Upper GI bleed: peptic ulcers: an aggressive surgical policy increased operations but had no clear effect on survival

Clinical bottom line (level 1b)

  1. Patients with a bleeding peptic ulcer who had surgery based on early intervention were more likely to have an operation than patients who had surgery based on delayed intervention (NNH = 3 at unknown)
  2. There is no clear effect on mortality.
Morris et al: British Medical Journal 1984; 288: 1277-1280
Expires October 2002

The study

Unblinded ?concealed randomised trial with intention-to-treat
Setting: acute hospital, UK

142 patients (aged , % male) admitted with upper GI bleeding or who bled while in hospital, and with an endoscopically-proven peptic ulcer

Excluded if
  • more than one possible source of bleeding
  • on anticoagulants
  • aged < 25
  • judged to ill to have surgery
  • malignancy suspected
  • undergone previous gastric surgery

Note:
  • All patients were managed by a 'bleeding team' and had endoscopy within 6 to 12 hours, followed by iv cimetidine 200 mg q6h until able to take 400 mg bd orally.
  • Patients were stratified for age (>60) and site of ulcer (duodenal/gastric) before randomisation.


Control Group: (n = 71, 71 analysed): delayed surgery: if 8 units of blood or plasma expander required to correct acute blood loss in 24 hours; 2 rebleeds in hospital or persistent bleeding requiring transfusion of 12 units in 48 hours or 16 units in 72 hours
Experimental Group: (n = 71, 71 analysed): early surgery: 4 units of blood or plasma expander required to correct blood loss in 24 hours, one rebleed in hospital, endoscopic stigmata (active bleeding, visible vessel, adherent clot, spots) or previous upper GI haemorrhage plus a 2 year history of dyspepsia
Operations were performed by senior registrars or consultants.
100% followed for 15 months

The evidence

All patients
Outcome Time to outcome CEREERRRR
(95% CI)		ARR
(95% CI)		NNT
(95% CI)
surgery performed weeks 15
(21.1%)		 42
(59.2%)		 -180%
(-360% to -72%)		 -38.0%
(-52.9% to -23.2%)		 -3
(-4 to -2)
death weeks 7
(9.86%)		 3
(4.23%)		 57%
(-59% to 88%)		 5.63%
(-2.73% to 14.0%)		 18
(NNT = 37 to infinity;
NNH = 7 to infinity)

Comments

  1. Deaths in the delayed group were due to massive haemorrhage, MI or stroke, sepsis, and respiratory infection.
  2. The study is too small to show a significant effect of the surgical policy on mortality. Subgroup analysis does not change the results.
  3. This study was performed before the era of endoscopic therapy

Citation

  1. Morris DL, Hawker PC, Brearley S, et al: optimal timing of operation for bleeding peptic ulcer: prospective randomised trial. British Medical Journal 1984; 288: 1277-1280
Contributor: Alan Townsend and Chris Ball, October 1999
Reviewer: .

Clinical Question.
Patient bleeding peptic ulcer upper GI bleed
Intervention or Exposure early surgery
Comparison delayed surgery
Outcome surgery, death