Use of end-tidal carbon dioxide detection to determine correct placement of nasogastric tube : a meta-analysis

Objective: To review the diagnostic accuracy of end-tidal carbon dioxide detection in detecting inadvertent airway intubation and verifying correct placement of nasogastric tubes. Design: We undertook a meta-analysis of diagnostic studies. Study selection: All clinical trials that evaluated the diagnostic accuracy of the colorimetric capnometsry or capnography in detecting inadvertent airway intubation and differentiating between respiratory and gastrointestinal tube placement in adults were included. Electronic databases including MEDLINE, CINAHL, EMBASE, All EBM Reviews, WanFang Data, China Journal Net, Chinese Medical Current Contents, and Index to Chinese Periodical Literature were searched from inception to July 2009. Data extraction and quality assessment: Data were extracted using a form piloted prior to use. Two reviewers independently extracted data relating to purpose of the trial, sample, measurements used, index test results and reference standard. Methodological quality of eligible trials was assessed independently by two reviewers using a modified version of the Quality Assessment of Diagnostic Accuracy Studies for assessing studies of diagnostic accuracy. The accuracy of diagnostic tests is presented in terms of sensitivity, specificity, predictive values, and likelihood ratios. Data synthesis: Nine clinical trials were eligible for inclusion in the meta-analysis. Eight trials were undertaken in intubated and mechanically ventilated patients and two trials also involved participants who were alert or awake. Eight involving a total of 456 nasogastric feeding tube placements investigated the diagnostic accuracy of either colorimetric capnometry or capnography to detect feeding tube placement. One trial involving 195 gastric tube insertions compared the diagnostic accuracy of simultaneous use of a colorimetric carbon dioxide detector and capnography to detect feeding tube placement. The use of colorimetric capnometry or capnography had a sensitivity ranging from 0.88 to 1.00, specificity 0.95 to 1.00, positive likelihood ratio 15.22 to 283.35, negative likelihood ratio 0.01 to 0.25. A summary receiver operator characteristics (SROC) curve was constructed and showed an area under the curve was 0.9959. Three trials reported significant cost savings using end-tidal carbon dioxide detectors. Conclusions: There is evidence to support the use of capnography or colorimetric capnometry for the identification of nasogastric feeding tube placement in mechanically ventilated patients.