Secondary Logo

Journal Logo

Clinical Cardiovascular

Regional Referral System for Patients with Acute Mechanical Support: Experience at The Cleveland Clinic Foundation

Gonzalez-Stawinski, Gonzalo V.*; Chang, Albert S. Y.*; Navia, Jose L.*; Banbury, Michael K.*; Buda, Tiffany*; Hoercher, Kathy*; Starling, Randall C.; Taylor, David O.; Smedira, Nicholas G.*

Author Information
doi: 10.1097/01.mat.0000225265.11371.ed
  • Free


The widespread use of mechanical assist devices in the community has allowed the stabilization of patients in shock after myocardial infarction and cardiac surgery.1–5 Although these devices are easy to implant, their successful management requires the commitment of a significant amount of time and resources. Transfer to “hub” facilities experienced with long-term ventricular assist device (VAD) support and heart transplantation increases therapeutic options available and may improve survival.6–9 However, this approach is costly, separates the patients from family and friends, and is often used in desperation. In an effort to identify which patients are best served by referral, we reviewed our experience with patients transferred on mechanical support.

Materials and Methods

Selection Criteria

A retrospective review of our cardiac assist device database identified patients transferred on mechanical cardiac support to The Cleveland Clinic Foundation. Excluded from this study were patients who were transferred with any kind of cardiac pathology and who had an intra-aortic balloon pump as their only method of circulatory support. Variables collected included demographics, type of surgical procedure at the referring institution (RI) and urgency (e.g., elective or emergent), number of days at RI, additional support devices in place, and short and long-term outcomes. The Social Security Death Index (SSDI) was used to confirm mortality in those patients who were eventually discharged. If the query using SSDI was negative, then the patient was assumed to be alive.


Complex surgical procedures were defined as open heart surgical procedures with surgical risks greater than primary coronary artery bypass or single valve replacement. Complications were defined as follows: neurologic or stroke if any significant motor or sensory deficit was present upon evaluation after the discontinuation of narcotics or sedatives; multisystem organ failure as renal, hepatic, and pulmonary failure in addition to the cardiac failure; renal failure (RF) when metabolic derangements persisted requiring the institution of continuous or intermittent hemodialysis; pneumonia required the presence of a positive sputum culture; limb ischemia defined as the absence of pulses in a single limb with clinical findings suggesting compromised circulation; deep vein thrombosis as a single swollen limb with Doppler ultrasound confirmation of a thrombus; and pulmonary embolus as the presence of a filling defect within a major or branch pulmonary artery as determined by contrast spiral computed tomography. Newly identified medical conditions were defined as medical conditions identified on arrival that were not described by the referring team or not considered by the accepting team at the time of transfer.

VAD Implantation

All VADs were implanted after full systemic heparinization and instituting cardiopulmonary bypass. For patients receiving a HeartMate device, pump pockets were created in the preperitoneal space behind the rectus muscle. Inflow to the VAD was achieved by an apical ventriculotomy, while for outflow the ascending aorta was used. At times the ABIOMED graft was cut to sew the HeartMate graft to it. A similar VAD cannulation strategy was used for patients receiving a Thoratec device.

All patients receiving HeartMate devices were given 325 mg aspirin alone for anticoagulation. Sometimes, HeartMate patients were treated with warfarin for other reasons (e.g., atrial fibrillation or deep vein thrombosis). Thoratec devices were managed with heparin, aspirin, and frequently another antiplatelet agent before transition to chronic warfarin, with a target International Normalized Ratio of 2.5 to 3.5.

Statistical Analysis

Univariate analysis was performed between survivors and nonsurvivors using standard t-tests. Statistical significance was defined as a p value less than 0.05.



Between January 1995 and September 2003, 39 patients were transferred for continuation of care. Most transfers occurred from within our region (26 patients arrived from institutions within Ohio). There were 29 men and 10 women. The average age for the entire group was 51 years (range 24–68 years).

Forty-six percent of patients had a history of coronary artery disease. Remote myocardial infarction was reported in 36% of patients, and 32% of patients had a history of diabetes mellitus. Other significant medical histories included a history of smoking in 29% of patients, prior percutaneous transluminal coronary angioplasty in 18% of patients, and a history of ventricular fibrillation/tachycardia in another 18 patients. Aortic valve disease occurred in 10% of patients. The average number of days, for the entire group, at the referring institution before transfer was 3 days (range 1–12 days).

Indications and Type of Support

Most patients (85%) required postcardiotomy support, and six patients (15%) suffered post–myocardial infarction cardiogenic shock. The procedure preceding support was primary coronary artery bypass grafting 52% of the time with 12 (31%) patients having more complex procedures (Table 1). Surgery was emergent in 62% of cases and an intra-aortic balloon pump was used in 67% of cases. The ABIOMED BVS 5000 was used in 33 patients (85%); 17 left VADs, 1 right VAD, and 15 biventricular assist devices. Six patients (15%) were supported with Biomedicus Rollerpump extracorporeal membrane oxygenator (ECMO). Central cannulation was used in all ECMO patients.

Table 1
Table 1:
Pretransfer Cardiac Surgical Procedures

Hospital Outcomes

Newly identified medical conditions were noted on arrival to the hub center in 26 patients. These included 7 patients (18%) in multisystem organ failure, 7 patients (18%) with neurologic complications, 4 patients (10%) with an open chest, 3 patients (8%) with coagulopathy, and 3 patients (8%) with pulmonary failure (Table 2).

Table 2
Table 2:
Newly Identified Medical Conditions on Arrival

There were no in-transit deaths; however, one patient was pronounced brain dead shortly after arrival and support was withdrawn after discussion with the family.

Postadmission complications occurred in 35 patients (90%) (Table 3). Neurologic complications after admission occurred in 9 patients (23%); 6 ischemic infarcts, 2 encephalopathies, and 1 cerebral edema. Complications associated with VAD implantation included 14 cases of bleeding, 4 cases of cannula malfunction, 4 cases of sternal wound infection, and 3 cases of abdominal wound infection.

Table 3
Table 3:
Postadmission Complications

There were a total of 15 survivors (38%, Figure 1). Among these survivors, 7 patients were successfully weaned from their mechanical support, and 8 were converted to a long-term device (LTD). Figure 2 depicts the algorithm used to decide who is converted to a LTD.

Figure 1.
Figure 1.:
Outcomes after admission to hub facility. Out of 39 patients transferred for continuous management of a ventricular assist device, 15 (38%) survived. Among the 15 survivors, 7 patients were successfully weaned from the device and 8 were converted to long-term devices. Six of these survivors were subsequently transplanted. Nonsurvivors were rarely weaned from their devices. LTD, patient converted to a long-term device; HTX, heart transplant; LTS, patient placed on long-term-support.
Figure 2.
Figure 2.:
Algorithm for the management of patients with temporary mechanical devices (TMD). Assessment of hemodynamic and neurologic status occurs concomitantly. Only patients who are stable both hemodynamically and neurologically are evaluated for weaning from the TMD. ICU, intensive care unit; CT, computed tomography; BI, brain injury; LTD, long-term device.

Six LTD patients underwent successful heart transplantation and two patients remained on long-term support. Average time from admission to transplantation was 112 days (range 48–188 days). The 30-day survival for this group of patients was 100%, with a 1-year survival of 87% (Table 4). With regard to the nonsurvivor group, 16 died while on mechanical support and 3 were converted to an LTD. However, two patients eventually died of sepsis and one patient died of multisystem organ failure. (Figure 1)

Table 4
Table 4:
Short-Term Outcomes among Survivors

Differences between Survivors and Nonsurvivors

Table 5 depicts differences between survivors and nonsurvivors. Survivors were younger (p = 0.03), were less likely to be female (p = 0.8), and had less comorbidity than nonsurvivors. In addition, survivors were less likely to have complex surgical procedures (7% vs. 50%, p < 0.001), require biventricular support (27% vs. 45%, p = 0.11), were transferred earlier (2 days vs. 4 days, p = 0.15), and traveled shorter distances (143 vs. 162 miles, p = 0.65). Finally, survivors were in better condition when compared with nonsurvivors. Not a single surviving patient presented in MOSF, only 1 patient was coagulopathic at admission, and only 1 patient in this subgroup had evidence of neurologic dysfunction. In contrast, nonsurvivors were sicker and were more likely to present in MSOF, neurologic deficits, and coagulopathy.

Table 5
Table 5:
Comparison of Admission Variables between Survivors and Nonsurvivors

Cost Associated with Care

The average charges for patients cared for during the study period was $178,060 (range $12,438 to $925,660). The average reimbursement was $106,061 (range $3,461 to $631,950). The average charges for survivors was $282,651 (range $88,359 to $925,660) and an average reimbursement for survivors was $220,791 (range $59,625 to $631,950). The average charges for nonsurvivors was $114,587 (range $12,438 to $463,949) with an average reimbursement $60,539 (range $0 to $303,704).


The purpose of regional referral networks is to provide large geographic areas with rapid access to expensive, resource-intensive, specialized care. This has been particularly effective in trauma and neonatal intensive care programs.10–14 Mechanical circulatory support has recently utilized this concept, and the majority of hospitals provide short-term resuscitation and support. For optimal outcomes, these hospitals must be affiliated with centers who are experienced in short-term and long-term mechanical circulatory support along with transplantation. Because not all cardiogenic shock patients will gain benefit from this strategy, it is important that the overall system understand factors associated with success.

First, the more complex the procedure, the less likely the patient is to survive. In a recent report by Doll et al.1 with 219 consecutive patients with refractory postoperative cardiogenic shock, overall survival was 24% for the entire group. Patients undergoing more than coronary revascularization had 30-day mortalities ranging from 82 to 95%. Similarly, in a study of 56 patients needing LVAD, Oz et al. 15 noted that a history of prior cardiac operation (i.e., redo) was associated with a twofold increase in mortality. In our patient cohort, a significant difference in mortality was detected when the complexity of cardiac procedures before VAD implantation was compared between survivors and nonsurvivors. Fifty percent of nonsurvivors had complex surgical procedures compared with 7% of survivors.

Many reports have validated that the presence of organ dysfunction at admission portrays a grim prognosis. Thus, it is not surprising that in our experience it was an evident factor associated with mortality. Orime et al. and Minev et al. reported that multisystem organ failure (MSOF) was a common cause of death in their cohort of patients supported with VAD.16,17 Similarly, in a report from Germany, MSOF was the major cause of death in postcardiotomy shock patients with mechanical cardiac support.18 Just one organ system can reduce survival significantly for this patient population. For instance, in the group of patients reported by McBride et al., 100% mortality was observed for those patients who had at least one organ failure in addition to the cardiovascular system.9 In our own experience, MSOF occurred in 30% of nonsurvivors at admission, with 41% of patients presenting with a single system affected.

Third, it seems that the longer the time between the initial surgical procedure and transfer to a hub facility, the lower the chance of survival. For instance, Helman et al.6 used a 72-hour cut-off when accepting patients in transfer for continued care of patients with a temporary assist device. In their opinion, this “golden period” was put in place to reduce the chance of complications, such as sepsis, associated with VAD use. Conversely, the experience of McBride and colleagues included 16 patients whose average time to transfer was 1.6 days (range 1 hour to 5 days). However, because these authors only waited hours from time of VAD implantation to transfer, time did not seem to play a role with regard to survival.9 In our experience, survivors waited a shorter time to be transferred when compared with nonsurvivors, although this difference did not reach statistical significance.

Fourth, some have suggested that biventricular failure portrays a worse prognosis than those with a single ventricle process despite the expeditious implantation of devices to allow ventricular recuperation. Entwistl et al.2 noted that in 17 patients in cardiogenic shock after acute myocardial infarctions, biventricular support was associated with 100% mortality whereas survival for univentricular support patients was 54%. A similar experience was reported by Golding et al.19 in 91 patients with postcardiotomy shock. In our own experience, 45% of patients with biventricular support died. This underscores the grim prognosis of this extent of disease, and suggests that if these patients are indeed to be saved, then transplantation is the only plausible alternative; therefore, patients are better served with devices that can be used as long-term support.

Study Limitations

In addition to the results presented herein, the limitations of this report need to be pointed out. First, this is a retrospective observational study, which carries with it all the limitations of a retrospective study design. Second, this study contains a small number of patients, which might render the magnitude of our findings insignificant if a larger and more homogenous group of patients were available for analysis. In addition, the limited patient number excludes the possibility of performing a multivariable analysis, which would allow the identification of predictive variables to construct a therapeutic algorithm.


As a policy, we accept almost all patients referred to our institution on short- term support. As more devices become available, how we manage and support patients will change. The focus must be on reducing MSOF and identifying ways to enhance myocardial recovery. This experience, supplemented with a large ECMO experience and a review of literature, can now identify patients with the greatest chance for survival. In general, a younger patient who undergoes a simple procedure, has no organ failure, and is a transplant candidate has a very good chance of surviving. Conversely, an older patient undergoing a complex operation with or without organ failure is much less likely to survive. This does not mean that transfer should be denied, but rather that a realistic estimation of survival should be presented to the referring physician and the family.


1. Doll N, Kiaii B, Borger M, et al.: Five-year results of 219 consecutive patients treated with extracorporeal membrane oxygenation for refractory postoperative cardiogenic shock. Ann Thorac Surg 77: 151–157, 2004.
2. Entwistle JW 3rd, Bolno PB, Holmes E, Samuels LE: Improved survival with ventricular assist device support in cardiogenic shock after myocardial infarction. Heart Surg Forum 6: 316–319, 2003.
3. Goldstein DJ, Oz MC: Mechanical support for postcardiotomy cardiogenic shock. Semin Thorac Cardiovasc Surg 12: 220–228, 2000.
4. Kitamura M, Aomi S, Hachida M, et al.: Current strategy of temporary circulatory support for severe cardiac failure after operation. Ann Thorac Surg 68: 662–665, 1999.
5. Samuels LE, Holmes EC, Thomas MP, et al.: Management of acute cardiac failure with mechanical assist: Experience with the ABIOMED BVS 5000. Ann Thorac Surg 71: S67–72 ; discussion S82–85, 2001.
6. Helman DN, Morales DL, Edwards NM, et al.: Left ventricular assist device bridge-to-transplant network improves survival after failed cardiotomy. Ann Thorac Surg 68: 1187–1194, 1999.
7. Helman DN, Oz MC: Developing a comprehensive mechanical support program. J Card Surg 16: 203–208, 2001.
8. Kherani AR, Cheema FH, Oz MC, et al.: Implantation of a left ventricular assist device and the hub-and-spoke system in treating acute cardiogenic shock: who survives? J Thorac Cardiovasc Surg 126: 1634–1635, 2003.
9. McBride LR, Lowdermilk GA, Fiore AC, et al.: Transfer of patients receiving advanced mechanical circulatory support. J Thorac Cardiovasc Surg 119: 1015–1020, 2000.
10. Chiu HS, Vogt JF, Chan LS, Rother CE: Regionalization of infant transports: The southern California experience and its implications. I: Referral pattern. J Perinatol 13: 288–296, 1993.
11. Cifuentes J, Bronstein J, Phibbs CS, et al.: Mortality in low birth weight infants according to level of neonatal care at hospital of birth. Pediatrics 109: 745–751, 2002.
12. Safford SD, Hayward TZ, Safford KM, et al.: A cost and outcomes comparison of a novel integrated pediatric air and ground transportation system. J Am Coll Surg 195: 790–795, 2002.
13. Sampalis JS, Denis R, Lavoie A, et al.: Trauma care regionalization: A process-outcome evaluation. J Trauma 46: 565–579 ; discussion 579–581, 1999.
14. Selevan JS, Fields WW, Chen W, et al.: Critical care transport: outcome evaluation after interfacility transfer and hospitalization. Ann Emerg Med 33: 33–43, 1999.
15. Oz MC, Rose EA, Levin HR: Selection criteria for placement of left ventricular assist devices. Am Heart J 129: 173–177, 1995.
16. Minev PA, El-Banayosy A, Minami K, et al.: Differential indication for mechanical circulatory support following heart transplantation. Inten Care Med 27: 1321–1327, 2001.
17. Orime Y, Shindo S, Shiono M, et al.: Experiences of postcardiotomy assist: Pneumatic ventricular assist device or venoarterial bypass with percutaneous cardiopulmonary support. Artif Organs 20: 721–723, 1996.
18. El-Banayosy A, Posival H, Minami K, et al.: Seven years of experience with the centrifugal pump in patients with cardiogenic shock. Thorac Cardiovasc Surg 43: 347–351, 1995.
19. Golding LA, Crouch RD, Stewart RW, et al.: Postcardiotomy centrifugal mechanical ventricular support. Ann Thorac Surg 54: 1059–1063 ; discussion 1063–1064, 1992.
Copyright © 2006 by the American Society for Artificial Internal Organs