Clinical and Economic Effectiveness of Percutaneous Ventricular Assist Devices for High-Risk Patients Undergoing Percutaneous Coronary Intervention

Abstract: Background. Comparative effectiveness research (CER) is taking a more prominent role in formalizing hospital treatment protocols and health-care coverage policies by having health-care providers consider the impact of new devices on costs and outcomes. CER balances the need for innovation with fiscal responsibility and evidence-based care. This study compared the clinical and economic impact of percutaneous ventricular assist devices (pVAD) with intraaortic balloon pumps for high-risk patients undergoing percutaneous coronary intervention (PCI). Methods. This study conducted a review of all comparative randomized control trials of the pVADS (Impella and TandemHeart) vs IABP for patients undergoing high-risk percutaneous coronary intervention (PCI). A retrospective analysis of the 2010 and 2011 Medicare MEDPAR data files was also performed to compare procedural costs and hospital length of stay (LOS). Readmission rates between the devices were also studied. Results. Based on available trials, there is no significant clinical benefit with pVAD compared to IABP. Use of pVADs is associated with increased length of Intensive Care Unit stay and a total longer LOS. The incremental budget impact for pVADs was $33,957,839 for the United States hospital system (2010-2011). Conclusions. pVADs are not associated with improved clinical outcomes, reduced hospital length of stay, or reduced readmission rates. Management of high-risk PCI and cardiogenic shock patients with IABP is more cost effective than a routine use of pVADS. Use of IABP as initial therapy in high-risk PCI and cardiogenic shock patients may result in savings of up to $2.5 billion annually of incremental costs to the hospital system. 

J INVASIVE CARDIOL 2015;27(3):148-154

Key words: intraaortic balloon pump, percutaneous left ventricular assist device, cost effectiveness, clinical outcomes

____________________________________

The total annual cost of cardiovascular disease in the United States is estimated to be $312.6 billion, with direct cardiac care consuming one-third of total expenditures.¹ As a result, Medicare and private payers are carefully evaluating patient care and the cost of this care on the total health-care budget as part of their comparative effectiveness research (CER) initiative. CER is designed to determine the clinical and cost effectiveness of similar treatments with the intent of improving outcomes while reducing cost. CER may guide payers to adjust coverage policies to ensure better value for every dollar spent while also providing doctors and patients with the information necessary to make more clinically effective, and therefore, more cost-effective health-care decisions. 

The objective of this study is to analyze the available clinical and economic effectiveness data of intraaortic balloon pump (IABP) and percutaneous ventricular assist devices (pVAD) for patients undergoing high-risk percutaneous coronary intervention (PCI). Given the cost of the devices, the care of this high-risk patient cohort may result in up to a 7-fold increase in hospital cost, or an incremental ~$66,000 per case, compared to the lowest risk patient. In addition, the hospital charges are more than 12 times the amount, or an incremental ~$498,000 per case, compared to the lowest-risk patient. The reason for the increased cost to the hospital and the payer is due to greater resource utilization because of complex comorbidities and higher procedural complications.² A major component of the costs associated with resource utilization can be linked to procedural supplies and equipment, which may include these devices. 

IABPs and pVADs are designed to provide temporary hemodynamic stabilization for patients suffering from acute left ventricular (LV) or right ventricular (RV) failure. Patients may present with heart failure due to an acute myocardial infarction (AMI) complicated by cardiogenic shock or a severely compromised LV and complex coronary artery disease. The mechanical support provided by these devices during ischemia and reperfusion can limit infarct size, sustain end-organ perfusion, and decrease LV filling pressures, thereby reducing LV pressure work and myocardial oxygen consumption.³ Use of hemodynamic support devices may decrease overall hospital costs by preventing the need for prolonged hospital stay or other complications of hemodynamic failure.

There are currently three types of percutaneous hemodynamic support devices, each with different capabilities and physical properties (Table 1). Nevertheless, the IABP is considered the gold standard and remains the most frequently used cardiac assist device due to its ease of use and clinical evidence. An IABP can provide increased cardiac output by increasing coronary perfusion during diastole and by reducing afterload. However, in certain patients, especially those with severe aortic insufficiency, an IABP may be contraindicated. Furthermore, a physician may choose a pVAD to increase cardiac output by virtue of the pVAD physically unloading the left ventricle and improving contractility. 

TandemHeart (CardiacAssist) and the Impella 2.5 (Abiomed) represent newer devices with the potential to increase the amount of circulatory support, but are more expensive and are associated with increased risks including delay to implementation, increased bleeding, hemolysis, and limb ischemia. Hospitals incur increased cost due to the price of the device (average selling price [ASP] for each of these devices: IABP $835; TandemHeart $21,131; Impella $23,645⁵), the indirect costs associated with delays in insertion, and costs associated with the treatment of complications. While there may be higher costs for pVADs, these costs may be mitigated in other ways, such as improved clinical outcomes or decreased length of hospital stay. 

Given the relative novelty of the Impella CP and the lack of any large randomized trials assessing its use, the Impella CP was not studied. The Impella 5.0 and Impella LD were not studied due to their primary use being in the surgical operating suite due to the large cannula size (21 Fr). While veno-arterial extracorporeal membrane oxygenation has been used during high-risk PCI,⁶ its role was not studied in this paper given the accompanying use of large-bore cannulae and lack of familiarity with its operation in the adult cardiac catheterization laboratory (CCL) arena.

Methods

We analyzed the available RCTs that compared pVADs with IABP in patients with cardiogenic shock and undergoing PCI to ascertain clinical effectiveness. The populations studied in these trials are described below. A meta-analysis was not performed since there were very few trials that involved patients in a randomized fashion.  

Economic analysis was performed by analyzing 2010–2011 Medicare and MEDPAR data. Patients who had received pVADs and IABP were collected and differences between the two groups were compared with univariate analysis. Furthermore, readmission data on these patients were studied and analyzed.   

Clinical analysis. Two randomized clinical trials compared TandemHeart (TH) against IABP in patients with cardiogenic shock. Despite notable improvements in hemodynamic parameters in patients treated with TH, 30-day survival was similar between both groups in each study, while patients receiving support with TH experienced more complications including bleeding and limb ischemia (Table 2).^7,8 Currently, there are no large randomized trials comparing TH with IABP in a high-risk PCI patient population.

A single randomized trial compared the Impella 2.5 against IABP in patients with cardiogenic shock (Table 3).⁹ In this study, the use of Impella was neither associated with improvements in hemodynamic parameters at 2 hours nor with increased 30-day survival. However, patients treated with the Impella 2.5 experienced a significantly higher incidence of hemolysis.  

The PROTECT II study was a prospective randomized clinical trial of hemodynamic support with the Impella 2.5 versus IABP in patients undergoing high-risk PCI. The primary endpoint was the rate of 30-day major adverse events (MAE) including: all-cause death, Q-wave or non-Q wave myocardial infarction, stroke or TIA, repeat revascularization (PCI or CABG), need for a cardiac or vascular operation, acute renal insufficiency, severe intraprocedural hypotension requiring therapy, cardiopulmonary resuscitation or ventricular tachycardia requiring cardioversion, aortic insufficiency and angiographic failure of PCI. A 90-day follow-up was required by protocol. The data safety monitoring board (DSMB) stopped the trial for futility based on inability to demonstrate differences in the primary endpoint after 452 of the 654 scheduled patients were enrolled. At the time of curtailment, the composite MAE was statistically equivalent at 30 and 90 days for the intent-to-treat population as well as the 30-day per-protocol population (Tables 4 and 5).¹⁰ However, the composite MAE of Impella-treated patients at 90 days was reduced compared to IABP patients in the per-protocol population without a reduction in any individual component of the MAE rate. There was a non-significant trend suggesting that Impella may be associated with a reduction in the need for repeat revascularization. Despite being a randomized trial, differences observed may have been due to increased operator use of rotational atherectomy (RA) in the Impella group (median passes per lesion 1 [IABP] vs 3 [Impella], P<.01; median passes per patient 2 [IABP] vs 5 [Impella], P<.01; median time per lesion 40 seconds [IABP] vs 60 seconds [Impella], P<.01; RA of left main artery 3.1% [IABP] vs 8.0% [Impella], P=.02).¹¹ Increasing use of RA did not result in a reduction of MAE between the groups.

Economic analysis. Economic analysis was performed by utilizing the 2010 and 2011 Medicare MEDPAR data files. The datasets were identified by using a series of diagnostic- and procedure-related ICD-9 codes. ICD-9 code surgical procedure code 00.66 was used to identify the PCI procedure and surgical procedure codes 37.61 and 37.68 were used to identify the IABP or TH/Impella, respectively. TH and Impella share the same procedure code; however, based on the 2012 IMS database, Impella represents 92% of the total between the two devices. We also examined if a co-occurrence existed with or without diagnosis code 785.51 to determine if the patient was in cardiogenic shock (Table 6). Lastly, we selected the procedures that were routed to MS-DRG 237 or 216 to capture the patients who have been identified as having major complications and comorbidities. A hospital cost and charge analysis was completed for this select group of patients, with an analysis on hospital reimbursement to determine the total budget impact of both devices. 

Based on these code groupings, we identified a total of 10,261 IABP patients and 1119 TH or Impella patients who met the criteria over the 2-year period. These patients were then divided into those with and without cardiogenic shock and both the cost and charge per procedure were compared (Table 7). There were significant differences in the total cost and charge per procedure that were primarily driven by an increase in supplies and equipment costs associated with the pVADs. We also observed notable differences in mean intensive and non-intensive care unit days for both groups. IABP patients averaged fewer days in an intensive care unit (7.11 vs 8.74 days; P<.001) and their total hospital stay was approximately 2 days shorter (8.50 days vs 10.70 days; P<.001). In addition to the higher costs observed in the TH and Impella group, there was an increase in associated costs including increased use of blood products, associated laboratory costs, and higher use of the operating room because of the longer length of stay.  

The total budget impact associated with pVADs can be quite substantial when viewed in the context of the entire United States health-care system. The incremental budget impact, based on a review of the available data, is approximately $34,000,000 to the hospital system and up to $109,000,000 to the payer system (Table 8).  

The impact will vary depending on payer-allowable charges or reimbursement from the payer to the hospital, which is affected by a number of different factors. The complexity of payment determination is not discussed in this article, but a previously published article did share allowable payments per claim from a national commercial health plan. In their analysis of high-risk PCI patients, the allowed payment per claim was $85,048 and $94,670 for the IABP and pVAD group, respectively.¹² Based on these figures, the hospital receives an additional payment of $9622 and the payer incurs an additional cost in the same amount. The additional payment to the hospital does not offset the incremental cost per procedure resulting from the use of a pVAD. 

A sensitivity analysis was also performed to better understand the total impact on the entire health-care system of utilizing the pVADS.  There are approximately 7000 pVADs and 75,000 IABP catheters used per year in the United States.¹³ Table 9 illustrates the total incremental hospital cost impact based on a progressive percent in migration from current IABP users to TH or Impella. The percent migration starts at 10%, which represents approximately the amount of migration that has already occurred and is projected to 100%.

Readmission analysis. One of the reported advantages of the pVADs is their potential reduction in readmission rates. Readmission rates are critically important to health-care providers and hospitals since an increase in readmission rates may result in reduced reimbursement to health-care providers. We accessed the IMS Health Database to review all PCI-related procedures (ICD-9 code 00.66) that utilized an IABP (ICD-9 code 37.61) or TH or Impella (ICD-9 code 37.68). IMS Health also had the ability to distinguish between TH and Impella when the 37.68 code was used so each device could be reviewed independently. We then followed those patients from January 2012 through March 2013, to determine if any cardiac-related readmissions occurred at 30, 60, or 90 days after the index procedure using all existing cardiac-related ICD-9 diagnosis codes. A total of 55,703 patients were identified and included in this analysis (IABP = 52,364 and TH or Impella = 3339). Of the TH or Impella group, 175 TH and 2559 Impella devices were identified. The groups were compared on the basis of device and if the patient underwent PCI with or without the presence of cardiogenic shock (ICD-9 code 785.51). The results yielded no statistical difference in cardiac readmissions between patients who were treated with IABP or the combined TH or Impella groups. There was also no statistical difference in cardiac readmissions between the IABP or Impella groups. Given the small number of TH devices identified we could not make any statistical comparisons, but the overall readmission rate for TH was comparable to IABP and Impella.

Discussion

This study investigated the clinical and cost effectiveness of pVADs compared with the use of IABP therapy in cardiogenic shock and high-risk PCI patients. A review of published data demonstrates no conclusive clinical benefit in trial patients treated with pVADs compared with IABP. 

Different algorithms have been suggested to assist physicians in deciding when to place a TH or Impella. Most physicians agree that an escalating approach is reasonable. In this approach, patients should initially receive vasopressor agents and an IABP. If a patient fails to stabilize, a TH or Impella can be considered.¹⁴ Use of pVADs has a theoretical benefit for improved left ventricular unloading and improved cardiac output. However, their larger cannulae sizes are associated with a higher rate of vascular complications, bleeding, and hemolysis. It is important that hospitals use this information to establish treatment protocols and payers adjust their coverage policies to reinforce clinician behavior to practice evidence-based medicine. Given clinical evidence and the cost-effectiveness data, use of pVADs as a first-line therapy may not be justified.

The Food and Drug Association (FDA) panel recently decided to reclassify IABP as a class II device because of its proven safety and efficacy profile, while maintaining the TH and Impella 2.5 as class III devices, considering them investigational and requiring a pre-market approval (PMA) study. Furthermore, the American College of Cardiology/American Heart Association guidelines recommend keeping these alternative left ventricular assist devices available for patients in refractory cardiogenic shock.¹⁵ 

Gregory et al recently reported that initial inpatient charges are lower with IABP,¹² and provide a sensitivity analysis that quantifies a 10% migration costing the payer an additional $603,415 based on a health plan that consists of approximately 25 million members and an additional $3,017,075 based on a 50% migration. These amounts reflect actual payer-allowed charges. 

The same authors published a value-based analysis of hemodynamic support strategies that utilized the PROTECT II study to establish an incremental cost-effectiveness ratio (ICER) for Impella.¹⁶ This model used the composite MAE to make projections; however, given the lack of difference in the individual MAE rates between the two groups, it is difficult to draw conclusions from this analysis. Despite these limitations, the study did confirm a significant reduction in the procedure cost and charge associated with cases that used IABP therapy instead of Impella. While no prospective randomized trial currently exists specifically addressing cost effectiveness, the PROTECT II study is the largest randomized trial to date that studied pVAD versus IABP and the secondary cost analysis of this trial suggest increased cost for the use of a pVAD in a relatively controlled patient population.

Use of pVADs in the high-risk PCI and cardiogenic shock setting does not confer additional clinical benefit over the use of IABP despite the additional costs of the device. Furthermore, use of pVADs does not appear to reduce hospital readmission in this high-risk patient cohort. The use of IABP for cardiogenic shock and high-risk PCI patients should be implemented as a first-line strategy with an escalating approach toward TH or Impella for refractory patients. This methodology can avoid up to $2.5 billion annually of incremental cost to the health-care system.  

Study limitations. This study focused only on the TH and Impella 2.5, and not the Impella CP, the Impella 5.0, or the Impella LD. Given that the Impella 5.0 and LD are primarily utilized in the operating room and not in the cardiac catheterization laboratory for the treatment of high-risk PCI patients, inclusion of these devices in this study is unlikely to change the results. The Impella CP is a relatively new device that has not been extensively studied, so therefore it was not included in this study.

This study relied on data from Medicare and did not analyze data from private payers. It is unknown whether the analysis of data from private payers would impact the findings of this study since the majority of the patients who receive this therapy are covered under Medicare. 

Furthermore, a univariate analysis was performed between the two groups with regard to the differences in the pVAD and IABP groups. Further details on the clinical characteristics of the two groups were not performed given that patients were identified based on their diagnosis. Physicians may choose a pVAD over an IABP because of a certain clinical situation that may not be necessarily reported in the data reported to Medicare. However, given the diagnosis, and associated modifiers, clinical differences between the two groups may not be that different, but a randomized, controlled trial would be the only way to fully control for this possible confounder.

References

1. Go AS, Mozaffarian D, Roger VL, et al. Heart disease and stroke statistics 2013. Circulation. 2013;127(1):e6-e245. Epub 2012 Dec 12.
2. Data source: 2010 MedPAR data released with FY2012 IPPS final rule.
3. Sarkar K, Kini AP. Percutaneous left ventricular support devices. Cardiol Clin. 2010;28(1):169-184.
4. Basra SS, Loyalka P, Kar B. Current status of percutaneous ventricular assist devices. Current Opin Cardiol. 2011;26(6):548-554. 
5. Data source: 2012 IMS health database.
6. Teirstein PS, Vogel RA, Dorros G, et al. Prophylactic versus standby cardiopulmonary support for high-risk percutaneous transluminal coronary angioplasty. J Am Coll Cardiol. 1993;21(3):590-596.
7. Thiele H, Sick P, Boudriot E, et al. Randomized comparison of IABP support with percutaneous left ventricular assist device in patients with revascularized AMI complicated by cardiogenic shock. Eur Heart J. 2005;26(13):1276-1283.
8. Burkhoff D, Cohen H, Brunckhorst C, O’Neill WW. A randomized multicenter clinical study to evaluate the safety and efficacy of the TandemHeart percutaneous assist device versus conventional therapy with IABP for treatment of cardiogenic shock. Am Heart J. 2006;152(3):469.e1-469.e8.
9. Seyfarth M, Sibbing D, Bauer I, et al. A randomized clinical trial to evaluate the safety and efficacy of a percutaneous left ventricular assist device vs. IABP for treatment of cardiogenic shock caused by myocardial infarction. J Am Coll Cardiol. 2008;52(19):1584-1588.
10. O’Neil WW, Kleiman N, Henriques J, et al, A prospective randomized clinical trial of hemodynamic support with Impella 2.5 versus IABP in patients undergoing high-risk PCI: the PROTECT II study. Circulation. 2012;126(14):1717-27.
11. United States of America Department of Health and Human Services, Food and Drug Administration: Center for Devices and Radiological Health Medical Devices Advisory Committee, Circulatory System Devices Panel; December 5, 2012.
12. Gregory D, Scotti DJ, de Lissovoy G, et al. A value based analysis of hemodynamic support strategies for high-risk heart failure patients undergoing a percutaneous coronary intervention. 2013;6(2):88-99.
13. Thiele H, Smalling RW, Schuler GC. Percutaneous left ventricular assist devices in acute myocardial infarction complicated by cardiogenic shock. Eur Heart J. 2007;28(17):2057-2063. Epub 2007 Jun 22.
14. O’Gara P, Kushner FG, Ascheim DD, et al. ACCF/AHA guideline for the management of ST-elevation myocardial infarction. Circulation. 2013;127(4):e362-e425. Epub 2012 Dec 17.
15. Gregory D, Scotti DJ. A budget impact model to estimate the cost dynamics of treating high-risk heart failure patients with advanced percutaneous cardiac assist devices: the payer perspective. J Managed Care Med. 2013;16(1):61-69.
16. Gregory D, Scott DJ, de Lissovoy G, et al. A value-based analysis of hemodynamic support strategies for high risk heart failure patients undergoing a percutaneous coronary intervention. Am Health Drug Benefits. 2013:6(2):88-99.

____________________________________

From the ¹Section of Cardiology, Department of Medicine, The University of Chicago, Chicago Illinois; ²Department of Radiology, The University of Chicago, Chicago Illinois; and the ³Section of Cardiology, Department of Medicine, Harbor-UCLA Medical Center, Torrance, California.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Shah reports a grant from Maquet Cardiovascular. The remaining authors report no conflicts of interest regarding the content herein.

Manuscript submitted May 19, 2014, provisional acceptance given August 21, 2014, final version accepted September 11, 2014.

Address for correspondence: Atman P. Shah, MD, The University of Chicago, 5841 South Maryland, MC 6080, Chicago, IL 60637. Email: ashah5@uchicago.edu