| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Michael E. DeBakey Department of Surgery, Division of Congenital Heart Surgery, Baylor College of Medicine, Houston, Texas
b Division of Pediatric Cardiology, Baylor College of Medicine, Houston, Texas
c Department of Physical Medicine and Rehabilitation, Baylor College of Medicine, Houston, Texas
Accepted for publication April 20, 2010.
* Address correspondence to Dr Morales, Division of Congenital Heart Surgery, Texas Children's Hospital, 6621 Fannin St, MC-WT 19345H, Houston, TX 77030 (Email: dlmorale{at}texaschildrenshospital.org).
Presented at the Forty-sixth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 25–27, 2010.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Methods: The most recent HCUP-KID (2006) was analyzed (n = 7.5 million). This database is a nationwide sampling of hospital discharges of patients less than 20 years old weighted to provide national estimates.
Results: In 2006, 187 children had a VAD implanted in the United States. Mean age was 13 ± 7 years. Forty patients (21%) were bridged to VAD by extracorporeal membrane oxygenation. Forty-nine patients (26%) were bridged to heart transplant. Fifty-six patients (30%) died in hospital. Eighty-six patients (46%) were discharged or transferred. Length of stay was 29 days (range, 0 to 285). Total cost was $174,743 (range, $4,230 to $1,911,588). Sixty-seven hospitals placed VADs; 66% of VADs (124) were implanted at large teaching hospitals (more than 500 beds), and 46% (85) were at high-volume hospitals (more than 5 VADs per year). High-volume, large teaching hospitals (10) had better survival (89% versus 61%; p < 0.001) and lower hospital cost ($236,000 ± $184,000 versus $300,000 ± $355,000; p = 0.002) compared with all other hospitals. On multivariate analysis, acute renal failure and extracorporeal membrane oxygenation were risk factors for mortality, whereas transplant and being at a high-volume large teaching hospital were protective factors from death.
Conclusions: Preliminary data suggest that the growing use of VADs in children may be better served in regard to resource utilization and outcomes if centralized to high-volume large teaching hospitals.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Despite the rapidly expanding interest in pediatric VADs, the yearly incidence and characteristics of VAD use for children in the United States is unknown. In light of this, an examination of various pediatric national databases was undertaken to determine which could yield the most comprehensive representation of pediatric VAD use in the United States. This exercise led to the Healthcare Cost and Utilization Project (HCUP) Kids' Inpatients Database (KID). The HCUP-KID project is sponsored by the Agency for Healthcare Research and Quality (AHRQ) and is used by many federal and professional organizations to create national policy and perform research because of its inclusiveness and ability to capture national trends in pediatric hospitalizations [4] Therefore, analysis of the HCUP-KID was performed to characterize the current utilization of VADs in children across the United States.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
To create a file for this analysis, only those KID discharges with International Classification of Diseases, Clinical Modification (ICD9-CM) procedure codes for VAD implantation (3752, 376, 3760, 3762, 3765, 3766 and 3768) [6] were included, irrespective of their occurrence related to other procedures during hospitalization. Alive or dead status at discharge was known for all VAD patients. All analyses are performed using weighted values in predictive analytics software complex samples, Statistical Package for the Social Sciences, version 17 (SPSS, Chicago, IL).
Cost estimates were calculated from charges using the AHRQ HCUP cost-to-charge ratio file. The file is based on all-payer, inpatient cost, and charge information from the detailed reports by hospitals to the Centers for Medicare and Medicaid Services. Hospital-specific cost to charge ratio was used to calculate costs, and were missing in 22% of the cohort, for whom the weighted average cost-to-charge ratio was used to calculate cost [7]. Logistic regression analysis was used to determine the risk factors associated with outcomes.
The HCUP-KID 2003 database was utilized only to determine overall incidence of VAD use and age distribution.
Hospital characteristics—teaching status, bed size, and rural hospital—used in the database are defined as follows. A hospital is defined as a teaching facility if it meets any one of the following three criteria: (1) residency training approval by the Accreditation Council for Graduate Medical Education; (2) membership in the Council of Teaching Hospitals; and (3) a ratio of full-time equivalent interns and residents to beds of 0.25 or higher. Bed size categories are defined nested with location and teaching status in Table 1. A rural hospital was defined by Core-Based Statistical Area codes from 2000 census data.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
Median length of hospital stay for all children undergoing VAD therapy was 29 days (range, 0 to 285). Total cost was $174,743 (range, $4,230 to $1,911,588).
Outcome at Discharge
During hospitalization, heart transplantation was performed in 49 patients (26%), and 56 patients (30%) died, including 3 heart transplant patients. Alive without a heart transplant was the discharge status for 46% (86) of the study cohort. The VAD patients with cardiomyopathy had a better survival (85%; 63 of 74) compared with VAD patients with congenital heart disease (65%; 26 of 40) and myocarditis (67%; 16 of 24; p = 0.027). Patients requiring ECMO and VAD support had a worse survival (40%; 16 of 40) than VAD patients who did not get ECMO (79%; 116 of 147; p < 0.001). Patients who had postcongenital heart surgery ECMO support and then a VAD had a worse survival (27%; 3 of 11) than VAD patients who did not get ECMO (p < 0.001). Patients who had VAD support after congenital heart surgery during the same admission had a worse survival rate than VAD patients who did not have congenital heart surgery (Fig 2). Multivariate logistical regression analysis revealed that acute renal failure and ECMO support were highly associated with hospital mortality, whereas heart transplantation and being at a high-volume large teaching hospital were highly associated with survival (Table 2).
|
|
Comparisons between high-volume large teaching hospitals and other hospitals are summarized in Table 3. High-volume large teaching hospitals had a better survival (p < 0.001) and a 21% decrease in cost compared with all other hospitals (p = 0.002). Survival at high-volume centers (72 of 85; 85%) regardless of hospital type had an improved survival compared with small-volume centers (60 of 103; 58%; p < 0.001). Whether a VAD implantation in a child occurred at a children's hospital, a general hospital with a children's unit, or an adult hospital did not affect the survival (p = 0.3). The length of stay and cost were higher and the age of VAD-supported children was younger at children's hospitals compared with general hospitals with a children's unit or an adult hospital (p < 0.001; Table 4).
|
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
Therefore, until very recently in the United States, clinicians, the government, industry, and hospitals had no answers to address this growing cohort of children with heart failure. However, the increasing recognition of this patient population over the past 5 years has caused a significant refocusing by these entities and reallocation of resources toward pediatric VAD therapy. Clinically, a significant advancement occurred in 2004 when the Berlin Heart EXCOR became the first pediatric specific VAD in the United States to gain widespread use as a bridge to transplant, especially for small children and infants. Despite the difficult logistical and regulatory obstacles to make it available for each implant, its use in North America increased significantly from two implants in 2004 to more than 80 implants in 2009. (Information shared at the 6th International Congress on Mechanical Circulatory Support in Berlin.) This obviously growing market and strong interest in the device led to the Berlin Heart EXCOR Investigational Device Exemption prospective study, which started in May 2007 and is now nearing completion. The government's understanding that this growing cohort of children with heart failure is an unmet need has led to a series of programs starting in 2004 to fund the development of a pediatric VAD especially for small children, namely, Pediatric Mechanical Circulatory Program, and Pumps for Kids, Infants, and Neonates (PumpKIN) [10].
Despite this growing interest, the demographics and fundamental outcomes for the use of VADs in children across the United States remain unknown. There have been several informative multi-institutional investigations into VAD use in children, but they have centered on a specific clinical group such as patients listed for transplantation [11]. As this new field begins to develop in the United States, the investigators feel that a broad nationwide understanding of VAD use in children is appropriate. Therefore, such data would have to come from a general nondiagnosis, therapy, or outcome-specific database that has basic clinical, demographic, and outcome data that were representative of national inpatient pediatric healthcare trends. The HCUP-KID was chosen because of its inclusiveness, available data, and sponsorship by the Department of Health and Human Services. The use of HCUP-KID by the Food and Drug Administration, National Institute of Health, Centers for Medicare and Medicaid Services, and AHRQ to create national policies and standards for pediatric healthcare in the United States instilled confidence that results derived from this database would be a fair and accurate assessment of VAD use in children across the United States.
The use of VADs in children is clearly increasing as demonstrated by the more than 30% increase between 2003 and 2006. Most notable between the years was a significant shift in the age groups receiving VADs in 2006 toward younger children, which is surely a result of the introduction of pediatric specific VADs to the United States (ie, Berlin Heart EXCOR) between 2003 and 2006. Clearly, however, the majority of VADS being placed in children is still in older adolescents. It is important to realize that the 187 children reported herein were not patients who received pediatric VADs but rather children who received VADs, the majority of which were adult or centrifugal temporary VADs. Therefore, like the reports by Blume and colleagues [12] and Davies and associates [11], this series does not represent the use of pediatric VADs in the United States. However, as more pediatric-specific VADs are introduced in the next several years, the landscape of VAD utilization in children will change. This series captures the beginning of the field of pediatric specific VADs in the United States since it documents the change between when there was no pediatric-specific VAD use in the United States (2003) to a time (2006) when the first pediatric-specific VADs were being implanted (ie, DeBakey VAD Child and Berlin Heart EXCOR) to any significant degree.
Children with cardiomyopathy appear to have superior outcomes with VAD therapy as compared with patients with congenital heart defects, which has been demonstrated before [11]. Clearly, those patients presenting with long-standing cardiomyopathy who gradually or acutely decompensate requiring VAD support are a much different cohort than those who have a congenital heart repair that require salvage VAD support postoperatively. The latter cohort has a significantly higher mortality rate, especially if they require ECMO to bridge them to a VAD. Regardless of diagnosis, the use of ECMO to bridge patients to VAD decreases survival significantly. This was clearly demonstrated in the multivariate analysis in which ECMO was highly associated with mortality. Even when ECMO is successful at bridging a child to heart transplant, which occurs less than 50% of the time, those patients have been shown to have significantly decreased early and late survival. This is in stark contrast to patients bridged to transplant with VADs whose early and late survival is similar to the less ill cohort of transplant recipients who did not require mechanical support [11].
Extracorporeal membrane oxygenation should not be used in children with heart failure because it is for cardiopulmonary support. ECMO should only be applied in children with heart failure if they are arresting or if the decision to use circulatory mechanical support has been delayed (ie, late presentation) to the point that the lungs have been compromised. Despite significant morbidity and mortality that is well documented [13, 14], ECMO has been the mainstay for pediatric postcardiopulmonary bypass cardiac failure and end-stage heart failure causing end-organ compromise for many years [15, 16]. However, as new pediatric VADs are introduced and results from pediatric specific VADs become consistently reliable, the field of mechanical circulatory support of children will mature to the point where ECMO will be reserved for cardiopulmonary failure. Pediatric heart failure, whether postbypass or long standing, will be treated with short- or long-term VADs, as it has been treated in the adult population for more than a decade.
The most successful programs implanting VADs in children are high-volume large teaching hospitals. They have significantly better survival at a lower cost and with a trend toward shorter length of hospital stay. These findings are not surprising as it takes not only VAD expertise in a heart center but also the availability of the multiple pediatric subspecialties (ie, anticoagulation team, infectious disease, psychiatry, and so forth) to care for these complex patients. Therefore, VADs were most commonly implanted in general hospitals with a pediatric unit, which can combine the clinical infrastructure to care for these children and the exposure to a mature adult VAD program. Also these programs would tend to have pediatric heart transplant programs, which was associated with improved survival in this series as well. Free-standing children's hospitals have the clinical infrastructure to care for children with VADs but only four were high-volume programs. However, despite this, overall survival at children's hospitals compared favorably to all other hospital types. Children's hospitals did have an increase in length of stay and cost. This most likely is a result of these hospitals caring for significantly younger patients (Table 4) whose device choices, for now, are limited to VADs that are not approved for hospital discharge. However, as more pediatric specific VADs, especially for the smaller children are introduced, the expertise in children's hospitals will grow.
This type of investigation has recognized limitations secondary to its reliance on a large administrative database that has limited clinical data. Data such as cause of death and device type, which could be critical in analysis of risk factors and determining ways to improve outcomes, are not available. The KID does not include a complete episode of care for a patient if they are transferred to another hospital. Medical coding inaccuracies are a significant concern even though there are very specific codes for the other types of mechanical circulatory support that could be confused with a VAD (ie, ECMO, intraaortic balloon pump). The results presented in this manuscript as representing US experience are calculated from KID estimations. These estimations are derived through a complex formula that is verified by multiple methods, is well accepted, and is used by many federal agencies to create national standards and policy. However, this database does not have the data on every US pediatric hospitalization in 2006. Therefore, even though probably the most inclusive and accurate method to demonstrate and characterize VAD use in children across the United States, these results remain estimations.
Even though still an infrequent therapy, the field of pediatric VAD therapy in the United States has begun and is growing. As in the adult VAD field, early data suggest that patient selection (ie, cardiomyopathy patients do significantly better than postcardiotomy patients) and VAD strategy (ie, avoiding ECMO bridge to VAD) are critical to outcomes. The preliminary data from this series also suggest that the growing use of VADs in children may be better served in regard to resource utilization and outcomes if centralized to high-volume large teaching hospitals. However, as the field continues to grow, access to such few hospitals will become an issue. Therefore, identifying the factors and practices in successful centers that might be transferable to other institutions to improve their outcomes will be increasingly important.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
The first relates to types of devices utilized. As you indicated, unfortunately, in this large database, as is often the case, the specific details of the devices were not identified. In the INTERMACS (Interagency Registry for Mechanically Assisted Circulatory Support) NHLBI-sponsored study of mechanical circulatory support in the United States, an increasing focus has been the separation of factors related to patient-specific causes of death, such as severity of illness, and the actual device type utilized in an attempt to better match the device with the patient in terms of potentially improving survival. And this, of course, is particularly important in pediatric patients because of large adult devices being inadequately supportive of children and therefore an emphasis on children-specific devices such as the Berlin Heart. So would you comment as you go forward with similar analyses how you might incorporate or expect the device types to affect risk factor analyses in large subsets of patients?
The second question relates to causes of death. Again, in this particular therapy, analyses about risk factors for death are particularly important in that they can be separated into two, patient-specific causes of death and device-related causes of death. For example, patients who are in cardiogenic shock with multiorgan dysfunction are likely to have a different set of predictors of mortality than device-related mortality such as thromboembolic complications. So in going forward, would you also comment perhaps on how we might expect differences in risk factor profiles in separating out these issues between device-related and patient-specific mortality?
I also think it is very important and commend you on the emphasis on identifying those factors and expertise in the small subset of hospitals that were very successful to a larger cohort of hospitals that may be applying this therapy because of, of course, the difficulties of triaging patients to a small number of centers who are critically ill and may not be successfully transported without mortality. A very nice analysis. Thank you very much.
DR MORALES: Thank you, Dr Kirklin. In terms of types of devices, it is true that the HCUP database does not provide that type of information, which is unfortunate. I think that there is a difference between devices or VADs in children and pediatric devices. Really, until 2004, there were no pediatric devices. With the introduction of the Berlin Heart and other pediatric-specific devices that are being developed, we will see is a shift in the data to younger patients. Therefore, risk factors will become more specific to children and not just adolescents, which in the past was the patient population of children with VADs. I think the population of children under 5 years who will be getting devices will continue to grow. Our data in the past for children with VADs have really only represented adolescents, and this is now starting to change, which is why I am looking forward to the 2009 HCUP data that become available in 2011. I am sure that they will again demonstrate that the field of VAD therapy in children is growing, that an increasing number of younger children are being supported, and that pediatric-specific VADs and not adult VADs in children are becoming the norm.
In terms of causes of death, it is true that the HCUP does not allow us to assess the different causes of death. I agree that device and patient risk factors for death are very different. There is a huge difference between a patient who has multisystem organ failure and is well supported but still perishes as compared with a patient who has a device-related death such as a massive cerebral embolus. I think this will not be something that the HCUP database will be able to address, but databases such as the INTERMACS database will probably help us investigate these issues. As pediatric devices get Food and Drug Administration approved and enter into the INTERMACS database, hopefully, we can start teasing out some of these issues.
DR ANASTASIOS POLIMENAKOS (Chicago, IL): I enjoyed your talk. I have a question. How did you define a high-volume center in your study? Is it the complexity score, the high-volume ECMO centers, or high-volume total surgical cases?
DR MORALES: We did this very simplistically. If you look at all the centers putting in VADs in children, the highest number of VADs placed in children was nine, and that was accomplished at two hospitals. Seeing nine as the maximum, we chose centers placing five or more devices as the cutoff because it represented the top 20% of hospitals placing VADs in children.
|
Acknowledgments |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentDiscussionAcknowledgmentsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. S. Almond, D. L. Morales, E. H. Blackstone, M. W. Turrentine, M. Imamura, M. P. Massicotte, L. C. Jordan, E. J. Devaney, C. Ravishankar, K. R. Kanter, et al. Berlin Heart EXCOR Pediatric Ventricular Assist Device for Bridge to Heart Transplantation in US Children Circulation, April 23, 2013; 127(16): 1702 - 1711. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Karamlou, M. Vafaeezadeh, A. M. Parrish, G. A. Cohen, K. F. Welke, L. Permut, and D. M. McMullan Increased extracorporeal membrane oxygenation center case volume is associated with improved extracorporeal membrane oxygenation survival among pediatric patients J. Thorac. Cardiovasc. Surg., February 1, 2013; 145(2): 470 - 475. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. B. Silverman and S. M. Quinn Successful ICU Care of a 14-Year-Old With Total Artificial Circulation Using Two Ventricular Assist Devices for 118 Days ICU Director, September 1, 2012; 3(5): 240 - 246. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Davies and M. Priest Invited Commentary Ann. Thorac. Surg., May 1, 2012; 93(5): 1590 - 1591. [Full Text] [PDF]
|
|
|
|
|
|
J. M. Chen, M. E. Richmond, K. Charette, H. Takayama, M. Williams, L. Gilmore, A. Garcia, and L. J. Addonizio A decade of pediatric mechanical circulatory support before and after cardiac transplantation J. Thorac. Cardiovasc. Surg., February 1, 2012; 143(2): 344 - 351. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. A. Hehir, R. A. Niebler, C. C. Brabant, J. S. Tweddell, and N. S. Ghanayem Intensive Care of the Pediatric Ventricular Assist Device Patient World Journal for Pediatric and Congenital Heart Surgery, January 1, 2012; 3(1): 58 - 66. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. R. Thiagarajan, C. S. Almond, D. S. Cooper, and D. L. S. Morales Ventricular Assist Devices for Mechanical Circulatory Support in Children World Journal for Pediatric and Congenital Heart Surgery, January 1, 2012; 3(1): 104 - 109. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. N. Dean, D. G. Hillman, K. E. McHugh, and H. P. Gutgesell Inpatient Costs and Charges for Surgical Treatment of Hypoplastic Left Heart Syndrome Pediatrics, November 1, 2011; 128(5): e1181 - e1186. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
