Comparison of blue dye visualization and glucose oxidase test strip methods for detecting pulmonary aspiration of enteral feedings in intubated adults.

STUDY OBJECTIVE To compare the relative utility of blue dye visualization with a glucose oxidase test strip method for detecting aspiration of enteral feedings. DESIGN Tracheally intubated adults were prospectively monitored for aspiration of enteral feedings. SETTING Intensive care units of two community hospitals in Michigan. INTERVENTIONS None. PATIENTS The experimental group consisted of 15 patients receiving enteral feedings. The control group included 14 patients not enterally fed. MEASUREMENTS AND RESULTS Blue food coloring was added to feeding formulas to obtain a visible blue color. At 8-h intervals, tracheal secretions were examined for blue discoloration, followed by measurement of glucose concentration using a calibrated glucose meter. Clinically significant aspiration was defined to require the following: (1) a bloodless positive glucose reading (> or = 20 mg/dl); (2) one or more signs of systemic inflammation; and (3) one or more signs of respiratory deterioration. Eight (53 percent) of 15 patients in the experimental group experienced at least one episode of presumptive aspiration as defined by either a bloodless positive glucose reading or visible blue discoloration of tracheal secretions. Clinically significant aspiration occurred in 5 (33 percent) of 15 patients in whom bloodless glucose readings were positive in 13 (19 percent) of 67 samples; among patients not developing this complication, glucose was found in only 3 (5 percent) of 60 samples; (p = 0.005). Inspecting tracheal secretions for blue dye usually failed to detect aspiration episodes identifiable by the glucose oxidase test strip method (relative sensitivity, 13 percent). Blue dye visualization performed no better among patients developing clinically significant aspiration (relative sensitivity, 15 percent). Patients who developed clinically significant aspiration received more of their enteral feedings in the supine position than patients without this complication (98 percent vs 21 percent; p < 0.001). CONCLUSIONS Inspecting tracheal secretions for blue discoloration failed to detect most episodes of enteral feeding aspiration. Glucose oxidase test strip methods should replace blue dye visualization for detecting aspiration of enteral feedings in intubated adults.