Upper GI bleed: Gastroccult testing of NG aspirate may help exclude it.

Clinical bottom line (level 4)

  1. Physician assessment of bile or blood in a nasogastric aspirate was not very helpful in determining if a patients is having a GI bleed.
  2. A whitish or yellow-green aspirate made GI bleeding less likely.
  3. A negative Gastroccult test made a GI bleed much less likely.
Cuellar : Archives of Internal Medicine 1990; 150: 1381-1384
Expires January 2003

The study

Setting: acute hospital, USA

62 patients (aged ?, ?% male) suspected acute upper GI bleeding, defined as the presence of an acute drop in haematocrit of at least 0.05, haematemesis, melena or the finding of gross blood within the stomach, as detected by aspiration via a nasogastric tube before endoscopy

Non-independent unblinded reference standard, applied in all patients from a consecutive appropriate spectrum.
Reference standard:
  • endoscopy
Diagnostic test:
  • physician assessment of nasogastric aspirate for blood
  • physician assessment of nasogastric aspirate for bile
  • appearance of nasogastric aspirate
  • Gastroccult test

The evidence

pre-test probability of upper GI bleed: 39%, (95% CI: 27% to 51%)

diagnostic test upper GI bleed no bleed LR+
(95% CI)		 post-test probability LR-
(95% CI)		 post-test probability
blood in aspirate 19 17 1.8
(1.2 to 2.7) 		53% 0.38
(0.16 to 0.87) 		19%
total 24 38


diagnostic test upper GI bleed no bleed LR+
(95% CI)		 post-test probability LR-
(95% CI)		 post-test probability
bile in aspirate 11 10 1.9
(0.94 to 3.7) 		52% 0.70
(0.46 to 1.1) 		29%
total 23 39


diagnostic test upper GI bleed no bleed LR+
(95% CI)		 post-test probability LR-
(95% CI)		 post-test probability
Gastroccult positive 23 7 5.2
(2.7 to 10) 		77% 0.051
(0.0075 to 0.35) 		3%
total 24 38


diagnostic test upper GI bleed no bleed LR
(95% CI)		 post-test probability
grossly bloody aspirate 16 10 2.5
(1.4 to 4.6) 		62%
slightly bloody aspirate 4 6 1.1
(0.33 to 3.4) 		40%
aspirate looked like coffee grounds 3 5 0.95
(0.25 to 3.6) 		38%
aspirate whitish or yellow-green 1 17 0.093
(0.013 to 0.66) 		6%
total 24 38

Comments

  1. Reference standard not applied blindly, but GI bleed is a fairly objective outcome.
  2. No specific criteria for physicians to use to assess nasogastric aspirate.

Citation

  1. Cuellar R, : Gastrointestinal tract hemorrhage. The value of a nasogastric hemorrhage. Archives of Internal Medicine 1990; 150: 1381-1384
Contributor: Alan Townsend, Sharon Straus and Chris Ball, July 2000
Reviewer:

Clinical Question.
Patient suspected GI bleed
Intervention or Exposure nasogastric aspirate colour, haemoccult test, guiac test
Outcome diagnosis of bleed