DETECTION OF POOLED SECRETIONS ABOVE ENDOTRACHEAL-TUBE CUFFS - VALUE OF PLAIN RADIOGRAPHS IN SHEEP CADAVERS AND PATIENTS

OBJECTIVE. Mechanical ventilation is thought to increase the risk of n osocomial pneumonia by permitting leakage of bacteria-laden gastro-oro pharyngeal secretions into the upper airways. The goal of this study w as (a) to validate radiographic signs of pooled secretions above endot racheal-tube cuffs (supracuff liquid) in an animal model and (b) to de termine whether suctionable pooled supracuff liquid can be identified on bedside radiographs of intubated patients. MATERIALS AND METHODS. D iagnostic criteria for supracuff liquid were initially validated by th ree radiologists interpreting 162 randomized radiographs made in an in tubated sheep cadaver. The primary criteria included (a) replacement o f the normal supracuff lucency with liquid opacity and (b) the formati on of a sharp interface between the lucency of the upper edge of the c uff below and the liquid above. Graded infusions of 0, 3, 8, 13, and 2 3 mi of saline were made in triplicate into the space above the cuff, and radiographs were evaluated for the presence or absence of saline. The validated diagnostic criteria were used by two radiologists to est imate the frequency with which pooled liquid was seen on portable ches t radiographs of 47 patients undergoing elective short-term postanesth etic mechanical ventilation. RESULTS. In the sheep-cadaver model, the diagnostic criteria for supracuff liquid allowed successful differenti ation between no liquid, a small amount of liquid (3-8 mi), and a larg e amount of liquid (13-23 mi; c2, p < .0001). In a clinical study, rad iographic signs of supracuff liquid were identified in 57% of 47 patie nts. In a small subset of patients (n = 18), the estimated liquid volu me (mean +/- SEM) was calculated to be 7.8 +/- 1.1 mt (range = 2.1-18. 4 mi). CONCLUSIONS. Radiography is a sensitive means of identifying sm all volumes of supracuff liquid above the inflated cuffs of endotrache al tubes. Potentially contaminating liquid pooled above the cuff of an endotracheal tube can be identified in about half of patients undergo ing short-term mechanical ventilation. Our results suggest the suction of the supracuff space may be a reasonable prophylactic maneuver agai nst nosocomial pneumonia. A much larger study is suggested to investig ate the actual relation between pooled supracuff liquid and the develo pment of nosocomial pneumonia.