ≥40 mL/h, was recorded over 7 days.

Results: At day 7, 38% (11 of 29) of patients were feed tolerant. Age, Acute Physiology and Chronic Health Evaluation scores, serum glucose concentrations, and creatinine clearance were comparable between successful and failed feeders. Both plasma erythromycin concentrations at 1 and 7 hrs after drug administration were significantly lower in successfully treated patients compared to treatment failures (1 hr: 3.7 ± 0.8 mg/L vs. 7.0 ± 1.0 mg/L, p = .02; and 7 hr: 0.7 ± 0.3 mg/L vs. 2.8 ± 0.6 mg/L, p = .01). There was a negative correlation between the number of days to failure of feeding and both the 1-hr (r = −.47, p = .049) and 7-hr (r = −.47, p = .050) plasma erythromycin concentrations. A 1-hr plasma concentration of >4.6 mg/L had 72% sensitivity and 72% specificity, and a 7-hr concentration of ≥0.5 mg/L had 83% sensitivity and 72% specificity in predicting loss of response to erythromycin.

Conclusions: In critically ill feed-intolerant patients, there is an inverse relationship between plasma erythromycin concentrations and the time to loss of clinical motor effect. This suggests that erythromycin binding to motilin receptors contributes to variations in the duration of prokinetic response. The use of lower doses of erythromycin and tailoring the dose of erythromycin according to plasma concentrations may be useful strategies to reduce erythromycin tachyphylaxis.