Long-term prognosis and a prediction model for acute bowel ischaemia following cardiac surgery

Bowel ischaemia following cardiac surgery is associated with a high postoperative mortality. No scoring system exists as yet to predict this complication following surgery. In addition, the long-term survival is not known. We sought to evaluate in-hospital outcomes and long-term outcomes in bowel ischaemia following cardiac surgery. We also sought to devise a simple risk prediction model for this catastrophic entity.

This was a retrospective study of data entered prospectively into our cardiac surgical database between July 1999 and May 2014. We compared the short- and long-term outcomes of patients who developed bowel ischaemia following cardiac surgery with those who did not develop bowel ischaemia using propensity-matched analysis. We developed a prediction model for bowel ischaemia from logistic regression.

In total, 13 853 patients underwent cardiac surgery. Of these, 85 had confirmed bowel ischaemia following surgery. The in-hospital mortality rate for those with bowel ischaemia was 60%, while in those without bowel ischaemia, the mortality rate was 3% (P < 0.0001). In those bowel ischaemia patients who had a laparotomy for corrective surgery, the in-hospital mortality was significantly less compared with those who did not have a laparotomy (39.2 vs 91.2%, P < 0.0001). The long-term survival for bowel ischaemia at 2, 6 and 10 years was 35% (±5), 31% (±5) and 26% (+/6), respectively. Multivariable analysis revealed that advanced age at surgery, peripheral vascular disease, intra-aortic balloon pump usage, NYHA IV and postoperative atrial fibrillation were the significant (P < 0.005) determinants of developing postoperative bowel ischaemia. We developed a model to predict bowel ischaemia and validated it within our population (c-index = 0.781).

We have shown that whilst bowel ischaemia carries a higher short-term mortality, the long-term mortality is not significantly greater for those few who survive to discharge. We have developed a simple prediction model to identify those at high risk of developing bowel ischaemia following cardiac surgery in order to optimize perioperative strategies in future.