Upper GI bleeds: second-look endoscopy had no clear effect on rebleeding.

Clinical bottom line (level 1b-)

  1. Patients with upper gastrointestinal bleeding who were given second-look endoscopy, had no clear difference in further bleeding than those not given a second-look.
  2. Patients given second-look endoscopy had no clear difference in mortality.
Villanueva et al: Gastrointestinal Endoscopy 1994; 40: 34-39
Expires October 2002

The study

Unblinded ?concealed randomised trial with intention-to-treat
Setting: general hospital, Spain

104 patients (aged mean 64 years, 69% male) upper gastrointestinal haemorrhage

Excluded if
  • <18 years old
  • no consent given


Control Group: (n = 52, 52 analysed): endoscopy with adrenaline (1:10,000) injected in aliquots of 1.0-2.0 mL, first around and then into the visible vessel or over and around the bleeding area, up to a total of 10.0-15.0 mL, but without second-look
Experimental Group: (n = 52, 52 analysed): second-look endoscopy following protocol as above
Following treatment, patients were admitted to the gastrointestinal bleeding unit and were given ranitidine (50 mg intravenously every 6 hours or 150 mg orally twice daily). Conventional supportive therapy was given as required.
100% followed for ?
Outcome notes:
  • further bleeding : actively bleeding ulcer at repeated endoscopy or vomiting of fresh blood or bloody aspirates after clear lavages by nasogastric tube or passage of fresh melaena plus : (a) signs of haemodynamic instability (systolic blood pressure <100 mmHg and pulse >100 beats per minute), (b) fall in haemoglobin level requiring transfusion to maintain the level around 9 g/dL.

The evidence

Outcome Time to outcome CEREERRRR
(95% CI)		ARR
(95% CI)		NNT
(95% CI)
further bleeding unknown 15
(28.9%)		 11
(21.2%)		 27.0%
(-44.0% to 63.0%)		 7.69%
(-8.89% to 24.3%)		 13
(NNT = 4 to infinity;
NNH = 11 to infinity)
death unknown 2
(3.85%)		 1
(1.92%)		 50.0%
(-435% to 95.0%)		 1.92%
(-4.50% to 8.35%)		 52
(NNT = 12 to infinity;
NNH = 22 to infinity)

Comments

  1. The trial is too small to show any clear difference in further bleeding or mortality between the two groups.

    Citation

    1. Villanueva C, Balanzo J, Torras X, et al: Value of second-look endoscopy after injection therapy for bleeding peptic ulcer: a prospective and randomized trial. Gastrointestinal Endoscopy 1994; 40: 34-39
    Contributor: Clare Wotton and Musab Hayatli, October 1999
    Reviewer:

    Clinical Question.
    Patient upper gastrointestinal bleeds
    Intervention or Exposure second-look endoscopy
    Comparison no second-look
    Outcome haemostasis, further haemorrhage, death