Background: The purpose of this study was to determine the extent to which pepsin and trypsin concentrations in feeding tube aspirates, in addition to pH, contribute to predicting feeding tube position.
Methods: Aspirates from 742 feeding tubes (nasogastric, n = 343; nasointestinal, n = 399) were tested for pH and enzyme concentrations. Also tested were aspirates from two feeding tubes inadvertently positioned into the lung (one in the pleural space and one in the tracheobronchial tree) and 146 samples of tracheobronchial and pleural fluids collected by other methods. Enzyme assays were conducted in a research laboratory. Results of the pH and enzyme tests were compared with radiologic reports of tube location.
Results: Gastric fluid had a mean low pH (4.06), a high mean pepsin concentration (349.1 micrograms/mL), and a low mean trypsin concentration (19.3 micrograms/mL). In contrast, intestinal fluid had a mean high pH (7.40), a high mean trypsin concentration (143.0 micrograms/mL), and a low mean pepsin concentration (24.2 micrograms/mL). Respiratory samples also had a high mean pH (7.89) but contained little or no pepsin or trypsin. Using a logistic regression equation with all three variables to differentiate between respiratory and gastrointestinal placement, it was possible to correctly classify 100% of the respiratory cases and 93.4% of the gastrointestinal cases. Another equation used to differentiate between gastric and intestinal sites was able to classify correctly 91.2% of the gastric cases and 91.5% of the intestinal cases.
Conclusions: The results clearly indicate that laboratory-determined enzyme concentrations in feeding tube aspirates are helpful in predicting tube location. Thus, it is desirable that inexpensive, simple bedside tests be developed so that they can be used in conjunction with pH measurements to help predict tube position.