Poor validity of residual volumes as a marker for risk of aspiration in critically ill patients.
BACKGROUND AND AIMS: Elevated residual volumes (RV), considered a marker for the risk of aspiration, are used to regulate the delivery of enteral tube feeding. We designed this prospective study to validate such use. METHODS: Critically ill patients undergoing mechanical ventilation in the medical, coronary, or surgical intensive care units in a university-based tertiary care hospital, placed on intragastric enteral tube feeding through nasogastric or percutaneous endoscopic gastrostomy tubes, were included in this study. Patients were fed Probalance (Nestle USA) to provide 25 kcal/kg per day (to which 10 yellow microscopic beads and 4.5 mL of blue food coloring per 1,500 mL was added). Patients were randomized to one of two groups based on management of RV: cessation of enteral tube feeding for RV >400 mL in study patients or for RV >200 mL in controls. Acute Physiology and Chronic Health Evaluation (APACHE) III, bowel function score, and aspiration risk score were determined. Bedside evaluations were done every 4 hrs for 3 days to measure RV, to detect blue food coloring, to check patient position, and to collect secretions from the trachea and oropharynx. Aspiration/regurgitation events were defined by the detection of yellow color in tracheal/oropharyngeal samples by fluorometry. Analysis was done by analysis of variance, Spearman's correlation, Student's t-test, Tukey's method, and Cochran-Armitage test. RESULTS: Forty patients (mean age, 44.6 yrs; range, 18-88 yrs; 70% male; mean APACHE III score, 40.9 [range, 12-85]) were evaluated (21 on nasogastric, 19 on percutaneous endoscopic gastrostomy feeds) and entered into the study. Based on 1,118 samples (531 oral, 587 tracheal), the mean frequency of regurgitation per patient was 31.3% (range, 0% to 94%), with a mean RV for all regurgitation events of 35.1 mL (range, 0-700 mL). The mean frequency of aspiration per patient was 22.1% (range, 0% to 94%), with a mean RV for all aspiration events of 30.6 mL (range, 0-700 mL). The median RV for both regurgitation and aspiration events was 5 mL. Over a wide range of RV, increasing from 0 mL to >400 mL, the frequency of regurgitation and aspiration did not change appreciably. Aspiration risk and bowel function scores did not correlate with the incidence of aspiration or regurgitation. Blue food coloring was detected on only three of the 1,118 (0.27%) samples. RV was < or =50 mL on 84.1% and >400 mL on 1.4% of bedside evaluations. Sensitivities for detecting aspiration per designated RV were as follows: 400 mL = 1.5%; 300 mL = 2.3%; 200 mL = 3.0%; and 150 mL = 4.5%. Low RV did not assure the absence of events, because the frequency of aspiration was 23.0% when RV was <150 mL. Raising the designated RV for cessation of enteral tube feeding from 200 mL to 400 mL did not increase the risk, because the frequency of aspiration was no different between controls (21.6%) and study patients (22.6%). The frequency of regurgitation was significantly less for patients with percutaneous endoscopic gastrostomy tubes compared with those with nasogastric tubes (20.3% vs. 40.7%, respectively; p = .046). There was no correlation between the incidence of pneumonia and the frequency of regurgitation or aspiration. CONCLUSIONS: Blue food coloring should not be used as a clinical monitor. Converting nasogastric tubes to percutaneous endoscopic gastrostomy tubes may be a successful strategy to reduce the risk of aspiration. No appropriate designated RV level to identify aspiration could be derived as a result of poor sensitivity over a wide range of RV. Study results do not support the conventional use of RV as a marker for the risk of aspiration.