Upper GI bleed: endoscopic therapies reduce rebleeding, need for surgery and death.

Clinical bottom line (level 1a)

  1. Patients with acute non-variceal upper GI bleeding who have endoscopic therapy are less likely to rebleed, require surgery or die.
Cook et al: Gastroenterology 1993; 102: 139-148
Expires July 2003

The study

Systematic review of randomised controlled trials of
  • Patients: acute non variceal upper GI haemorrhage
  • Intervention: endoscopic therapies (laser, thermal contact devices- monopolar, multipolar and heater probe- injection therapy) compared with control
  • Outcome: further bleeding (persistent or recurrent), surgery, death


    • Articles found in ? using MEDLINE, Science Citations Index, 1966 to 1991 (search terms: hemorrhage (gastrointestinal), endoscopy and clinical trials ) and manual search of reference lists, contacting experts in the field and other bibliographic databases

    Selection criteria: as above
    Appraisal criteria: detailed in text; performed by the two authors independently
    Articles excluded if:

    30 trials involving 2412 patients
    There was no significant heterogeneity.

    The evidence

    Outcome Time to outcome CER OR
    (95% CI)    		 NN?
    (% CI)
    any endoscopy- further bleeding unknown /
    (%)    		 0.38
    (0.32 to 0.45)
    any endoscopy- surgery unknown /
    (%)    		 0.36
    (0.28 to 0.45)
    any endoscopy- death unknown /
    (%)    		 0.55
    (0.40 to 0.76)
    thermal contact- further bleeding unknown /
    (%)    		 0.32
    (0.22 to 0.41)
    thermal contact- surgery unknown /
    (%)    		 0.31
    (0.19 to 0.43)
    thermal contact- death unknown /
    (%)    		 0.67
    (0.39 to 1.14)
    laser- further bleeding unknown /
    (%)    		 0.58
    (0.38 to 0.69)
    laser- surgery unknown /
    (%)    		 0.58
    (0.40 to 0.80)
    laser- death unknown /
    (%)    		 0.49
    (0.30 to 0.81)
    injection- further bleeding unknown /
    (%)    		 0.23
    (0.12 to 0.45)
    injection- surgery unknown /
    (%)    		 0.18
    (0.11 to 0.32)
    injection- death unknown /
    (%)    		 0.50
    (0.22 to 1.12)

    • Complication rates were:
      • bleeding requiring surgery: 0.4% (laser), 0.39% (thermal), 0.4% (injection)
      • perforation: 0.9% (laser), 0.7% (thermal), 0% (injection)
      • these figures are lower than the rate of ~1% reported from large series.

    Comments

    1. No control rates were provided nor length of follow-up. NNTs cannot be calculated from this data.

    Citation

    1. Cook DJ, et al: Endoscopic therapy for acute non variceal upper gastrointestinal hemorrhage: a meta-analysis. Gastroenterology 1993; 102: 139-148
    Contributor: Alan Townsend, Sharon Straus and Chris Ball, July 2000
    Reviewer:

    Clinical Question.
    Patient upper GI bleed
    Intervention or Exposure endoscopy
    Outcome rebleed, death, surgery