Upper gastrointestinal bleeding following major surgical procedures: Prevalence, etiology, and outcome

T. R. Pollard, W. H. Schwesinger, C. P. Page, P. R. Schauer, K. R. Sirinek

Research output: Contribution to journalArticlepeer-review

6 Scopus citations


With continuing improvements in the medical therapy of peptic ulcer disease, the incidence of primary upper gastrointestinal bleeding (UGIB) has markedly declined. Nonetheless, in a subset of surgical patients, secondary UGIB following a major operation may still be a source of substantial morbidity. To further elucidate this problem, we reviewed 103 cases of overt UGIB following all major surgical procedures conducted in two hospitals between July 1982 and June 1994. The prevalence of postoperative UGIB during this period was 0.39%. The mean interval between initial operation and UGIB was 16 days (range 1-55 days) and there was a high incidence of associated sepsis (26%). The source of bleeding was defined endoscopically in all cases and included gastritis (69.9%), solitary ulcers (17.5%), and other causes (12.6%). Postoperative UGIB (nonvariceal) was most commonly seen following portacaval shunting operations but mortality rates were highest in patients who developed UGIB after cardiovascular operations. We conclude that: (1) postoperative UGIB has become a relatively uncommon but still formidable clinical problem; (2) erosive gastritis continues to be the major source of UGIB but acute ulcers, varices, and other causes contribute to the total; and (3) postoperative UGIB is most likely to be fatal in cardiovascular patients and those who develop concurrent sepsis and/or multiorgan failure.

Original languageEnglish (US)
Pages (from-to)75-78
Number of pages4
JournalJournal of Surgical Research
Issue number1
StatePublished - Jul 15 1996

ASJC Scopus subject areas

  • Surgery


Dive into the research topics of 'Upper gastrointestinal bleeding following major surgical procedures: Prevalence, etiology, and outcome'. Together they form a unique fingerprint.

Cite this