- Email: [email protected]
Transcatheter Versus Surgical Pulmonary Valve Replacement in Repaired Tetralogy of Fallot
Accepted Manuscript
Trans-catheter Versus Surgical Pulmonary Valve Replacement in Repaired Tetralogy of Fallot Joshua A. Daily MDMEd , Xinyu Tang PhD , Michael Angtuaco MD , Elijah Bolin MD , Sean M. Lang MD , R. Thomas Collins II MD PII: DOI: Reference:
S0002-9149(18)31024-5 10.1016/j.amjcard.2018.04.028 AJC 23269
To appear in:
The American Journal of Cardiology
Received date: Revised date: Accepted date:
16 January 2018 4 April 2018 6 April 2018
Please cite this article as: Joshua A. Daily MDMEd , Xinyu Tang PhD , Michael Angtuaco MD , Elijah Bolin MD , Sean M. Lang MD , R. Thomas Collins II MD , Trans-catheter Versus Surgical Pulmonary Valve Replacement in Repaired Tetralogy of Fallot, The American Journal of Cardiology (2018), doi: 10.1016/j.amjcard.2018.04.028
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Trans-catheter Versus Surgical Pulmonary Valve Replacement in Repaired Tetralogy of Fallot
MDa, b, Sean M. Langa, b, MD, R. Thomas Collins, II, MDc, d
Affiliations: aArkansas Children’s Hospital, Little Rock, AR b
Stanford University School of Medicine, Palo Alto, CA
d
AN US
c
University of Arkansas for Medical Sciences, Little Rock, AR
CR IP T
Joshua A. Dailya, b, MD, MEd, Xinyu Tang, PhDa, b, Michael Angtuacoa, b, MD, Elijah Bolin,
Lucile Packard Children’s Hospital Stanford, Palo Alto, CA
Conflict of Interest Statement: The authors have no financial relationships relevant to this article
Corresponding Author:
ED
M
to disclose.
PT
Joshua A. Daily, MD, MEd
Arkansas Children’s Hospital
CE
1 Children’s Way, Slot 512-3 Little Rock, AR 72202
AC
Phone: 501-364-2012
Email: [email protected]
Running Head: Trans-catheter Versus Surgical Pulmonary Valve Replacement
1
ACCEPTED MANUSCRIPT
Abstract Trans-catheter pulmonary valve replacement (TC-PVR) is an alternative to surgical PVR (SPVR) in repaired Tetralogy of Fallot (TOF). The purpose of this study is to compare in-hospital outcomes, hospital costs, and projected 5-year total costs of S-PVR to TC-PVR in patients with
CR IP T
repaired TOF. We performed a multi-center, retrospective cohort study of children and adults
with TOF ≥ 8 years of age who underwent PVR from January 1, 2010 to December 31, 2016 at 46 centers contributing to the Pediatric Health Information Systems database. Baseline
characteristics, in-hospital outcomes, and costs were compared between the two groups. Projected
AN US
5-year costs were calculated by combining cost data with published re-intervention rates. A total of 194 TC-PVR and 1072 S-PVR were performed. The baseline characteristics of the TC-PVR and S-PVR groups were not significantly different with the exception of greater age in the TCPVR group (median age of 17 years vs 15 years, p value <0.001). Discharge mortality, hospital
M
charges and estimated cost, surgical complication rates, and acute kidney failure were not significantly different between the groups. Intensive care unit (ICU) use, ICU length of stay
ED
(LOS), mechanical ventilation use, extracorporeal membrane oxygenation use, and total LOS were lower with TC-PVR than S-PVR. Projected 5-year costs were greater with TC-PVR
PT
compared to S-PVR ($64,762 vs $56,536) due to the cost of the trans-catheter pulmonary valve
CE
and higher re-intervention rates. In conclusion, in spite of longer LOS and greater in-hospital resource utilization for patients with TOF undergoing S-PVR compared to TC-PVR, mortality
AC
and in-hospital costs are the same, and projected 5-year costs are less. Key Words: Trans-catheter Pulmonary Valve Replacement, Tetralogy of Fallot
2
ACCEPTED MANUSCRIPT
Initial repair of Tetralogy of Fallot (TOF) often results in severe pulmonary valve insufficiency that leads to a chronic volume load on the right ventricle and the need for multiple pulmonary valve replacements (PVR) throughout a patient’s lifetime. Surgical PVR (S-PVR) was long the standard of care for severe pulmonary insufficiency in repaired TOF 1,2. Recently, trans-catheter
CR IP T
PVR (TC-PVR) has been developed as an alternative to S-PVR which theoretically avoids the morbidity, mortality, and costs associated with open-heart surgery. Previous attempts to compare the cost of S-PVR and TC-PVR have been made with conflicting results 3-7. These comparisons utilized heterogeneous cohorts with multiple types of congenital heart disease, small patient
AN US
numbers, single centers, and/or charge data instead of cost data. In addition, attempts at estimating the 5-year total costs of both approaches have been based on theoretical reintervention rates. Recently, the 7-year results from the United States Melody Valve
Investigational Device Exemption study were published, allowing for a more accurate estimation
M
of 5-year cost of TC-PVR 8. The purpose of this current study was to compare in-hospital outcomes, hospital costs, and projected 5-year costs for patients with TOF who undergo either
ED
TC-PVR or S-PVR.
PT
Methods
With the approval of the Institutional Review Board of the University of Arkansas for
CE
Medical Sciences, data were obtained from the Pediatric Health Information Systems (PHIS), an administrative database of 46 not-for-profit, tertiary care pediatric hospitals in the United States
AC
affiliated with the Children’s Health Corporation of America. The PHIS data include detailed, deidentified information on each inpatient’s demographics, diagnoses, procedures, medications, outcomes, and cost. Data quality assurance is ongoing and data from the individual hospitals are accepted when classified errors for a given quarter occur less frequently than a criterion threshold of 2%.
3
ACCEPTED MANUSCRIPT
The study design was a multi-center, retrospective cohort investigation of all children and adults ≥ 8 years of age with a diagnosis of TOF (International Classification of Diseases, Ninth Revision [ICD-9], code 745.2 or International Classification of Diseases, Tenth Revision [ICD10], code Q21.3) who underwent either S-PVR or TC-PVR at any of the PHIS hospitals from
CR IP T
January 1, 2010 to December 31, 2016. Since hospital reporting transitioned from ICD-9 to ICD10 on October 1, 2015, ICD-9 codes were used prior to this date, and ICD-10 codes were used
after. Patients were assigned to the S-PVR group if the following codes were used: 35.25, 35.92, 33475, 02RH08Z, 02RH0JZ, 02RH0KZ, 02RH07Z 021K0KP, 021K09Q, 021K08P, 021K08Q,
AN US
021K08R, 021K09P, 021K09R, 021K0AP, 021K0AQ, 021K0AR, 021K0JP, 021K0JQ,
021K0JR, 021K0KQ, 021K0KR, 021K0ZP, 021K0ZQ, or 021K0ZR. Patients were assigned to the TC-PVR group if the following codes were used: 35.07, 35.08, 35.09, 33477, 0262T, 02RH48Z, 02RH38Z, 02RH3JZ, 02RH4KZ, 02RH4JZ, 02RH37Z, 02RH3KZ, 02RH37H,
M
02RH38H, 02RH3JH, 02RH3KH, 02RH47Z, or 02RH4JZ. After review of multiple instructions for billing of TC-PVR, it was noted that some included the use of ICD-9 code 35.26 which was
ED
also commonly used for S-PVR 9,10. Patients with an ICD-9 code of 35.26 were reviewed in detail by the investigators, and the following schema was developed to place them in the appropriate
PT
groups. Patients with ICD-9 code 35.26 were placed in the S-PVR group if the Operative Room Charge Flag was present and procedure code 3961 (Extracorporeal circulation auxiliary to open
CE
heart surgery) was utilized. Conversely, patients were placed in the TC-PVR group if the Operative Room Charge Flag was not present and procedure code 3723 (combined right and left
AC
heart catheterization) was used. Data were extracted from the PHIS database by direct query and included age, gender,
race, insurance payor (private, Medicaid, and other), median household income by patient’s zip code, and presence of a genetic syndrome (758.1, 758.2, 758.0, 758.31, 758.32, 758.33, 758.9, 758.6, 758.89, 279.11, Q91.7, Q91.3, Q90.9, Q93.4, Q93.81, Q93.88, Q99.9, Q96.9, Q99.8, or
4
ACCEPTED MANUSCRIPT
D82.1). Discharge mortality, intensive care unit (ICU) use, hospital length of stay (LOS), ICU LOS, mechanical ventilation, mechanical ventilation days, extracorporeal membrane oxygenation (ECMO), acute kidney failure (586, 584.9, N17.0, N17.1, N17.2, N17.8, or N17.9), and surgical complications were also extracted. The codes that triggered the surgical complication flag were
CR IP T
additionally evaluated and divided into the following categories: cardiac, respiratory, hemorrhage/ hematoma, infection/ fever, and other. The use of mechanical ventilation referred
only to use during the ICU stay (mechanical ventilation in the operating room or catheterization laboratory was not considered). A day was counted as one mechanical ventilation day if a patient
AN US
received mechanical ventilation in the ICU at any point during that day. Estimated total costs were also extracted. The PHIS database calculates estimated total cost by converting charges using ratios of cost-to-charges and adjusting for regional differences (based on wage/price indices). All costs were adjusted for inflation to 2016 US (United States) dollars using the
M
Consumer Pricing Indexes Data.
Descriptive statistics were expressed as median with interquartile ranges (IQRs: 1st
ED
quartile (Q1)-3rd quartile (Q3)) for continuous variables and percentage (count) for categorical variables. The distributions of continuous variables were compared between unmatched groups
PT
using the Wilcoxon rank-sum test. The proportions of categorical variables were compared
CE
between unmatched groups using the chi-square test. Boxplot was drawn without outliers for estimated cost in 2016 US dollars over time among TC-PVR and S-PVR respectively. Outliers
AC
were defined as values less than (Q1 - 1.5×IQR) or greater than (Q3 + 1.5×IQR). Propensity score matching was performed between the TC-PVR and S-PVR groups. A
multivariable logistic regression model was fitted to estimate the probability of TC-PVR as a function of discharge year, admission age, insurance payor, and genetic diagnosis. Using the linear propensity score calculated based on the model, TC-PVR patients were matched to S-PVR patients at a 1:2 ratio. The 1:2 ratio was determined by evaluating the crude sample size ratio 5
ACCEPTED MANUSCRIPT
between TC-PVR and S-PVR by discharge year. Generalized linear mixed model was used for comparing baseline characteristics and outcomes between matched TC-PVR and S-PVR. Appropriate distribution and link functions were used for different variable types. For example, Gaussian distribution and identity link was used for continuous variables, and binomial
CR IP T
distribution and logit link was used for binary variables. A random effect was included in the models to account for clustering after matching. P-values ≤ 0.05 were used to indicate statistical significance.
Projected 5-year total costs were estimated using re-intervention data for both the S-PVR
AN US
and TC-PVR. To determine the S-PVR re-operation rate, the results of two recent meta-analyses of S-PVR after operative TOF were averaged for a 5-year repeat PVR rate of 7.2% 1,2. For the purposes of comparison, it was assumed that all patients in the S-PVR group who required a repeat PVR underwent S-PVR. Balloon angioplasty or valvuloplasty rates were assumed to be 3%
M
at 5 years based on a previous study 11. Trans-catheter pulmonary valve failure rates were based on the five-year results of the United States Melody Valve Investigational Device Exemption trial . Of the 150 patients who underwent initial successful TC-PVR in the Investigational Device
ED
8
Exemption trial, at a median follow-up of 4.5 years, 12 (8%) had undergone surgery, 6 (4%) had
PT
undergone catheter dilation, and 20 (13.3%) underwent a second TC-PVR. These rates of re-
CE
intervention were used to estimate the 5-year costs for TC-PVR. We assumed that the cost for a second procedure was the same as the first. To estimate the cost of catheter dilation, the costs of
AC
the Melody valve®, Ensemble delivery system®, and one night hospitalization were subtracted from the estimated total costs for a hospitalization for TC-PVR. Wage loss was calculated as the product of estimated days of work lost and median household income based on the zip code of the patient. Total days of work lost were determined by adding the patient’s LOS to the recommended time away from work, which was assumed to be 1 week for TC-PVR and 3 weeks for S-PVR. For patients < 18-years-old, it was assumed that the patient’s parent was out of work
6
ACCEPTED MANUSCRIPT
for the recovery time. To convert median household income to median income of a single earner, the ratio of 1.28 earners per household was used based on 2010 census data. One- and two-way sensitivity analysis was performed. For one-way sensitivity analysis, the total PVR rate (including both surgical redo and trans-catheter re-implantation) after TC-PVR
CR IP T
was allowed to vary. Based on the results of the IDE study, PVRs after TC-PVR were assumed to be 38% (8/21) surgical and 62% (13/21) trans-catheter. The cost of PVR after TC-PVR was
assumed to be the same as the cost of an initial TC-PVR or S-PVR. For purposes of comparison, the surgical redo rate after S-PVR was assumed to be of 7.2% based on two prior meta-
AN US
analyses1,2. For two-way sensitivity analysis, both the surgical redo rate and the trans-catheter reimplantation rate after TC-PVR were allowed to vary to evaluate the effect that changing reintervention rates would have on the preference of S-PVR verses TC-PVR.
M
Results
From 2010 to 2016, a total of 1256 patients with TOF underwent PVR at participating
ED
PHIS centers. Five of these patients (0.4%) were excluded because they could not be assigned to either the TC-PVR or S-PVR group. Of the remaining 1251 patients, 1236 patients contributed
PT
one admission, and 15 patients contributed two admissions each. Of the 15 patients with two
CE
admissions, 11 patients underwent S-PVR twice, three patients underwent S-PVR followed by TC-PVR, and one patient underwent TC-PVR followed by S-PVR. In the entire cohort, there
AC
were 194 admissions with TC-PVR and 1072 admissions with S-PVR. The baseline characteristics of both groups of patients, with and without propensity score matching, can be found in Table 1. A comparison of in-hospital outcomes between TC-PVR and S-PVR can be found in Table 2. Discharge mortality, adjusted billed charges, estimated cost, acute kidney failure, and surgical complication (any) rate were not significantly different between the two groups. Cardiac 7
ACCEPTED MANUSCRIPT
surgical complications were more common with TC-PVR, and respiratory, hemorrhage/hematoma, infection/fever, and other complications were more common with SPVR. Admission to the ICU, ICU LOS, mechanical ventilation use, ECMO use, and total hospital LOS were lower with TC-PVR than S-PVR.
CR IP T
The median estimated cost for admission for TC-PVR and S-PVR are found in Figure A. When adjusted for inflation to 2016 US dollars, estimated cost did not change significantly for either group over the study period. Projected five-year costs can be found in Table 3. The
projected 5-year costs were greater for TC-PVR than S-PVR both from the perspective of the
AN US
payor (wage loss not considered) and from the perspective of society (wage loss included). These differences were primarily due to the higher re-intervention rates for TC-PVR compared to SPVR used in the model. The results of one- and two-way sensitivity analyses are found in Figures B and C. In both, S-PVR is more financially favorable than TC-PVR over a range of possible re-
M
intervention rates.
ED
Discussion
There were three important findings in this multi-center study investigating the largest
PT
cohort to date of patients undergoing TC-PVR versus S-PVR after initial palliation of TOF. First, LOS and hospital resource utilization were lower with TC-PVR than S-PVR. Second, mortality
CE
and in-hospital costs were the same for S-PVR and TC-PVR. Third, due to the identical inhospital costs combined with increased re-intervention rates with TC-PVR, the projected 5-year
AC
cost of TC-PVR is greater than the 5-year cost of S-PVR. Our finding of greater LOS and hospital resource utilization in patients with TOF
following S-PVR compared to TC-PVR is similar to both a prior study of only 30 patients with TC-PVR 6 and an additional study which included multiple types of congenital heart disease in addition to TOF 7. The promise of shorter hospitalizations, less ICU use, and less mechanical 8
ACCEPTED MANUSCRIPT
ventilation use with TC-PVR are borne out in the current study. However, our findings show that the hope of reduced mortality and lower costs with TC-PVR have not yet been realized. Our finding that in-hospital mortality is less than 1% in patients with TOF following SPVR is similar to prior studies 1,6. However, we found no difference in in-hospital mortality
CR IP T
following TC-PVR. This is a novel finding, as the only other paper to compare in-hospital
mortality in patients with TOF following TC-PVR and S-PVR included only 30 patients with TCPVR6. The likely explanation for the identical in-hospital mortality is that mortality with S-PVR is so low that it would be extremely difficult for TC-PVR to improve upon it.
AN US
In spite of increased hospital resource-utilization, we found that TC-PVR provides no inhospital cost savings over S-PVR. This finding is in keeping with other smaller studies, most of which included patients with multiple types of congenital heart disease3,5-7. A few prior studies with small cohorts have found possible in-hospital cost savings with TC-PVR; however, those
M
studies were based on either charges or indirect costs4,12. While other studies have projected total
ED
5-year costs with mixed results3,4, to our knowledge, our study is the first to utilize actual reintervention rates in the model. The recent publication of the 7-year results of the United States
PT
Melody Valve Investigational Device Exemption trial 8 has allowed for a more accurate estimation of total 5-year costs. Utilizing these re-intervention rates, we found that projected 5-
CE
year costs were greater with TC-PVR compared to S-PVR. One potential advantage of TC-PVR is that shorter hospital stays and more rapid
AC
recovery will allow patients and/or their parents to miss less work and thus decrease the cost to society. However, our study has shown that even when wage losses were included, TC-PVR remains more expensive when considered from a perspective of total 5-year cost. As with any new method or technology, with time there is the potential for the cost of TC-PVR to decrease. However, when accounting for inflation, the cost of TC-PVR remained stable from 2010 to 2016
9
ACCEPTED MANUSCRIPT
and is no different than the cost of S-PVR, primarily due to the high, fixed cost of the Ensemble delivery system® and the Melody valve® which was the primary trans-catheter pulmonary valve in use during the study period. However, if new companies enter the market, the device cost could come down, making TC-PVR more cost-effective. Additionally, with changes in technique
CR IP T
such as pre-stenting, the re-intervention rates appear to be decreasing13 which would further reduce total costs of TC-PVR. However, based on our results, the 5-year re-intervention rate for TC-PVR would have to decrease to <6% for TC-PVR to be more cost effective than S-PVR.
Our total projected 5-year costs did not account for differences in outpatient follow-up
AN US
and testing between patients with S-PVR and TC-PVR. Suggested follow-up after TC-PVR
includes clinical evaluation every six months; chest x-rays and echocardiograms at one month, three months, six months, one year following valve implantation, and then annually; fluoroscopy at six months; cardio-pulmonary exercise testing at six months and then annually; and cardiac
M
magnetic resonance imaging (MRI) and computed tomographic pulmonary angiography at six months14. In comparison, after S-PVR, published guidelines suggest annual clinical evaluations,
ED
echocardiograms every two years, MRI every three years, and intermittent cardio-pulmonary exercise testing15,16. In actual practice, those with TC-PVR may receive less intensive follow-up
PT
than is recommended, but they almost certainly receive more follow-up and testing than those
CE
with S-PVR. Additionally, our estimation did not directly account for the increased incidence of endocarditis associated with TC-PVR compared to S-PVR17. While the surgical and trans-catheter
AC
interventions due to endocarditis were accounted for in the re-intervention rates used in the model, admissions for medical management of endocarditis were not accounted for. This study has several limitations. As with any administrative database, the classification
of patients relies on accurate recording by the billing physicians and billing coders. The PHIS is a database of primarily diagnostic and procedure codes and thus does not include all clinically relevant information; for example, results of imaging studies and exercise testing are not 10
ACCEPTED MANUSCRIPT
included. In addition, the PHIS database is restricted to pediatric hospitals, and therefore the results of this study may not be generalizable to adults with TOF who undergo PVR outside of pediatric institutions. The timeframe of our study included the transition from ICD-9 to ICD-10. It is possible that identification of patients and outcomes changed with this transition. There is
CR IP T
also a possibility that some of the patients underwent TC-PVR because they were considered too high risk for S-PVR, and thus differences in baseline characteristics of the patients may explain
some of the study outcomes. It is possible that some of the patients who underwent S-PVR were not candidates for TC-PVR because of the limitations of performing TC-PVR in a native outflow
AN US
track. Due to the nature of the PHIS database, neither the indications for PVR nor the rationale
for choosing TC-PVR versus S-PVR could be determined. In addition, the projected 5-year costs are based on multiple assumptions including re-intervention rates that were published from early experiences of TC-PVR, and thus may not accurately represent actual 5-year costs.
M
In conclusion, both S-PVR and TC-PVR in patients with repaired TOF carry low inhospital morbidity and mortality. In spite of longer LOS and greater in-hospital resource
PT
those for TC-PVR.
ED
utilization after S-PVR, in-hospital costs are the same, and projected 5-year costs are less than
AC
CE
Acknowledgement: None
11
ACCEPTED MANUSCRIPT
1. Cavalcanti PEF, Sá MPBO, Santos CA, Esmeraldo IM, de Escobar RR, de Menezes AM, de Azevedo OM, de Vasconcelos Silva FP, de Albuquerque Lins RF, de Carvalho Lima R. Pulmonary valve replacement after operative repair of tetralogy of Fallot: meta-analysis and meta-regression of 3,118 patients from 48 studies. J Am Coll Cardiol 2013;62:2227-2243.
CR IP T
2. Cheung EW-Y, Wong WH-S, Cheung Y-F. Meta-analysis of pulmonary valve replacement after operative repair of tetralogy of fallot. Am J Cardiol 2010;106:552-557.
3. Gatlin SW, Kim DW, Mahle WT. Cost analysis of percutaneous pulmonary valve
AN US
replacement. Am J Cardiol 2011;108:572-574.
4. Vergales JE, Wanchek T, Novicoff W, Kron IL, Lim DS. Cost‐analysis of percutaneous pulmonary valve implantation compared to surgical pulmonary valve replacement. Cathet Cardiovasc Interven 2013;82:1147-1153.
M
5. Steinberg ZL, Jones TK, Verrier E, Stout KK, Krieger EV, Karamlou T. Early outcomes in
ED
patients undergoing transcatheter versus surgical pulmonary valve replacement. Heart 2017:heartjnl-2016-310776.
PT
6. O’Byrne ML, Glatz AC, Mercer-Rosa L, Gillespie MJ, Dori Y, Goldmuntz E, Kawut S, Rome JJ.
CE
Trends in pulmonary valve replacement in children and adults with tetralogy of fallot. Am J Cardiol 2015;115:118-124.
AC
7. O'Byrne ML, Gillespie MJ, Shinohara RT, Dori Y, Rome JJ, Glatz AC. Cost comparison of transcatheter and operative pulmonary valve replacement (from the Pediatric Health Information Systems Database). Am J Cardiol 2016;117:121-126. 8. Cheatham JP, Hellenbrand WE, Zahn EM, Jones TK, Berman DP, Vincent JA, McElhinney DB. Clinical and hemodynamic outcomes up to 7 years after transcatheter pulmonary valve
12
ACCEPTED MANUSCRIPT
replacement in the US melody valve investigational device exemption trial. Circulation 2015:CIRCULATIONAHA. 114.013588. 9. Blue Cross of Idaho. Transcatheter Pulmonary Valve Implantation, 2011. Available at: https://www.bcidaho.com/providers/medical_policies/sur/mp_701131.asp. Accessed June
CR IP T
20, 2017.
10. Hawaii Medical Service Association. Transcatheter Pulmonary Valve Implantation, 2013. Available at:
ion_012315.pdf. Accessed June 20, 2017.
AN US
https://hmsa.com/portal/provider/MM.06.022_Transcatheter_Pulmonary_Valve_Implantat
11. Oosterhof T, Meijboom FJ, Vliegen HW, Hazekamp MG, Zwinderman AH, Bouma BJ, van Dijk AP, Mulder BJ. Long-term follow-up of homograft function after pulmonary valve
M
replacement in patients with tetralogy of Fallot. Eur Heart J 2006;27:1478-1484.
ED
12. Andresen B, Mishra V, Lewandowska M, Andersen JG, Andersen MH, Lindberg H, Døhlen G, Fosse E. In-hospital cost comparison between percutaneous pulmonary valve
PT
implantation and surgery. Eur J Cardiothorac Surg 2016:ezw378. 13. Cabalka AK, Hellenbrand WE, Eicken A, Kreutzer J, Gray RG, Bergersen L, Berger F,
CE
Armstrong AK, Cheatham JP, Zahn EM, McElhinney DB. Relationships Among Conduit Type,
AC
Pre-Stenting, and Outcomes in Patients Undergoing Transcatheter Pulmonary Valve Replacement in the Prospective North American and European Melody Valve Trials. JACC Cardiovasc Interv 2017;10:1746-1759. 14. Zahn EM, Hellenbrand WE, Lock JE, McElhinney DB. Implantation of the Melody transcatheter pulmonary valve in patients with a dysfunctional right ventricular outflow tract conduit: early results from the US clinical trial. J Am Coll Cardiol 2009;54:1722-1729. 13
ACCEPTED MANUSCRIPT
15. Krishnamurthy R, Taylor AM, Warnes CA. Multimodality imaging guidelines for patients with repaired tetralogy of fallot: a report from the American Society of Echocardiography. J Am Soc Echocardiogr 2014;27:111-141. 16. Warnes CA, Williams RG, Bashore TM, Child JS, Connolly HM, Dearani JA, del Nido P,
CR IP T
Fasules JW, Graham TP, Hijazi ZM. ACC/AHA 2008 guidelines for the management of adults with congenital heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Develop
Guidelines on the Management of Adults With Congenital Heart Disease) Developed in
AN US
Collaboration With the American Society of Echocardiography, Heart Rhythm Society, International Society for Adult Congenital Heart Disease, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. J Am Coll Cardiol
M
2008;52:e143-e263.
17. Van Dijck I, Budts W, Cools B, Eyskens B, Boshoff DE, Heying R, Frerich S, Vanagt WY,
ED
Troost E, Gewillig M. Infective endocarditis of a transcatheter pulmonary valve in
AC
CE
PT
comparison with surgical implants. Heart 2014:heartjnl-2014-306761.
14
AN US
CR IP T
ACCEPTED MANUSCRIPT
Figure A: Estimated Cost for Pulmonary Valve Replacement Admissions Over Time
M
TC-PVR = trans-catheter pulmonary valve replacement, S-PVR = surgical pulmonary valve replacement, 2016US$ = United States dollars adjusted to inflation to year 2016
AC
CE
PT
ED
The middle line in the box corresponds to the median. The lower and upper hinges correspond to the 1st and 3rd quartiles. Interquartile range (IQR) is the distance between the 1st and 3rd quartiles. The upper whisker extends from the hinge to the largest value no further than 1.5 × IQR from the hinge, and the lower whisker extends from the hinge to the smallest value at most 1.5×IQR from the hinge.
15
M
AN US
CR IP T
ACCEPTED MANUSCRIPT
ED
Figure B: Five-year Cost with Varying Pulmonary Valve Replacement Rates
PT
One-way sensitivity analysis was performed allowing the 5-year rate of pulmonary valve replacement (including both surgical redo and trans-catheter re-implantation) after TC-PVR to vary. The 5-year rate of pulmonary valve replacement following S-PVR was kept constant and assumed to be 7.2% based on prior meta-analyses.
AC
CE
TC-PVR = trans-catheter pulmonary valve replacement, S-PVR = surgical pulmonary valve replacement, IDE = United States Melody Valve Investigational Device Exemption Trial
16
M
AN US
CR IP T
ACCEPTED MANUSCRIPT
ED
Figure C: Two-way Sensitivity Analysis on Surgical Redo and Trans-catheter Reimplantation Rates
PT
Two-way sensitivity analysis was performed allowing both the surgical redo rate and the transcatheter re-implantation rate after TC-PVR to vary.
CE
TC-PVR = trans-catheter pulmonary valve replacement, S-PVR = surgical pulmonary valve replacement, IDE = United States Melody Valve Investigational Device Exemption Trial
AC
Table 1: Study population of patients undergoing pulmonary valve replacement
Entire Cohort Trans-catheter
Surgical
After Matching Tran scathe ter
Surgi cal
17
ACCEPTED MANUSCRIPT
n=194
n=1072
pvalue
n=19 1
n=38 2
pval ue‡
<0.00 1
17 (IQR: 1324)
16 (IQR: 1222)
0.50
0.87
110 (58% )
236 (62% )
0.33
27 (15% )
50 (15% )
0.84
17 (IQR: 13-24)
Age (years)
15 (IQR: 11-19)
111 (57%)
Male
620 (58%)
27 (15%)
Hispanic
120 (13%)
CR IP T
†
0.44
0.94
131 (73%)
Black
20 (11%)
Asian
6 (3%)
128 (73% )
259 (72% )
117 (11%)
20 (11% )
46 (13% )
44 (4%)
6 (3%)
17 (5%)
27 (14%)
133 (12%)
22 (12% )
40 (11% )
48,186 (IQR: 37,218-63,560)
48,386 (IQR: 38,600-65,133)
48,27 5 (IQR: 37,27 163,43 4)
47,71 3 (IQR: 39,28 563,77 6)
ED PT
Other
0.31
AC
CE
Median household income (2016US$)
0.83
739 (72%)
M
White
AN US
Race
0.44
Payor
0.79
0.93
Private
97 (51%)
596 (56%)
97 (51% )
200 (52% )
Medicaid
67 (35%)
340 (32%)
67 (35% )
131 (34% )
18
ACCEPTED MANUSCRIPT
Other
Genetic diagnosis
27 (14%)
133 (12%)
25 (13%)
142 (13%)
0.89
27 (14% )
51 (13% )
25 (13% )
40 (10% )
0.33
CR IP T
IQR = interquartile range (1st quartile – 3rd quartile), 2016US$ = United States dollars adjusted to inflation to year 2016 Summary statistics were expressed as median (IQR) for continuous variables and percentage (count) for categorical variables. †
AN US
P-values were obtained from comparisons between TC-PVR and S-PVR using Wilcoxon rank-sum test for continuous variables and chi-square test for categorical variables. ‡
P-values were obtained from comparisons between matched TC-PVR and SPVR using generalized linear mixed model.
M
Table 2: Comparison of outcomes between trans-catheter and surgical pulmonary valve replacement
ED
Entire Cohort
n=194
Surgical
n=1072
CE
PT
Trans-catheter
2011
2012
AC
pvalue †
Tran scathe ter
Surgi cal
n=19 1
n=38 2
<0.0 01
Discharge year
2010
After Matching
pvalu e‡ 0.95
6 (3%)
156 (15%)
6 (3%)
17 (4%)
7 (4%)
164 (15%)
7 (4%)
11 (3%)
25 (13%)
144 (13%)
25 (13%
44 (12%
19
ACCEPTED MANUSCRIPT
)
)
159 (15%)
30 (16% )
55 (14% )
162 (15%)
35 (18% )
66 (17% )
2013 30 (15%)
36 (19%)
2015 37 (19%)
150 (14%)
53 (27%)
137 (13%)
1 (1%)
9 (1%)
Discharge mortality
145,413 (IQR: 98,486-203,860)
75 (20% )
53 (28% )
114 (30% )
0.64
1 (1%)
3 (1%)
0.73
148,020 (IQR: 111,594-192,647)
0.99
147,0 75 (IQR: 98,20 5205,2 02)
159,7 15 (IQR: 117,6 21201,8 79)
0.24
53,00 7 (IQR: 38,82 764,77 0)
52,20 8 (IQR: 43,44 667,09 1)
0.20
M
Adjusted billed charge (2016US$)
35 (18% )
AN US
2016
CR IP T
2014
52,935 (IQR: 39,356-64,637)
52,201 (IQR: 42,627-66,712)
0.32
CE
PT
ED
Estimated cost (2016US$)
AC
Length of stay
Intensive care unit
Intensive care unit length of stay
Mechanical ventilation
1 (IQR: 1-2)
4 (IQR: 4-6)
<0.0 01
1 (IQR: 1-2)
5 (IQR: 4-6)
<0.0 01
93 (48%)
934 (87%)
<0.0 01
92 (48% )
342 (90% )
<0.0 01
0 (IQR: 0-1)
2 (IQR: 1-3)
<0.0 01
0 (IQR: 0-1)
2 (IQR: 1-3)
0.00 2
21 (11%)
589 (55%)
<0.0
21 (11%
57
<0.0
20
ACCEPTED MANUSCRIPT
01
)
(218
01
0 (IQR: 0-0)
1 (IQR: 0-1)
<0.0 01
0 (IQR: 0-0)
1 (IQR: 0-1)
<0.0 01
Extracorporeal membrane oxygenation
0 (0%)
21 (2%)
0.04 9
0 (0%)
6 (2%)
0.08
Acute kidney failure
3 (2%)
26 (2%)
0.45
3 (2%)
12 (3%)
0.28
Surgical complication (any)
94 (48%)
565 (53%)
0.28
93 (49% )
193 (51% )
0.68
81 (42% )
111 (29% )
0.00 2
8 (4%)
Hemorrhage/ hematoma
9 (5%)
M
Respiratory
2 (1%)
ED
Infection/ fever
7 (4%)
Other
352 (33%)
0.01
AN US
82 (42%)
Cardiac
CR IP T
Mechanical ventilation days
262 (24%)
<0.0 01
8 (4%)
89 (23% )
<0.0 01
75 (7%)
0.22
9 (5%)
24 (6%)
0.45
44 (4%)
0.04
2 (1%)
19 (5%)
0.03
0.01
7 (4%)
44 (12% )
0.00 3
105 (10%)
†
CE
PT
IQR = interquartile range (1st quartile - 3rd quartile), 2016US$ = United States dollars adjusted to inflation to year 2016. Summary statistics were expressed as median (IQR) for continuous variables and percentage (count) for categorical variables. P-values were obtained from comparisons between TC-PVR and S-PVR using Wilcoxon rank-sum test for continuous variables and chi-square test for categorical variables. ‡
AC
P-values were obtained from comparisons between matched TC-PVR and SPVR using generalized linear mixed model.
Table 3: Projected 5-Year Cost of Pulmonary Valve Replacement
21
ACCEPTED MANUSCRIPT
Surgical
Initial total hospital cost
$52,935
$52,201
Surgical redo cost (rate)
$4,176 (8%)
$3,758 (7.2%)
$769 (4%)
$577 (3%)
$6,882 (13%)
$0 (0%)
Valvuloplasty/angioplasty cost (rate) Trans-catheter pulmonary valve re-implantation cost (rate)
CR IP T
Trans-catheter
$1,241
$2,987
Projected total 5-year cost to payor (wage loss not included)
$64,762
$56,536
Projected total 5-year cost to society (wage loss included)
$66,003
$59,523
AC
CE
PT
ED
M
AN US
Total Wage Loss
22
Trans-catheter Versus Surgical Pulmonary Valve Replacement in Repaired Tetralogy of Fallot Joshua A. Daily MDMEd , Xinyu Tang PhD , Michael Angtuaco MD , Elijah Bolin MD , Sean M. Lang MD , R. Thomas Collins II MD PII: DOI: Reference:
S0002-9149(18)31024-5 10.1016/j.amjcard.2018.04.028 AJC 23269
To appear in:
The American Journal of Cardiology
Received date: Revised date: Accepted date:
16 January 2018 4 April 2018 6 April 2018
Please cite this article as: Joshua A. Daily MDMEd , Xinyu Tang PhD , Michael Angtuaco MD , Elijah Bolin MD , Sean M. Lang MD , R. Thomas Collins II MD , Trans-catheter Versus Surgical Pulmonary Valve Replacement in Repaired Tetralogy of Fallot, The American Journal of Cardiology (2018), doi: 10.1016/j.amjcard.2018.04.028
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Trans-catheter Versus Surgical Pulmonary Valve Replacement in Repaired Tetralogy of Fallot
MDa, b, Sean M. Langa, b, MD, R. Thomas Collins, II, MDc, d
Affiliations: aArkansas Children’s Hospital, Little Rock, AR b
Stanford University School of Medicine, Palo Alto, CA
d
AN US
c
University of Arkansas for Medical Sciences, Little Rock, AR
CR IP T
Joshua A. Dailya, b, MD, MEd, Xinyu Tang, PhDa, b, Michael Angtuacoa, b, MD, Elijah Bolin,
Lucile Packard Children’s Hospital Stanford, Palo Alto, CA
Conflict of Interest Statement: The authors have no financial relationships relevant to this article
Corresponding Author:
ED
M
to disclose.
PT
Joshua A. Daily, MD, MEd
Arkansas Children’s Hospital
CE
1 Children’s Way, Slot 512-3 Little Rock, AR 72202
AC
Phone: 501-364-2012
Email: [email protected]
Running Head: Trans-catheter Versus Surgical Pulmonary Valve Replacement
1
ACCEPTED MANUSCRIPT
Abstract Trans-catheter pulmonary valve replacement (TC-PVR) is an alternative to surgical PVR (SPVR) in repaired Tetralogy of Fallot (TOF). The purpose of this study is to compare in-hospital outcomes, hospital costs, and projected 5-year total costs of S-PVR to TC-PVR in patients with
CR IP T
repaired TOF. We performed a multi-center, retrospective cohort study of children and adults
with TOF ≥ 8 years of age who underwent PVR from January 1, 2010 to December 31, 2016 at 46 centers contributing to the Pediatric Health Information Systems database. Baseline
characteristics, in-hospital outcomes, and costs were compared between the two groups. Projected
AN US
5-year costs were calculated by combining cost data with published re-intervention rates. A total of 194 TC-PVR and 1072 S-PVR were performed. The baseline characteristics of the TC-PVR and S-PVR groups were not significantly different with the exception of greater age in the TCPVR group (median age of 17 years vs 15 years, p value <0.001). Discharge mortality, hospital
M
charges and estimated cost, surgical complication rates, and acute kidney failure were not significantly different between the groups. Intensive care unit (ICU) use, ICU length of stay
ED
(LOS), mechanical ventilation use, extracorporeal membrane oxygenation use, and total LOS were lower with TC-PVR than S-PVR. Projected 5-year costs were greater with TC-PVR
PT
compared to S-PVR ($64,762 vs $56,536) due to the cost of the trans-catheter pulmonary valve
CE
and higher re-intervention rates. In conclusion, in spite of longer LOS and greater in-hospital resource utilization for patients with TOF undergoing S-PVR compared to TC-PVR, mortality
AC
and in-hospital costs are the same, and projected 5-year costs are less. Key Words: Trans-catheter Pulmonary Valve Replacement, Tetralogy of Fallot
2
ACCEPTED MANUSCRIPT
Initial repair of Tetralogy of Fallot (TOF) often results in severe pulmonary valve insufficiency that leads to a chronic volume load on the right ventricle and the need for multiple pulmonary valve replacements (PVR) throughout a patient’s lifetime. Surgical PVR (S-PVR) was long the standard of care for severe pulmonary insufficiency in repaired TOF 1,2. Recently, trans-catheter
CR IP T
PVR (TC-PVR) has been developed as an alternative to S-PVR which theoretically avoids the morbidity, mortality, and costs associated with open-heart surgery. Previous attempts to compare the cost of S-PVR and TC-PVR have been made with conflicting results 3-7. These comparisons utilized heterogeneous cohorts with multiple types of congenital heart disease, small patient
AN US
numbers, single centers, and/or charge data instead of cost data. In addition, attempts at estimating the 5-year total costs of both approaches have been based on theoretical reintervention rates. Recently, the 7-year results from the United States Melody Valve
Investigational Device Exemption study were published, allowing for a more accurate estimation
M
of 5-year cost of TC-PVR 8. The purpose of this current study was to compare in-hospital outcomes, hospital costs, and projected 5-year costs for patients with TOF who undergo either
ED
TC-PVR or S-PVR.
PT
Methods
With the approval of the Institutional Review Board of the University of Arkansas for
CE
Medical Sciences, data were obtained from the Pediatric Health Information Systems (PHIS), an administrative database of 46 not-for-profit, tertiary care pediatric hospitals in the United States
AC
affiliated with the Children’s Health Corporation of America. The PHIS data include detailed, deidentified information on each inpatient’s demographics, diagnoses, procedures, medications, outcomes, and cost. Data quality assurance is ongoing and data from the individual hospitals are accepted when classified errors for a given quarter occur less frequently than a criterion threshold of 2%.
3
ACCEPTED MANUSCRIPT
The study design was a multi-center, retrospective cohort investigation of all children and adults ≥ 8 years of age with a diagnosis of TOF (International Classification of Diseases, Ninth Revision [ICD-9], code 745.2 or International Classification of Diseases, Tenth Revision [ICD10], code Q21.3) who underwent either S-PVR or TC-PVR at any of the PHIS hospitals from
CR IP T
January 1, 2010 to December 31, 2016. Since hospital reporting transitioned from ICD-9 to ICD10 on October 1, 2015, ICD-9 codes were used prior to this date, and ICD-10 codes were used
after. Patients were assigned to the S-PVR group if the following codes were used: 35.25, 35.92, 33475, 02RH08Z, 02RH0JZ, 02RH0KZ, 02RH07Z 021K0KP, 021K09Q, 021K08P, 021K08Q,
AN US
021K08R, 021K09P, 021K09R, 021K0AP, 021K0AQ, 021K0AR, 021K0JP, 021K0JQ,
021K0JR, 021K0KQ, 021K0KR, 021K0ZP, 021K0ZQ, or 021K0ZR. Patients were assigned to the TC-PVR group if the following codes were used: 35.07, 35.08, 35.09, 33477, 0262T, 02RH48Z, 02RH38Z, 02RH3JZ, 02RH4KZ, 02RH4JZ, 02RH37Z, 02RH3KZ, 02RH37H,
M
02RH38H, 02RH3JH, 02RH3KH, 02RH47Z, or 02RH4JZ. After review of multiple instructions for billing of TC-PVR, it was noted that some included the use of ICD-9 code 35.26 which was
ED
also commonly used for S-PVR 9,10. Patients with an ICD-9 code of 35.26 were reviewed in detail by the investigators, and the following schema was developed to place them in the appropriate
PT
groups. Patients with ICD-9 code 35.26 were placed in the S-PVR group if the Operative Room Charge Flag was present and procedure code 3961 (Extracorporeal circulation auxiliary to open
CE
heart surgery) was utilized. Conversely, patients were placed in the TC-PVR group if the Operative Room Charge Flag was not present and procedure code 3723 (combined right and left
AC
heart catheterization) was used. Data were extracted from the PHIS database by direct query and included age, gender,
race, insurance payor (private, Medicaid, and other), median household income by patient’s zip code, and presence of a genetic syndrome (758.1, 758.2, 758.0, 758.31, 758.32, 758.33, 758.9, 758.6, 758.89, 279.11, Q91.7, Q91.3, Q90.9, Q93.4, Q93.81, Q93.88, Q99.9, Q96.9, Q99.8, or
4
ACCEPTED MANUSCRIPT
D82.1). Discharge mortality, intensive care unit (ICU) use, hospital length of stay (LOS), ICU LOS, mechanical ventilation, mechanical ventilation days, extracorporeal membrane oxygenation (ECMO), acute kidney failure (586, 584.9, N17.0, N17.1, N17.2, N17.8, or N17.9), and surgical complications were also extracted. The codes that triggered the surgical complication flag were
CR IP T
additionally evaluated and divided into the following categories: cardiac, respiratory, hemorrhage/ hematoma, infection/ fever, and other. The use of mechanical ventilation referred
only to use during the ICU stay (mechanical ventilation in the operating room or catheterization laboratory was not considered). A day was counted as one mechanical ventilation day if a patient
AN US
received mechanical ventilation in the ICU at any point during that day. Estimated total costs were also extracted. The PHIS database calculates estimated total cost by converting charges using ratios of cost-to-charges and adjusting for regional differences (based on wage/price indices). All costs were adjusted for inflation to 2016 US (United States) dollars using the
M
Consumer Pricing Indexes Data.
Descriptive statistics were expressed as median with interquartile ranges (IQRs: 1st
ED
quartile (Q1)-3rd quartile (Q3)) for continuous variables and percentage (count) for categorical variables. The distributions of continuous variables were compared between unmatched groups
PT
using the Wilcoxon rank-sum test. The proportions of categorical variables were compared
CE
between unmatched groups using the chi-square test. Boxplot was drawn without outliers for estimated cost in 2016 US dollars over time among TC-PVR and S-PVR respectively. Outliers
AC
were defined as values less than (Q1 - 1.5×IQR) or greater than (Q3 + 1.5×IQR). Propensity score matching was performed between the TC-PVR and S-PVR groups. A
multivariable logistic regression model was fitted to estimate the probability of TC-PVR as a function of discharge year, admission age, insurance payor, and genetic diagnosis. Using the linear propensity score calculated based on the model, TC-PVR patients were matched to S-PVR patients at a 1:2 ratio. The 1:2 ratio was determined by evaluating the crude sample size ratio 5
ACCEPTED MANUSCRIPT
between TC-PVR and S-PVR by discharge year. Generalized linear mixed model was used for comparing baseline characteristics and outcomes between matched TC-PVR and S-PVR. Appropriate distribution and link functions were used for different variable types. For example, Gaussian distribution and identity link was used for continuous variables, and binomial
CR IP T
distribution and logit link was used for binary variables. A random effect was included in the models to account for clustering after matching. P-values ≤ 0.05 were used to indicate statistical significance.
Projected 5-year total costs were estimated using re-intervention data for both the S-PVR
AN US
and TC-PVR. To determine the S-PVR re-operation rate, the results of two recent meta-analyses of S-PVR after operative TOF were averaged for a 5-year repeat PVR rate of 7.2% 1,2. For the purposes of comparison, it was assumed that all patients in the S-PVR group who required a repeat PVR underwent S-PVR. Balloon angioplasty or valvuloplasty rates were assumed to be 3%
M
at 5 years based on a previous study 11. Trans-catheter pulmonary valve failure rates were based on the five-year results of the United States Melody Valve Investigational Device Exemption trial . Of the 150 patients who underwent initial successful TC-PVR in the Investigational Device
ED
8
Exemption trial, at a median follow-up of 4.5 years, 12 (8%) had undergone surgery, 6 (4%) had
PT
undergone catheter dilation, and 20 (13.3%) underwent a second TC-PVR. These rates of re-
CE
intervention were used to estimate the 5-year costs for TC-PVR. We assumed that the cost for a second procedure was the same as the first. To estimate the cost of catheter dilation, the costs of
AC
the Melody valve®, Ensemble delivery system®, and one night hospitalization were subtracted from the estimated total costs for a hospitalization for TC-PVR. Wage loss was calculated as the product of estimated days of work lost and median household income based on the zip code of the patient. Total days of work lost were determined by adding the patient’s LOS to the recommended time away from work, which was assumed to be 1 week for TC-PVR and 3 weeks for S-PVR. For patients < 18-years-old, it was assumed that the patient’s parent was out of work
6
ACCEPTED MANUSCRIPT
for the recovery time. To convert median household income to median income of a single earner, the ratio of 1.28 earners per household was used based on 2010 census data. One- and two-way sensitivity analysis was performed. For one-way sensitivity analysis, the total PVR rate (including both surgical redo and trans-catheter re-implantation) after TC-PVR
CR IP T
was allowed to vary. Based on the results of the IDE study, PVRs after TC-PVR were assumed to be 38% (8/21) surgical and 62% (13/21) trans-catheter. The cost of PVR after TC-PVR was
assumed to be the same as the cost of an initial TC-PVR or S-PVR. For purposes of comparison, the surgical redo rate after S-PVR was assumed to be of 7.2% based on two prior meta-
AN US
analyses1,2. For two-way sensitivity analysis, both the surgical redo rate and the trans-catheter reimplantation rate after TC-PVR were allowed to vary to evaluate the effect that changing reintervention rates would have on the preference of S-PVR verses TC-PVR.
M
Results
From 2010 to 2016, a total of 1256 patients with TOF underwent PVR at participating
ED
PHIS centers. Five of these patients (0.4%) were excluded because they could not be assigned to either the TC-PVR or S-PVR group. Of the remaining 1251 patients, 1236 patients contributed
PT
one admission, and 15 patients contributed two admissions each. Of the 15 patients with two
CE
admissions, 11 patients underwent S-PVR twice, three patients underwent S-PVR followed by TC-PVR, and one patient underwent TC-PVR followed by S-PVR. In the entire cohort, there
AC
were 194 admissions with TC-PVR and 1072 admissions with S-PVR. The baseline characteristics of both groups of patients, with and without propensity score matching, can be found in Table 1. A comparison of in-hospital outcomes between TC-PVR and S-PVR can be found in Table 2. Discharge mortality, adjusted billed charges, estimated cost, acute kidney failure, and surgical complication (any) rate were not significantly different between the two groups. Cardiac 7
ACCEPTED MANUSCRIPT
surgical complications were more common with TC-PVR, and respiratory, hemorrhage/hematoma, infection/fever, and other complications were more common with SPVR. Admission to the ICU, ICU LOS, mechanical ventilation use, ECMO use, and total hospital LOS were lower with TC-PVR than S-PVR.
CR IP T
The median estimated cost for admission for TC-PVR and S-PVR are found in Figure A. When adjusted for inflation to 2016 US dollars, estimated cost did not change significantly for either group over the study period. Projected five-year costs can be found in Table 3. The
projected 5-year costs were greater for TC-PVR than S-PVR both from the perspective of the
AN US
payor (wage loss not considered) and from the perspective of society (wage loss included). These differences were primarily due to the higher re-intervention rates for TC-PVR compared to SPVR used in the model. The results of one- and two-way sensitivity analyses are found in Figures B and C. In both, S-PVR is more financially favorable than TC-PVR over a range of possible re-
M
intervention rates.
ED
Discussion
There were three important findings in this multi-center study investigating the largest
PT
cohort to date of patients undergoing TC-PVR versus S-PVR after initial palliation of TOF. First, LOS and hospital resource utilization were lower with TC-PVR than S-PVR. Second, mortality
CE
and in-hospital costs were the same for S-PVR and TC-PVR. Third, due to the identical inhospital costs combined with increased re-intervention rates with TC-PVR, the projected 5-year
AC
cost of TC-PVR is greater than the 5-year cost of S-PVR. Our finding of greater LOS and hospital resource utilization in patients with TOF
following S-PVR compared to TC-PVR is similar to both a prior study of only 30 patients with TC-PVR 6 and an additional study which included multiple types of congenital heart disease in addition to TOF 7. The promise of shorter hospitalizations, less ICU use, and less mechanical 8
ACCEPTED MANUSCRIPT
ventilation use with TC-PVR are borne out in the current study. However, our findings show that the hope of reduced mortality and lower costs with TC-PVR have not yet been realized. Our finding that in-hospital mortality is less than 1% in patients with TOF following SPVR is similar to prior studies 1,6. However, we found no difference in in-hospital mortality
CR IP T
following TC-PVR. This is a novel finding, as the only other paper to compare in-hospital
mortality in patients with TOF following TC-PVR and S-PVR included only 30 patients with TCPVR6. The likely explanation for the identical in-hospital mortality is that mortality with S-PVR is so low that it would be extremely difficult for TC-PVR to improve upon it.
AN US
In spite of increased hospital resource-utilization, we found that TC-PVR provides no inhospital cost savings over S-PVR. This finding is in keeping with other smaller studies, most of which included patients with multiple types of congenital heart disease3,5-7. A few prior studies with small cohorts have found possible in-hospital cost savings with TC-PVR; however, those
M
studies were based on either charges or indirect costs4,12. While other studies have projected total
ED
5-year costs with mixed results3,4, to our knowledge, our study is the first to utilize actual reintervention rates in the model. The recent publication of the 7-year results of the United States
PT
Melody Valve Investigational Device Exemption trial 8 has allowed for a more accurate estimation of total 5-year costs. Utilizing these re-intervention rates, we found that projected 5-
CE
year costs were greater with TC-PVR compared to S-PVR. One potential advantage of TC-PVR is that shorter hospital stays and more rapid
AC
recovery will allow patients and/or their parents to miss less work and thus decrease the cost to society. However, our study has shown that even when wage losses were included, TC-PVR remains more expensive when considered from a perspective of total 5-year cost. As with any new method or technology, with time there is the potential for the cost of TC-PVR to decrease. However, when accounting for inflation, the cost of TC-PVR remained stable from 2010 to 2016
9
ACCEPTED MANUSCRIPT
and is no different than the cost of S-PVR, primarily due to the high, fixed cost of the Ensemble delivery system® and the Melody valve® which was the primary trans-catheter pulmonary valve in use during the study period. However, if new companies enter the market, the device cost could come down, making TC-PVR more cost-effective. Additionally, with changes in technique
CR IP T
such as pre-stenting, the re-intervention rates appear to be decreasing13 which would further reduce total costs of TC-PVR. However, based on our results, the 5-year re-intervention rate for TC-PVR would have to decrease to <6% for TC-PVR to be more cost effective than S-PVR.
Our total projected 5-year costs did not account for differences in outpatient follow-up
AN US
and testing between patients with S-PVR and TC-PVR. Suggested follow-up after TC-PVR
includes clinical evaluation every six months; chest x-rays and echocardiograms at one month, three months, six months, one year following valve implantation, and then annually; fluoroscopy at six months; cardio-pulmonary exercise testing at six months and then annually; and cardiac
M
magnetic resonance imaging (MRI) and computed tomographic pulmonary angiography at six months14. In comparison, after S-PVR, published guidelines suggest annual clinical evaluations,
ED
echocardiograms every two years, MRI every three years, and intermittent cardio-pulmonary exercise testing15,16. In actual practice, those with TC-PVR may receive less intensive follow-up
PT
than is recommended, but they almost certainly receive more follow-up and testing than those
CE
with S-PVR. Additionally, our estimation did not directly account for the increased incidence of endocarditis associated with TC-PVR compared to S-PVR17. While the surgical and trans-catheter
AC
interventions due to endocarditis were accounted for in the re-intervention rates used in the model, admissions for medical management of endocarditis were not accounted for. This study has several limitations. As with any administrative database, the classification
of patients relies on accurate recording by the billing physicians and billing coders. The PHIS is a database of primarily diagnostic and procedure codes and thus does not include all clinically relevant information; for example, results of imaging studies and exercise testing are not 10
ACCEPTED MANUSCRIPT
included. In addition, the PHIS database is restricted to pediatric hospitals, and therefore the results of this study may not be generalizable to adults with TOF who undergo PVR outside of pediatric institutions. The timeframe of our study included the transition from ICD-9 to ICD-10. It is possible that identification of patients and outcomes changed with this transition. There is
CR IP T
also a possibility that some of the patients underwent TC-PVR because they were considered too high risk for S-PVR, and thus differences in baseline characteristics of the patients may explain
some of the study outcomes. It is possible that some of the patients who underwent S-PVR were not candidates for TC-PVR because of the limitations of performing TC-PVR in a native outflow
AN US
track. Due to the nature of the PHIS database, neither the indications for PVR nor the rationale
for choosing TC-PVR versus S-PVR could be determined. In addition, the projected 5-year costs are based on multiple assumptions including re-intervention rates that were published from early experiences of TC-PVR, and thus may not accurately represent actual 5-year costs.
M
In conclusion, both S-PVR and TC-PVR in patients with repaired TOF carry low inhospital morbidity and mortality. In spite of longer LOS and greater in-hospital resource
PT
those for TC-PVR.
ED
utilization after S-PVR, in-hospital costs are the same, and projected 5-year costs are less than
AC
CE
Acknowledgement: None
11
ACCEPTED MANUSCRIPT
1. Cavalcanti PEF, Sá MPBO, Santos CA, Esmeraldo IM, de Escobar RR, de Menezes AM, de Azevedo OM, de Vasconcelos Silva FP, de Albuquerque Lins RF, de Carvalho Lima R. Pulmonary valve replacement after operative repair of tetralogy of Fallot: meta-analysis and meta-regression of 3,118 patients from 48 studies. J Am Coll Cardiol 2013;62:2227-2243.
CR IP T
2. Cheung EW-Y, Wong WH-S, Cheung Y-F. Meta-analysis of pulmonary valve replacement after operative repair of tetralogy of fallot. Am J Cardiol 2010;106:552-557.
3. Gatlin SW, Kim DW, Mahle WT. Cost analysis of percutaneous pulmonary valve
AN US
replacement. Am J Cardiol 2011;108:572-574.
4. Vergales JE, Wanchek T, Novicoff W, Kron IL, Lim DS. Cost‐analysis of percutaneous pulmonary valve implantation compared to surgical pulmonary valve replacement. Cathet Cardiovasc Interven 2013;82:1147-1153.
M
5. Steinberg ZL, Jones TK, Verrier E, Stout KK, Krieger EV, Karamlou T. Early outcomes in
ED
patients undergoing transcatheter versus surgical pulmonary valve replacement. Heart 2017:heartjnl-2016-310776.
PT
6. O’Byrne ML, Glatz AC, Mercer-Rosa L, Gillespie MJ, Dori Y, Goldmuntz E, Kawut S, Rome JJ.
CE
Trends in pulmonary valve replacement in children and adults with tetralogy of fallot. Am J Cardiol 2015;115:118-124.
AC
7. O'Byrne ML, Gillespie MJ, Shinohara RT, Dori Y, Rome JJ, Glatz AC. Cost comparison of transcatheter and operative pulmonary valve replacement (from the Pediatric Health Information Systems Database). Am J Cardiol 2016;117:121-126. 8. Cheatham JP, Hellenbrand WE, Zahn EM, Jones TK, Berman DP, Vincent JA, McElhinney DB. Clinical and hemodynamic outcomes up to 7 years after transcatheter pulmonary valve
12
ACCEPTED MANUSCRIPT
replacement in the US melody valve investigational device exemption trial. Circulation 2015:CIRCULATIONAHA. 114.013588. 9. Blue Cross of Idaho. Transcatheter Pulmonary Valve Implantation, 2011. Available at: https://www.bcidaho.com/providers/medical_policies/sur/mp_701131.asp. Accessed June
CR IP T
20, 2017.
10. Hawaii Medical Service Association. Transcatheter Pulmonary Valve Implantation, 2013. Available at:
ion_012315.pdf. Accessed June 20, 2017.
AN US
https://hmsa.com/portal/provider/MM.06.022_Transcatheter_Pulmonary_Valve_Implantat
11. Oosterhof T, Meijboom FJ, Vliegen HW, Hazekamp MG, Zwinderman AH, Bouma BJ, van Dijk AP, Mulder BJ. Long-term follow-up of homograft function after pulmonary valve
M
replacement in patients with tetralogy of Fallot. Eur Heart J 2006;27:1478-1484.
ED
12. Andresen B, Mishra V, Lewandowska M, Andersen JG, Andersen MH, Lindberg H, Døhlen G, Fosse E. In-hospital cost comparison between percutaneous pulmonary valve
PT
implantation and surgery. Eur J Cardiothorac Surg 2016:ezw378. 13. Cabalka AK, Hellenbrand WE, Eicken A, Kreutzer J, Gray RG, Bergersen L, Berger F,
CE
Armstrong AK, Cheatham JP, Zahn EM, McElhinney DB. Relationships Among Conduit Type,
AC
Pre-Stenting, and Outcomes in Patients Undergoing Transcatheter Pulmonary Valve Replacement in the Prospective North American and European Melody Valve Trials. JACC Cardiovasc Interv 2017;10:1746-1759. 14. Zahn EM, Hellenbrand WE, Lock JE, McElhinney DB. Implantation of the Melody transcatheter pulmonary valve in patients with a dysfunctional right ventricular outflow tract conduit: early results from the US clinical trial. J Am Coll Cardiol 2009;54:1722-1729. 13
ACCEPTED MANUSCRIPT
15. Krishnamurthy R, Taylor AM, Warnes CA. Multimodality imaging guidelines for patients with repaired tetralogy of fallot: a report from the American Society of Echocardiography. J Am Soc Echocardiogr 2014;27:111-141. 16. Warnes CA, Williams RG, Bashore TM, Child JS, Connolly HM, Dearani JA, del Nido P,
CR IP T
Fasules JW, Graham TP, Hijazi ZM. ACC/AHA 2008 guidelines for the management of adults with congenital heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Develop
Guidelines on the Management of Adults With Congenital Heart Disease) Developed in
AN US
Collaboration With the American Society of Echocardiography, Heart Rhythm Society, International Society for Adult Congenital Heart Disease, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. J Am Coll Cardiol
M
2008;52:e143-e263.
17. Van Dijck I, Budts W, Cools B, Eyskens B, Boshoff DE, Heying R, Frerich S, Vanagt WY,
ED
Troost E, Gewillig M. Infective endocarditis of a transcatheter pulmonary valve in
AC
CE
PT
comparison with surgical implants. Heart 2014:heartjnl-2014-306761.
14
AN US
CR IP T
ACCEPTED MANUSCRIPT
Figure A: Estimated Cost for Pulmonary Valve Replacement Admissions Over Time
M
TC-PVR = trans-catheter pulmonary valve replacement, S-PVR = surgical pulmonary valve replacement, 2016US$ = United States dollars adjusted to inflation to year 2016
AC
CE
PT
ED
The middle line in the box corresponds to the median. The lower and upper hinges correspond to the 1st and 3rd quartiles. Interquartile range (IQR) is the distance between the 1st and 3rd quartiles. The upper whisker extends from the hinge to the largest value no further than 1.5 × IQR from the hinge, and the lower whisker extends from the hinge to the smallest value at most 1.5×IQR from the hinge.
15
M
AN US
CR IP T
ACCEPTED MANUSCRIPT
ED
Figure B: Five-year Cost with Varying Pulmonary Valve Replacement Rates
PT
One-way sensitivity analysis was performed allowing the 5-year rate of pulmonary valve replacement (including both surgical redo and trans-catheter re-implantation) after TC-PVR to vary. The 5-year rate of pulmonary valve replacement following S-PVR was kept constant and assumed to be 7.2% based on prior meta-analyses.
AC
CE
TC-PVR = trans-catheter pulmonary valve replacement, S-PVR = surgical pulmonary valve replacement, IDE = United States Melody Valve Investigational Device Exemption Trial
16
M
AN US
CR IP T
ACCEPTED MANUSCRIPT
ED
Figure C: Two-way Sensitivity Analysis on Surgical Redo and Trans-catheter Reimplantation Rates
PT
Two-way sensitivity analysis was performed allowing both the surgical redo rate and the transcatheter re-implantation rate after TC-PVR to vary.
CE
TC-PVR = trans-catheter pulmonary valve replacement, S-PVR = surgical pulmonary valve replacement, IDE = United States Melody Valve Investigational Device Exemption Trial
AC
Table 1: Study population of patients undergoing pulmonary valve replacement
Entire Cohort Trans-catheter
Surgical
After Matching Tran scathe ter
Surgi cal
17
ACCEPTED MANUSCRIPT
n=194
n=1072
pvalue
n=19 1
n=38 2
pval ue‡
<0.00 1
17 (IQR: 1324)
16 (IQR: 1222)
0.50
0.87
110 (58% )
236 (62% )
0.33
27 (15% )
50 (15% )
0.84
17 (IQR: 13-24)
Age (years)
15 (IQR: 11-19)
111 (57%)
Male
620 (58%)
27 (15%)
Hispanic
120 (13%)
CR IP T
†
0.44
0.94
131 (73%)
Black
20 (11%)
Asian
6 (3%)
128 (73% )
259 (72% )
117 (11%)
20 (11% )
46 (13% )
44 (4%)
6 (3%)
17 (5%)
27 (14%)
133 (12%)
22 (12% )
40 (11% )
48,186 (IQR: 37,218-63,560)
48,386 (IQR: 38,600-65,133)
48,27 5 (IQR: 37,27 163,43 4)
47,71 3 (IQR: 39,28 563,77 6)
ED PT
Other
0.31
AC
CE
Median household income (2016US$)
0.83
739 (72%)
M
White
AN US
Race
0.44
Payor
0.79
0.93
Private
97 (51%)
596 (56%)
97 (51% )
200 (52% )
Medicaid
67 (35%)
340 (32%)
67 (35% )
131 (34% )
18
ACCEPTED MANUSCRIPT
Other
Genetic diagnosis
27 (14%)
133 (12%)
25 (13%)
142 (13%)
0.89
27 (14% )
51 (13% )
25 (13% )
40 (10% )
0.33
CR IP T
IQR = interquartile range (1st quartile – 3rd quartile), 2016US$ = United States dollars adjusted to inflation to year 2016 Summary statistics were expressed as median (IQR) for continuous variables and percentage (count) for categorical variables. †
AN US
P-values were obtained from comparisons between TC-PVR and S-PVR using Wilcoxon rank-sum test for continuous variables and chi-square test for categorical variables. ‡
P-values were obtained from comparisons between matched TC-PVR and SPVR using generalized linear mixed model.
M
Table 2: Comparison of outcomes between trans-catheter and surgical pulmonary valve replacement
ED
Entire Cohort
n=194
Surgical
n=1072
CE
PT
Trans-catheter
2011
2012
AC
pvalue †
Tran scathe ter
Surgi cal
n=19 1
n=38 2
<0.0 01
Discharge year
2010
After Matching
pvalu e‡ 0.95
6 (3%)
156 (15%)
6 (3%)
17 (4%)
7 (4%)
164 (15%)
7 (4%)
11 (3%)
25 (13%)
144 (13%)
25 (13%
44 (12%
19
ACCEPTED MANUSCRIPT
)
)
159 (15%)
30 (16% )
55 (14% )
162 (15%)
35 (18% )
66 (17% )
2013 30 (15%)
36 (19%)
2015 37 (19%)
150 (14%)
53 (27%)
137 (13%)
1 (1%)
9 (1%)
Discharge mortality
145,413 (IQR: 98,486-203,860)
75 (20% )
53 (28% )
114 (30% )
0.64
1 (1%)
3 (1%)
0.73
148,020 (IQR: 111,594-192,647)
0.99
147,0 75 (IQR: 98,20 5205,2 02)
159,7 15 (IQR: 117,6 21201,8 79)
0.24
53,00 7 (IQR: 38,82 764,77 0)
52,20 8 (IQR: 43,44 667,09 1)
0.20
M
Adjusted billed charge (2016US$)
35 (18% )
AN US
2016
CR IP T
2014
52,935 (IQR: 39,356-64,637)
52,201 (IQR: 42,627-66,712)
0.32
CE
PT
ED
Estimated cost (2016US$)
AC
Length of stay
Intensive care unit
Intensive care unit length of stay
Mechanical ventilation
1 (IQR: 1-2)
4 (IQR: 4-6)
<0.0 01
1 (IQR: 1-2)
5 (IQR: 4-6)
<0.0 01
93 (48%)
934 (87%)
<0.0 01
92 (48% )
342 (90% )
<0.0 01
0 (IQR: 0-1)
2 (IQR: 1-3)
<0.0 01
0 (IQR: 0-1)
2 (IQR: 1-3)
0.00 2
21 (11%)
589 (55%)
<0.0
21 (11%
57
<0.0
20
ACCEPTED MANUSCRIPT
01
)
(218
01
0 (IQR: 0-0)
1 (IQR: 0-1)
<0.0 01
0 (IQR: 0-0)
1 (IQR: 0-1)
<0.0 01
Extracorporeal membrane oxygenation
0 (0%)
21 (2%)
0.04 9
0 (0%)
6 (2%)
0.08
Acute kidney failure
3 (2%)
26 (2%)
0.45
3 (2%)
12 (3%)
0.28
Surgical complication (any)
94 (48%)
565 (53%)
0.28
93 (49% )
193 (51% )
0.68
81 (42% )
111 (29% )
0.00 2
8 (4%)
Hemorrhage/ hematoma
9 (5%)
M
Respiratory
2 (1%)
ED
Infection/ fever
7 (4%)
Other
352 (33%)
0.01
AN US
82 (42%)
Cardiac
CR IP T
Mechanical ventilation days
262 (24%)
<0.0 01
8 (4%)
89 (23% )
<0.0 01
75 (7%)
0.22
9 (5%)
24 (6%)
0.45
44 (4%)
0.04
2 (1%)
19 (5%)
0.03
0.01
7 (4%)
44 (12% )
0.00 3
105 (10%)
†
CE
PT
IQR = interquartile range (1st quartile - 3rd quartile), 2016US$ = United States dollars adjusted to inflation to year 2016. Summary statistics were expressed as median (IQR) for continuous variables and percentage (count) for categorical variables. P-values were obtained from comparisons between TC-PVR and S-PVR using Wilcoxon rank-sum test for continuous variables and chi-square test for categorical variables. ‡
AC
P-values were obtained from comparisons between matched TC-PVR and SPVR using generalized linear mixed model.
Table 3: Projected 5-Year Cost of Pulmonary Valve Replacement
21
ACCEPTED MANUSCRIPT
Surgical
Initial total hospital cost
$52,935
$52,201
Surgical redo cost (rate)
$4,176 (8%)
$3,758 (7.2%)
$769 (4%)
$577 (3%)
$6,882 (13%)
$0 (0%)
Valvuloplasty/angioplasty cost (rate) Trans-catheter pulmonary valve re-implantation cost (rate)
CR IP T
Trans-catheter
$1,241
$2,987
Projected total 5-year cost to payor (wage loss not included)
$64,762
$56,536
Projected total 5-year cost to society (wage loss included)
$66,003
$59,523
AC
CE
PT
ED
M
AN US
Total Wage Loss
22