Author + information
- Published online September 25, 2017.
- aDepartment of Cardiovascular Medicine, Cleveland Clinic, Cleveland, Ohio
- bHeart and Vascular Institute, Kaufman Center for Heart Failure, Cleveland Clinic, Cleveland, Ohio
- ↵∗Address for correspondence:
Dr. Randall C. Starling, Department of Cardiovascular Medicine, Cleveland Clinic, 9500 Euclid Avenue, Desk J3-4, Cleveland, Ohio 44195.
“Happiness is a place between too little and too much.”
—Finnish Proverb (1)
Cowger et al. (2), in this issue of JACC: Heart Failure, report the relationship between 1-year survival and left ventricular assist device (LVAD) center volume based upon an Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) registry database study. The medical literature is replete with reports indicating that the quality of outcomes is related to center volume and surgeon experience (3–5). The authors sought to determine if low-volume centers are inferior to higher-volume centers. Surprisingly, they found that “very low” (≤10 cases) LVAD centers and “high volume” (>50 cases) implantations/year had inferior outcomes, with an adjusted mortality rate of 32% higher in very-low volume centers and 17% higher in high volume centers compared with medium volume centers. The “sweet spot” for LVAD outcomes was medium volume centers performing 31 to 50 implantations per year. The authors noted that high volume centers might have a referral bias (more ill patients); patients deemed “too risky” who were turned down from other centers; more deaths were attributed to multi-organ failure; and patients had worse renal function. These factors likely contributed to the increased mortality in high volume centers. The authors concluded that the Centers for Medicare and Medicaid Services stipulation for reimbursement might be “too lenient” at ≥10 LVAD implantations over a 3-year period.
Although we acknowledge that this study was expertly performed, we urge caution when interpreting the results and the authors’ stated conclusions. All registries are fallible and weakened by missing data and lack of important variables that were not accounted for when designed. There are 2 concerns: 1) the conclusion that the Centers for Medicare and Medicaid Services limit for LVAD implantations per year should be revised; and 2) the finding that survival is lower in high volume centers. Access to care is important, and limiting centers purely based on a volume threshold would be unjust. A low volume center with experienced personnel and careful judgment might achieve excellent results. We believe a volume threshold should serve as a guide, but that demonstrated excellent outcomes should trump a volume threshold. This is a good value proposition both for patients and payers. Conversely, low volume and poor outcomes are not sustainable, and this is the model followed by the United Network for Organ Sharing when assessing heart transplantation programs.
Why did high volume centers have inferior survival rates compared to medium volume centers? This is a statistical finding based on registry data. The authors did not include the distribution of survival among high volume centers; certainly, there is variability, and other factors may have influenced outcome that were not accounted for in the INTERMACS registry. We provide in Table 1 putative variables that influence risk and outcomes that were not accounted for in the INTERMACS registry. High volume centers frequently are asked to accept patients from other centers that are considered “too high risk” for the center requesting referral. As indicated by the authors, surgeon experience is important and was not accounted for in their analysis (6). The importance of the experience level of the cardiologist and transplantation coordinator for heart transplantation outcomes and was first shown 25 years ago (3). Logically, the experience of both the cardiologist and the LVAD coordinator is important for LVAD outcomes. We believe staffing ratios are critically important for LVAD outcomes. Typically, smaller-volume programs have favorable patient/caregiver ratios. Conversely, high volume programs are subject to lower patient/caregivers ratios. Staff turnover and changing manuals of operation and technology mandate ongoing in-service education and competency testing to assure high-quality care and outcomes. Dire LVAD complications including device thrombosis, infected LVADs, refractory gastrointestinal bleeding, and severe aortic insufficiency often require urgent device exchange, complex procedures (transcatheter aortic valve replacement), or cardiac transplantation. The LVAD center must be prepared to provide care for LVAD patients with serious complications; geographic variability in donor waiting times can clearly influence strategy for care (LVAD exchange versus urgent transplantation) and outcomes. Many LVAD patients with serious complications necessitating LVAD exchange hence are referred to high-volume centers with extensive experience.
Somewhat unexpected was the hazard ratio for centrifugal flow versus axial flow LVAD (hazard ratio: 0.82; p = 0.012). The ENDURANCE (A Clinical Trial to Evaluate The HeartWare Ventricular Assist System) trial reported by Rogers et al. (7) showed equivalent survival with centrifugal flow versus axial flow LVAD over 24 months (p = 0.17). Differences reflected in this INTERMACS cohort may reflect noninvestigational sites and a “learning curve” for centers with a new device.
The best outcomes in this INTERMACS registry analysis were in medium volume LVAD centers performing 31 to 50 implantations per year. This is not surprising in the context of the infrastructure of a typical advanced heart failure center. Most centers this size offer both heart transplantation and LVAD therapy and have multiple heart failure cardiac surgeons and cardiologists. Patients likely will receive the best therapy option versus what the only option the center has to offer. LVAD patients are labor and resource intensive and require frequent hospitalizations for adverse events (8). The medium volume center is likely to be well staffed but not over stressed due to an overwhelming patient volume. The multiplier effect of an expanding LVAD center should not be underestimated as far as the demands on the caregivers and infrastructure (Figure 1). This becomes exaggerated with a more complex, higher-risk patient population with longer intensive care unit (ICU) stay, longer hospitalization post-implantation, and more frequent re-hospitalizations for adverse events.
In summary, we do not believe this study proves that low volume centers have worse outcomes uniformly. We believe the quality of a center should be determined based on outcomes and not specific volume thresholds. Of course, maintaining competency for a complex procedure requires familiarity only garnered by experience and process improvement gained by a continuous volume of cases. We suggest caution when interpreting the finding that high-volume centers have worse survival compared to medium volume centers; acknowledging numerous unaccounted for variables. However, increased risk clearly occurs at both ends of the spectrum and outcomes may be influenced by very different variables in high versus very-low volume LVAD centers.
↵∗ Editorials published in JACC: Heart Failure reflect the views of the authors and do not necessarily represent the views of JACC: Heart Failure or the American College of Cardiology.
Dr. Starling has received research funding from Abbott and Medtronic; is on an advisory panel for Medtronic; and is a member of the INTERMACS Finance Committee. Dr. Xanthopoulos has reported that he has no relationships relevant to the contents of this paper to disclose.
- 2017 American College of Cardiology Foundation
- ↵Tall Clover Farm. A Place Between Too Little and Too Much. Available at: http://tallcloverfarm.com/5486/a-place-between-too-little-and-too-much. Accessed July 5, 2017.
- Cowger J.A.,
- Stulak J.M.,
- Shah P.,
- et al.
- Laffel G.L.,
- Barnett A.I.,
- Finkelstein S.,
- Kaye M.P.
- Pettit S.J.,
- Jhund P.S.,
- Hawkins N.M.,
- et al.
- Rogers J.G.,
- Pagani F.D.,
- Tatooles A.J.,
- et al.
- Smedira N.G.,
- Hoercher K.J.,
- Lima B.,
- et al.