This section describes safety outcomes from the published literature that the committee considered as part of the evidence about this procedure. For more detailed information on the evidence, see the interventional procedure overview.
5.1 In a non-randomised retrospective comparative study of 578 patients (United Network of Organ Sharing [UNOS] database analysis), rates of deaths were not significantly different between total artificial heart (TAH) and biventricular assisted device (BIVAD) groups: 10% (22/212) compared with 12% (45/366), p=0.7 while on support; and 30% (64/212) compared with 30% (111/366), p=0.9 after transplantation. The causes of death while on device support (infection, multi-organ failure, stroke or haemorrhage) and after heart transplantation (acute rejection, infection, cardiac arrest, multi-organ failure and stroke) for both groups were also similar. In a non-randomised prospective comparative study of 130 patients, rates of death before transplantation were 21% (17/81) in the TAH group compared with 54% (19/35) in the control group (p value not reported). Causes of the 17 deaths before transplantation in the TAH group were multi-organ failure (7), procedural or technical complications (4), bleeding (2), sepsis (2), congestive heart failure (1) and pulmonary oedema (1). In the same group after transplantation, there were 6 deaths (3 graft failure, 1 sepsis, 1 procedural or technical complication and 1 multi-organ failure). Rates of deaths in a non-randomised retrospective comparative study of 148 patients were not significantly different between TAH and BIVAD support groups while on support (37% compared with 39%; p=0.87). Death occurred in 32% (32/101) of patients in a case series of 101 patients with TAH implantation; 70% were within the first 14 days. Causes of deaths were multi-organ failure (13), pneumonia or pulmonary oedema (6), sepsis (5), neurologic injury (4, including 1 stroke, 1 hypoxic damage from hypotension and 2 intracranial haemorrhage), pancreatic abscess (1), small intestinal ischaemia (1), disseminated intravascular coagulopathy (1), and disseminated coccidiodomycosis (1).
5.2 Bleeding events were reported in 62% (59/95) of patients in the TAH group in the non-randomised prospective comparative study of 130 patients. Of these events, 50% occurred during TAH implantation and were mainly tamponade or mediastinal bleeding (needing surgery and blood transfusion) within 21 days. Two patients died from bleeding: 1 during TAH implantation and 1 during heart transplantation). Bleeding (from various sites) at a mean of 4 days was reported in 43% (43/101) of patients in the case series of 101 patients. Reoperations for haemorrhage (mediastinal explorations) were done in 25% (25/101) of patients. The surgical revision rate for bleeding or haematoma was significantly lower in the TAH group than the BIVAD group in the non-randomised comparative study (23% [17/81] compared with 42% [28/67] respectively; p=0.03). Surgical re-exploration for bleeding, haematoma or infection was reported in 39% (35/90) of patients while on device support in the case series of 90 patients. Haemorrhagic events (at a median time of 249 days) were reported in 14% (7/47) of patients in the case series of 47 patients with TAH support for more than 1 year. These included cerebral haemorrhage (n=3), subarachnoid haemorrhage (n=2), and gastrointestinal bleeding (n=2).
5.3 Device malfunction events were reported in 17% (16/95) of patients in the TAH group in the non-randomised prospective comparative study of 130 patients. One patient died because of perforation in 1 of the layers of the device's left ventricular diaphragm on day 124. Fitting complications were reported in 5% (5/95) of patients in the TAH group in the same study; 2 patients died because of poor fitting and 3 patients had repeat surgery to reposition the device. Device failure was reported in 10% (5/47) of patients in the case series of 47 patients with TAH support for more than 1 year; there were 3 membrane ruptures (2 patients died), 1 air hole in the driveline and 1 lower pump output. Device malfunction (technical problems with the alarm and computer monitoring system needing console change) was reported in 1 patient during hospitalisation in the case series of 27 patients. Malfunctions were reported in 25% (3/12) of patients discharged home after TAH implantation. In 2 patients, an air leak occurred in the driveline to the ventricle, which was sealed with a silicon band. In 1 patient, alarm triggering was caused by auricular compression during some movements (stretching and yawning) and the patient was advised to avoid them.
5.4 The renal failure rate was significantly higher in the TAH support group than the BIVAD support group after implantation (24% [52/212] compared with 10% [35/366]; p<0.0001) and after transplantation (26% [56/212] compared with 14% [49/366]; p=0.0001) in the non-randomised retrospective comparative study (UNOS database analysis). Renal failure (needing postoperative renal replacement therapy) was reported in 64% (30/47) of patients in the case series of 47 patients with TAH support for more than 1 year. Recovery of renal function occurred in 73% (22/47) of patients but therapy continued in 17% (8/47) of patients.
5.5 The infection rate after implantation was significantly lower in the TAH group than the BIVAD group (22% [46/212] compared with 28% [104/366]; p=0.005) in the non-randomised retrospective comparative study (UNOS database analysis). Infections (7 blood, 50 respiratory, 5 mediastinal, 28 genitourinary, 12 gastrointestinal, 17 driveline and 6 in catheters) were reported in 77% (73/95) of patients in the TAH group in the non-randomised prospective comparative study of 130 patients. Infections contributed to death in 7 patients and delayed transplantation in 5 patients. Infections (commonly in the lungs and urinary tract) were reported in 64% (64/101) of patients in the case series of 101 patients. Rates of postoperative mediastinitis (TAH 12% [9/81] compared with BIVAD 5% [3/67]; p=0.14) or driveline infections (1 patient with TAH compared with 2 patients with BIVAD; p=0.61) were similar in both groups during support in the non-randomised comparative study of 148 patients.
5.6 Neurologic events (10 stroke, 4 transient ischaemic attack, 5 anoxic encephalopathy, 1 metabolic encephalopathy, 4 seizure and 1 syncope) were reported in 27% (26/95) of patients in the TAH group in the non-randomised prospective comparative study of 130 patients. Transplantation was delayed in 6 patients. Neurologic events were reported in 16% (16/101) of patients in the case series of 101 patients. Strokes were reported in 8% (8/101) of patients (5 within 9 days). Peripheral emboli (1 celiac artery, 2 spleen, 1 superior mesenteric artery, 2 kidney, 2 retina) were reported in 8% (8/101) of patients, 4 of whom died. Strokes rates were not significantly different between the TAH and BIVAD groups during support in the non-randomised prospective comparative study of 148 patients (TAH 12% [9/81] compared with BIVAD 24% [16/67]; p=0.08). Strokes were reported in 10% (9/90) of patients while on device support in the case series of 90 patients with TAH implantation. Thromboembolic events (at a median time of 500 days) were reported in 19% (9/47) of patients in the case series of 47 patients with TAH support for more than 1 year; 6 had a transient ischaemic attack and 3 had a major cerebrovascular accident with hemiparesis or aphasia.
5.7 Technical or procedural problems (obstruction of the mechanical tricuspid valve of the TAH because of migration of the central venous catheter) were reported in 3% (3/95) of patients in the TAH group in the non-randomised prospective comparative study of 130 patients. This caused death in 1 patient. Catheter entrapment of a central line in the tricuspid valve was reported in 2 patients in the case series of 101 patients with TAH implantation. Both these patients had irreversible brain damage from device arrest and died.
5.8 In addition to safety outcomes reported in the literature, specialist advisers are asked about anecdotal adverse events (events which they have heard about) and about theoretical adverse events (events which they think might possibly occur, even if they have never done so). For this procedure, specialist advisers did not describe any anecdotal or theoretical adverse events.