Introduction As survival of patients with left ventricular assist devices (LVADs) and cardiac transplantation improves, surgeons must increasingly decide whether and when to operate on them.

Method We conducted a retrospective review of abdominal and colorectal operations performed on patients with heart transplants or LVADs at a tertiary care hospital in Boston, MA between 2003–2013. Data were collected on comorbidities, anticoagulation and immunosuppression status, intra-and post-operative complications, need for blood product transfusion, type of surgery performed, 24-hr and thirty day mortality, and time post-transplant or LVAD insertion. Statistical testing for significance was done using the two-sided Fisher’s exact test.

Results Over the 11-year period, 13 patients with LVADs and 32 patients with heart transplants underwent a total of 67 colorectal (22%) and other abdominal (78%) operations. The median time between LVAD insertion or heart transplant and abdominal surgery was 175 days and 765 days respectively. There was a tendency towards a higher proportion of elective (vs emergency) cases in heart transplant (74%) vs LVAD patients (53%) (p = 0.13). LVAD patients were more likely to be anticoagulated, and had an 18% incidence of postoperative bleeding, with a significantly greater proportion of patients requiring blood transfusion within the first postoperative week (47%) compared to transplant patients (8%) (p = 0.001). Venous thromboembolic (VTE) events were also significantly more common in the LVAD group (21%) vs transplant group (0%) (p = 0.01). Subgroup analysis revealed the incidence of VTE was significantly decreased in the LVAD group bridged for surgery (p = 0.0008). Despite all transplant patients being on immunosuppression medication, and 46% on steroids, there was no statistically significant difference in surgical site infections between the transplant (6%) and LVAD (12%) groups (p = 0.59). Mortality within 24 h tended to be higher in the LVAD group (13%) vs transplant group (0%), with borderline significance (p = 0.06). 30-day mortality was significantly higher in the LVAD group (21%) vs transplant group (0%) (p = 0.01).

Conclusion Surgical outcomes were better in patients who had undergone heart transplantation compared to those with LVADs, especially with respect to the need for postoperative transfusion, the incidence of VTE, and 24-hr and 30-day mortality. The observed tendency to perform fewer elective cases in LVAD patients compared to transplant patients may be a reflection of clinical intuition of this finding. Surgeons should assess these higher-risk patients on a case-by-case basis. Surgeons should also bridge anticoagulation for LVAD patients whenever possible to reduce the risk of VTE.

Disclosure of interest None Declared.