Confirming nasogastric tube placement: Is the colorimeter as sensitive and specific as X-ray? A diagnostic accuracy study.

BACKGROUND: The effect of delivering enteral nutrition or medications via a nasogastric tube that is inadvertently located in the tracheobronchial tract can cause respiratory complications. Although radiographic examination is accepted as the gold standard for confirming the position of patients' enteral tubes, it is costly, involves risks of radiation, and is not failsafe. Studies using carbon dioxide sensors to detect inadvertent nasogastric tube placements have been conducted in intensive care settings. However, none involved patients in general wards.

OBJECTIVE: The objective of this study was to ascertain the diagnostic measure of colorimeter, with radiographic examination as the reference standard, to confirm the location of nasogastric tubes in patients.

DESIGN: A prospective observational study of a diagnostic test.

SETTING: This study was conducted in the general wards of an approximately 1100-bed acute care tertiary hospital of an Academic Medical Center in Singapore.

PARTICIPANTS: Adult patients with nasogastric tubes admitted to the general wards were recruited into the study.

METHODS: The colorimeter was attached to the nasogastric tube to detect for the presence of carbon dioxide, suggestive of a tracheobronchial placement. The exact location of the nasogastric tube was subsequently confirmed by a radiographic examination.

RESULTS: A total of 192 tests were undertaken. The colorimeter detected carbon dioxide in 29 tested nasogastric tubes, of which radiographic examination confirmed that four tubes were located in the tracheobronchial tract. The colorimeter failed to detect carbon dioxide in one nasogastric tube that was located in the tracheobronchial tract, thus, demonstrating a sensitivity of 0.80 [95% CI (0.376, 0.964)]. The colorimeter detected absence of carbon dioxide in 163 tested nasogastric tubes in which radiographic examination confirmed 160 gastrointestinal and one tracheobronchial placements, demonstrating a specificity of 0.865 [95% CI (0.808, 0.907)]. The colorimeter detected one tracheobronchial nasogastric tube placement that the radiographic examination was misinterpreted.

CONCLUSION: The study found that the use of the colorimeter in the general ward setting was not 100% sensitive or specific in ascertaining the location of a nasogastric tube as previously reported by many studies undertaken in intensive care settings. This is the first study on the use of a colorimeter to confirm the placement of a nasogastric tube in adult patients in the general ward setting. More research on the use of a colorimeter in the general ward setting and its potential use in certain processes for confirming the placement of a nasogastric tube is warranted.