Gastrointestinal bleeding rates in recipients of nonpulsatile and pulsatile left ventricular assist devices

Pulsatile and nonpulsatile left ventricular assist devices are effective in managing congestive heart failure. Despite early evidence for clinical efficacy, the long-term impact of nonpulsatile flow on end-organ function remains to be determined. Our goal was to compare rates of gastrointestinal bleeding in nonpulsatile and pulsatile device recipients.

In a retrospective review of 101 left ventricular assist device recipients (55 nonpulsatile, 46 pulsatile) from October 31, 2003, to June 1, 2007, at a single center, gastrointestinal bleeding was defined as guaiac-positive stool with hemoglobin drop requiring transfusion of at least 2 units of packed red blood cells. To assess bleeding risk outside the initial postoperative course, any patients with a device in place for 15 days or less was excluded.

Twelve nonpulsatile and 3 pulsatile left ventricular assist device recipients had gastrointestinal bleeding 16 days or longer after device implantation. The event rates were 63 events/100 patient-years for nonpulsatile devices and 6.8 events/100 patient-years for pulsatile devices (P = .0004). This difference persisted for bleeding occurring 31 days or longer after device implantation, with 46.5 events/100 patient-years for nonpulsatile devices versus 4.7 events/100 patient-years for pulsatile devices (P = .0028). Mortalities were similar between groups (15% nonpulsatile vs 17% pulsatile, P = .6965).

Patients with nonpulsatile left ventricular assist devices appear to have a higher rate of gastrointestinal bleeding events than do pulsatile left ventricular assist device recipients. Further prospective evaluation is needed to determine potential etiologies and strategies for reducing gastrointestinal bleeding in this population.