Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Congenital heart defect repair with ADAPT tissue engineered pericardium scaffold: An early-stage health economic model

  • Vladica M. Veličković ,

    Contributed equally to this work with: Vladica M. Veličković, Oleg Borisenko

    Roles Conceptualization, Data curation, Formal analysis, Methodology, Software, Validation, Visualization, Writing – review & editing

    vladica.velickovic@edu.umit.at

    Affiliations Synergus AB, Health Economics and Evidence Synthesis Department, Stockholm, Sweden, Institute of Public Health, Medical Decision Making and Health Technology Assessment, Department of Public Health, Health Services Reseaech and Health Technology Assessment, UMIT - University for Health Sciences, Medical Informatics and Technology, Hall i.T., Austria

  • Oleg Borisenko ,

    Contributed equally to this work with: Vladica M. Veličković, Oleg Borisenko

    Roles Conceptualization, Methodology, Project administration, Supervision, Validation, Writing – review & editing

    Affiliation Synergus AB, Health Economics and Evidence Synthesis Department, Stockholm, Sweden

  • Mikael Svensson ,

    Roles Conceptualization, Methodology, Supervision, Validation, Writing – review & editing

    ‡ These authors also contributed equally to this work.

    Affiliation Health Metrics, The Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden

  • Tim Spelman ,

    Roles Validation, Writing – original draft, Writing – review & editing

    ‡ These authors also contributed equally to this work.

    Affiliations Synergus AB, Health Economics and Evidence Synthesis Department, Stockholm, Sweden, Centre for Population Health, Burnet Institute, Melbourne, Australia, School of Public Health and Preventive Medicine, Monash University, Melbourne, Australia

  • Uwe Siebert

    Roles Methodology, Validation, Writing – review & editing

    ‡ These authors also contributed equally to this work.

    Affiliation Institute of Public Health, Medical Decision Making and Health Technology Assessment, Department of Public Health, Health Services Reseaech and Health Technology Assessment, UMIT - University for Health Sciences, Medical Informatics and Technology, Hall i.T., Austria

Congenital heart defect repair with ADAPT tissue engineered pericardium scaffold: An early-stage health economic model

  • Vladica M. Veličković, 
  • Oleg Borisenko, 
  • Mikael Svensson, 
  • Tim Spelman, 
  • Uwe Siebert
PLOS
x

Abstract

Objective

The objective of this study was to evaluate the cost effectiveness of tissue engineered bovine tissue pericardium scaffold (CardioCel) for the repair of congenital heart defects in comparison with surgery using xenogeneic, autologous, and synthetic patches over a 40-year time horizon from the perspective of the UK National Health Service.

Methods

A six-state Markov state-transition model to model natural history of disease and difference in the interventional effect of surgeries depending on patch type implanted. Patches differed regarding their probability of re-operation due to patch calcification, based on a systematic literature review. Transition probabilities were based on the published literature, other clinical inputs were based on UK registry data, and cost data were based on UK sources and the published literature. Incremental cost-effectiveness ratio (ICER) was determined as incremental costs per quality adjusted life years (QALY) gained. We used a 40-year analytic time-horizon and adopted the payer perspective. Comprehensive sensitivity analyses were performed.

Results

According to the model predictions, CardioCel was associated with reduced incidence of re-operation, increased QALY, and costs savings compared to all other patches. Cost savings were greatest compared to synthetic patches. Estimated cost savings associated with CardioCel were greatest within atrioventricular septal defect repair and lowest for ventricular septal defect repair. Based on our model, CardioCel relative risk for re-operations is 0.938, 0.956and 0.902 relative to xenogeneic, autologous, and synthetic patches, respectively.

Conclusion

CardioCel was estimated to increase health benefits and save cost when used during surgery for congenital heart defects instead of other patches.

Introduction

Congenital heart defects (CHD) are associated with considerable morbidity and mortality globally, with an incidence ranging from 8 to 13 per 1000 live births [13]. The ongoing need for repeat and revision surgery coupled with a lifelong burden of associated disease and comorbidity translates into significant public health impact and healthcare costs, which characteristically extend well beyond childhood. Of the $2.6 billion USD in hospital costs associated with birth defects in the United States during 2004, $1.4 billion (54%) was directly associated with the management and treatment of structural cardiovascular defects [2]. The costs of identifying critical congenital heart defects in neonates have previously been reported at just over $20,000 USD per newborn detected [4, 5] whilst median palliation costs have been reported at up to $99,000 USD [6].

A diverse range of cardiovascular patches for the correction of congenital defects in neonates and pediatric patients have been tested in clinical trials including the use of synthetic, autologous, and biological material [712]. Of the biologics, patches derived from bovine pericardium have previously demonstrated to be effective for the closure of several common anomalies [7, 8, 10]. Whilst the complication rate associated with such procedures is relatively moderate [13], calcification of patches with subsequent dehiscence and failure is a known risk of biologics [14], in addition to infection, inflammation, and bleeding [15].

The ADAPT tissue-engineered bovine pericardium bio-scaffold (CardioCel) is distinct from comparable patches by the removal of calcium-binding phospholipids and antigens during manufacture, both of which are known to promote the calcification observed with other patches [13, 16]. Recent data from the extension study of the Phase II trial of CardioCel in CHD repair reported no calcification or repeat surgery across 14 patients followed-up for between five and eight years [unpublished data, [17]]. A single-centre, prospective cohort study of 30 pediatric patients receiving the CardioCel scaffold observed no patch-associated morbidity within one month of insertion and no echocardiographic evidence of calcification, bleeding, or failure at 18 to 36 months’ post-surgery [13]. A recent retrospective German review of 37 pediatric and adult patients similarly reported no patch-related complication or signs of tissue failure [18]. In addition, Prabhu et al. [19] in histologic evaluation of explanted CardioCel in CHD population confirms that in top of the fact that calcification did not occur, CardioCel demonstrate evidence of remodelling and neointima formation.

Whilst the early data suggests a potential advantage favoring CardioCel in terms of calcification, remodeling, and thrombus formation relative to synthetic patches, and lack of calcification, surface thickening, and structural leak relative to biological patches [15], the potential cost-effectiveness of CardioCel has not yet been investigated. The objective of this study was to evaluate the potential cost-effectiveness of CardioCel repair of congenital cardiac defects in comparison with xenogeneic, autologous, and synthetic patches.

Materials & methods

Model description

A Markov state-transition cohort model was developed and used to estimate the cost-effectiveness of CardioCel versus other patches (Fig 1). Patients enter the model at the time of the index patch repair surgery, after which they may stay alive with no re-operation, undergo repeat operation due to calcification of patch or other causes, or die.

The premise of the analysis is that due to better tissue properties, CardioCel can reduce the frequency of reoperations caused by calcification and correspondingly can increase survival and quality of life of patients with a lower cost to the health care system.

The ideal solution which can provide a full insight into outcomes after patch implantation will be to follow patients from index surgery through the complete lifetime horizon with all relevant events properly analyzed. However, in real world settings this is not feasible both for organisational and financial reasons. On the other hand, the simulation methods framework allows for simulating the lives of virtual patients from the index event until death taking into account all relavant events and synthesize all relevant available evidence (prospective and retrospective primary data, published secondary data sources, and real-world data sources from registries). One of the most frequently used methods for this purpose is the Markov state-transition model, and in line with methodological recommendations we use this method in order to adequately address the research question [20, 21]. We also followed the international ISPOR-SMDM guidelines on decision-analytic modelling [22]. The evaluation is reported in line with “Consolidated Health Economic Evaluation Reporting Standards” (CHEERS) [23].

Clinical inputs

The target population includes a mix of pediatric patients with six congenital cardiovascular anomalies, treated with the most common procedures involving the use of patches (Table 1). Distribution of surgical methods was informed by the 2014–15 Admitted Care Hospital Episode Statistics (HES), which collects data from all hospital admissions in England [24].

thumbnail
Table 1. Congenital heart defects and most common surgical procedures involving use of patches.

https://doi.org/10.1371/journal.pone.0204643.t001

Three groups of comparator patches were used: xenogeneic, autologous, and synthetic patches. Synthetic patches included expanded polytetrafluoroethylene (ePTFE) and fabric patches, whilst both decellularized xenogeneic and tissue-engineered patches were considered in the xenogeneic patch group. Utilization of each patch type for different surgeries was based on the 15-year analysis of HES [24].

The three-tier approach was taken in relation to the estimation of the frequency of reoperations. Firstly, disease-specific freedom from re-operation was informed from Monro et al. [25] and Sakurai et al [26]. Secondly, the disease-specific and patch-related etiological fraction of reoperations was estimated via the single source (tetralogy of Fallot (ToF) and coarctation of the aorta (CoA)) or meta-analysis of large, international studies reporting the cause of reoperation (atrioventricular septal defect (AVSD), ventricular septal defect (VSD), and transposition of great arteries (TGA)). Thirdly, within a fraction of patch-related reoperations, the fraction of reoperations due to patch calcification was estimated from explanted patch histological studies. Only patch calcification-related reoperations differ between compared patches. In the case of aortic stenosis, the assumption was made that 100% of reoperation were patch-related due to the specific position and active role of the patch. The baseline value of risk of re-operations in the absence of calcifications were estimated by adjusting these patch-related reoperations

Survival analysis involved calculating of the age-, gender- and disease-specific survival, and additional adjustment for procedure-specific 30-day mortality for both index and redo surgeries. The UK life tables [27] were utilised as a starting point with several subsequent adjustments in order to inform survival in the model. Conceptual steps of that adjustment are presented in Table 2.

The first relative risk between mortality of the normal population and patients with CHD was informed from a population-based study from Finland [28]. This study is selected as the best possible source for several reasons: (i) this is the one of the largest CHD patient population-based studies internationally; (ii) reporting survival starting from the index surgery for four surgical decades (this is particularly important due to the significant differences in survival in recent decades); (iii) reported follow-up survival time is from 20 years for the more recent surgical decade up to 45 for oldest one (this give a unique opportunity to use older surgical decades as an external validation source when extrapolating recent surgical decades from 20 years of observed follow-up to 40 years after index surgery). The data from Fig 2 from Nieminen et al. [28] were extracted in order to estimate the relative risk between the general population and CHD disease mortality. In the first step, this relative risk was applied to the UK life tables and average survival for the UK population with CHD was estimated. Additionally, by extracting data from the survival curves in Fig 3 from Nieminen et al [28] the disease specific relative risk was determined as the ratio between average CHD mortality and specific disease mortality. However, due to the limitation of observed survival in the most recent surgical decade of 20 years, after extraction from the original source, data are first extrapolated up to 40 years after index surgery and then the relative ratio was estimated. Exceptions are aortic stenosis and atrioventricular septal defect because they are not reported in Nieminen et al [28] and therefore an alternative source is used for survival estimation. Aortic stenosis survival was informed by Alexiou et al. [29] as a most appropriate available source from the UK reporting survival 25 years after index surgery. In the case of atrioventricular septal defect currently, there is a lack of studies with sufficient follow-up coming from the UK. Therefore, as a most appropriate source, but still relevant for the UK, Ginde et al [30] reporting survival of patients with atrioventricular septal defect after index surgery in the US was selected. Details regarding extrapolation of survival data are reported in the Figs A—I and Tables A—E in S3 File).

thumbnail
Fig 2. Observed and extrapolated survival.

CHD—congenital heart defects, AS—aortic valve stenosis, VSD—ventricular septal defect, AVSD—atrioventricular septal defect, ToF—tetralogy of Fallot, TGA—transposition of great arteries, CoA—coarctation of the aorta.

https://doi.org/10.1371/journal.pone.0204643.g002

thumbnail
Fig 3. Observed and extrapolated freedom from reoperation data inputs.

CHD—congenital heart defects, AS—aortic valve stenosis, VSD—ventricular septal defect, AVSD—atrioventricular septal defect, ToF—tetralogy of Fallot, TGA—transposition of great arteries, CoA—coarctation of the aorta.

https://doi.org/10.1371/journal.pone.0204643.g003

The approach in the selection of an extrapolation model was done systematically, similar to the case of reoperation extrapolations and in line with current NICE recommendations [31]. Furthermore, survival in the model was additionally adjusted for procedure-specific operation mortality (30-day mortality) informed by the UK Central Cardiac Audit Database [32] as presented in Table 3.

Re-operation and survival trends (Table 4) were then extrapolated out to a 40-year time horizon using a variety of parametric and non-parametric functions (Figs 2 and 3). Model fit was assessed through comparison of Akaike (AIC) and Bayesian (BIC) Information Criterion, coefficients of determination, visual inspection of residuals, and comparison against external benchmarks [31].

Utilities

The utilities employed in the analysis were calculated using the approach described by Mistry et al [34] and Knowles [35]. Quality-adjusted life-years (QALYs) were estimated in two steps. In the first step, patients with different CHD types were classified into four different heart failure categories as presented in Table 5.

In the second step, age-dependent utility values were assigned to the corresponding group of heart failure patients (Table 6.)

thumbnail
Table 6. Proxy utility values attached to different age group in patients with CHD diseases.

https://doi.org/10.1371/journal.pone.0204643.t006

A short-term (one month) disutility of surgical intervention was also used. Firstly, the base value of 50% reduction in utility was calculated, based on Orlando et al [40]. Next, this base value was adjusted using relative severity of surgery, based on the Aristotle complexity score (ACS) [41] (Table 7). The base value of 50% reduction in quality of life was corresponding to the aortic valve replacement with ACS of 8.5.

thumbnail
Table 7. Basic Aristotle complexity score and short-term (monthly) disutility’s.

https://doi.org/10.1371/journal.pone.0204643.t007

Cost inputs

The cost of index surgery, reoperations, follow-up management (hospital stay, outpatient visits, tests, examinations, and drugs) was included in the analysis.

The UK expert inputs for resource use by defined CHD class were based on Mistry et al [34] and are presented in Table 8.

The cost of index surgery and reoperations was based on weighted average of Health Care Resource Groups (HRGs) for pediatric cardiac surgery procedures (Table 6). Analysis was performed using “HRG4+ 2015/16 Reference Costs Grouper” [42] in order to distinguish between different procedures used for index and redo surgeries. Unit cost of follow-up management is also presented in Table 9.

thumbnail
Table 9. Unit costs for resource use and cost of index surgery and reoperations.

https://doi.org/10.1371/journal.pone.0204643.t009

Uncertainty analysis

Uncertainty about data inputs was evaluated in deterministic and probabilistic sensitivity analysis per every CHD disease included into the analysis. Deterministic sensitivity analysis included varying each model parameter individually while holding other variables fixed at base-case values. Results are presented as Tornado diagrams. Monte Carlo probabilistic sensitivity analysis with 10,000 iterations were performed, where a random value for each variable is drawn based on a pre-determined distribution. The beta distribution was used for effectiveness and utility parameters; gamma distribution was used for cost and disutility parameters, normal distribution was used for age, and log-normal distribution was used for risk ratios.

Analysis

In line with NICE recommendations, all costs and outcomes beyond the first year were discounted at 3.5% annually [46, 47]. Cardiac patch was considered cost-effective if the incremental cost-effectiveness ratio (ICER, which is calculated by dividing difference in cost between two arms by difference in QALY) was below the lower-bound willingness-to-pay threshold of £20,000/QALY [47]. Cardiac patch was considered a dominant strategy when associated expected effectiveness were greater and costs were smaller than the expected effectiveness and costs (also referred to as strong dominance or simple dominance).

Statistics analysis was performed using Stata 14 (StataCorp. 2015. Stata Statistical Software: Release 14. College Station, TX). The model was developed in Microsoft Excel 2016 (Microsoft Corp., Redmond, Washington, DC, USA).

Results

Base case analysis

According to model predictions, CardioCel was associated with lower incidence of re-operation, higher QALY and life years (LY), and lower overall costs when compared to any of the comparator patches (Table 10). CardioCel was dominating all three comparative patches.

As patches are mutually exclusive treatments options, and therefore the base-case results are additionally reported (Tables A—L in S4 File) by calculating the incremental costs, effects, and ICER relative to the next most effective patch.

A similar pattern was observed for the model estimated results disaggregated by indication. Model estimations confirms that CardioCel again dominated all three comparator patch groups with cost savings maximized for AS across all three-patch type, whilst the smallest savings were estimated for VSD repair (Tables A—L in S4 File). Across all three comparator patch groups, cost savings were as follows (in order of decreasing savings): AS, AVSD, TGA, CoA, ToF, and VSD.

The model projections in relation to re-operation relative risk reduction was greatest in the AVSD with CardioCel 15%, 11%, and 23% for xenogeneic, autologous, and synthetic patches respectively. By comparison, the model prediction in relation to re-operation relative risk reduction associated with the use of CardioCel in AS repair was comparatively small with 1.6%, 1,2%, and 2.6% reductions estimated for xenogeneic, autologous, and synthetic patches respectively.

Sensitivity analyses

Deterministic sensitivity analysis showed that the model most sensitive parameters are short-term operative mortality, the cost of reoperations, and utility values for CHD disability. CardioCel was cost saving according to model estimates and dominated all comparators within all investigated diseases.

Probabilistic sensitivity analysis demonstrated that CardioCel produced clinical benefits (additional QALYs) and had a cost-saving effect irrespective to an indication or chosen comparator among all comparators. Detailed disease-specific results together with breakdowns by reoperation aetiology and cost of surgeries versus the cost of follow-up fractions are presented in the Tables A—L and Figs A—AJ in S4 File).

Discussion

Our model based economic evaluation informed by preliminary clinical outcome results suggests that the ADAPT tissue engineered bovine tissue pericardium scaffold (CardioCel), according to the simulation results, consistently dominates any of the xenogeneic, autologous, and synthetic comparator patches both in terms of being cost-saving and associated with a QALY gain across a 40-year time horizon.

This was observed across all indications in our model, although the magnitude of the advantage varied with the underlying anomaly. Predicted savings and QALYs gained were highest when CardioCel was used in the repair of congenital AVSD, relative to all three of the comparator patch groups. In contrast, CardioCel for ventricular septal defect and aortic stenosis repair was associated with the smallest predicted cost savings. Whilst this pattern was consistent across all three patch types, savings and QALYs were both highest when CardioCel was compared against synthetic patches, followed by xenogeneic, and then autologous patches according to the model predictions.

As short and long-term health outcomes associated with congenital cardiovascular defects improve and the associated survival gain leads to a population shift towards more adult patients [48, 49], so too is the potential gain in cost savings associated with avoiding reoperation secondary to patch failure or breakdown. In real terms, the primary model estimated cost savings associated with CardioCel for congenital cardiovascular anomaly repair are modest, with maximum savings per patient of £ 372 GBP over 40 years (CardioCel vs. synthetic patches in AVSD). However, the estimated relative reduction in re-operation risk was as high as 23% (CardioCel vs. synthetic patches, indication = AVSD) suggesting clinically significant gains in sub-groups of CHD patients, particularly those diagnosed with AVSD.

Limitations

There are several limitations to this model-based economic evaluation. The presented analysis was informed by a variety of randomized and non-randomised data sources including national, administrative datasets and expert opinions in addition to trial data. Several of the data used in the primary analysis are limited by moderate to high risk of bias, secondary to retrospective designs and small sample size. Furthermore, estimation of the change of re-operation incidence due to the calcifications and the identification of factors associated with reoperation assumed that the overall risk of reoperation is the simple sum of its individual factors, which does not account for any correlation or interaction between subsets of these factors. Whilst the causes of re-operation are many and complex, there are limited data available that quantify the relationships between these explanatory variables and subsequent re-operation risk.

Further on, simulation of cohorts with state-transition models, can potentially exclude some important correlates of health and cost outcomes in a CHD population. In general, the ideal method is to randomize patients to appropriate comparative arms and follow all outcomes over time. However, comparing several different patches with 40 years of follow-up is not practical and feasible due to patch technology advancement over time and associated study costs. Therefore, despite all limitations, the decision-analytic simulation approach is the best reasonable alternative and similar conclusions have been reached in several other types of outcomes research in this field [5052].

The focus of the present evaluation was upon the calcification patch properties only and their risk of subsequent reoperations. However, calcifications may also be associated with arrhythmias [53] and heart failure [54], each associated with their own unique suite of costs. The published data supporting these associations, however, is currently limited. Excluding the additional costs saved from avoiding these other conditions may underestimate the true benefit of CardioCel. Beyond reducing calcification, CardioCel has also been associated with the absence of surface thickening, thrombus formation, structural leak, and residual leak [13]. However, those properties and their effects on health and cost outcomes were not taken into consideration in these analyses, biasing our results against CardioCel.

The histological studies used for the adjustment of the patch related reoperations due to the calcifications are based on small sample sizes. The decision to use histological studies instead clinical studies, despite small sample size, was made after a detailed systematic review of the literature. Namely, in order to distinguish between different patch types, we conduct a systematic literature search in MEDLINE via PubMed, Cochrane, and EMBASE for the patch specific calcifications rates in CHD patients. The following keywords were used: “prostheses and implants", "transplants”,” biocompatible materials”, “patch”, “scaffold”, “cord”, *graft”, “septal occluder”, “bovine”, “porcine”, “equine”, “synthetic”, “biological”, “tissue”, “tissue-engineered”, “pericardium”, “homograft”, “homologous”, “autologous”, “xenograft” “heterograft”, “allograft”, “autograft”, “isograft”, "glutaraldehyde crosslinking", "glutaraldehyde fixation”. In total, 2026 articles were identified after deduplication. All studies identified by the search can be divided into two groups depending on the method used for assessing patch calcifications: (i) studies reporting echocardiography or other types of radiology assessment of implanted patches; and (ii) studies reported a histological examination of explanted patches.

In the early phase of our research, the focus was on clinical studies, which reported echocardiography or other types of radiology assessment of implanted patches. In total, 49 studies were identified. Only few studies were eligible for full-text review. The primary reason for exclusion of studies was that they reported calcification rates for a prosthesis, conduits, and valved conduits and not for patch or scaffolds. However, detailed analyses of the studies showed that there was no possibility to differentiate between patch type and rates of calcification, mostly due to differences in reporting and an infeasibility to distinguish between clinically consequential and insignificant calcifications. There is a clear need for the standardization in assessment and categorization of calcification severity, and currently evidence synthesis of the patch specific calcification rates according to their severity is not feasible from clinical studies.

However, the second approach (using histological studies) allows for more accurate quantification of the causal link between calcification and reoperation due to the reason that analyses were done on the explanted patches. Therefore, although the sample size of histological studies is much smaller, the relation between calcification and reoperation more accurately characterized. In such trade-off situation, we decide to use results of histological studies in the model, as it provides more reliable information regarding the clinical significance of calcifications in relation to reoperations. Unfortunately, other solutions cannot be expected in near to mid future. An international effort will be needed with linkage of several registries as well as their harmonization and standardization of the measurements.

In addition, modelled clinical pathway in the case of the isolated congenital aortic stenosis is the open surgical valvotomy in which in addition patch is used for the valve reconstruction. Subsequent reoperations are assumed to be exclusively repairs (reconstructions) with the same patch types (not replacements with prothesis). Although we are aware of a small number of infants in the projected clinical pathway in our model, we are forced to model in this way due to the need to isolate the patch effect which is not modified by other non-patch surgical approaches, both at the level of index surgery, but for the reoperations as well. This is the trade-off decision we must make to have a fair comparison of patches. Effect of switching to another patch or other surgical approaches in case of the reoperations was not possible to statistically control due to the lack of sufficient primary data sources. Therefore, the infrequent rather than standard of care pathway is modelled. To stay in line with current clinical practise and to limit potential bias in final results, AS indication was modelled using adequately low prevalence inputs in the overall CHD cohort. In addition, conservative assumption in relation to patch-related re-operations (16%) in this indication were taken into account to future minimise the impact of AS sub-cohort to final results. However, by developing new patch technologies as CardioCel, it is highly possible that clinical practise will shift toward using AS repairs not only as a temporary solution toward surgical replacement but rather as a long-term solution in both pediatric and adult patients.

The current evidence level in relation to differences in short and middle term outcomes among different patch classes lack direct head to head comparison. Future comparative studies are needed for more robust quantification of the patch calcification impact on re-operations, arrhythmias, and heart failure. Additional research is needed to shed light on the underlying causal relations of other important patch properties, as surface thickening, thrombus formation, structural leak, and residual leak, with mid and long-term outcomes. However, having in mind the differences in outcomes among different CHD diseases, which dictated the need for disease specific comparison and consequently small target populations, the realistic and feasible approach should be retrospective comparison, ideally based on registry data.

Conclusion

A Markov state-transition model was developed to evaluate the cost effectiveness of the surgical repair of congenital heart diseases with bovine tissue engineered scaffold CardioCel compared with patches of xenogeneic, autologous, and synthetic origin. Model results demonstrated that CardioCel reduces reoperations, prolongs life, improves quality of life and saves costs compared xenogeneic, autologous, and synthetic patches. However, future research is needed to better quantify crucial causal relations links between patch properties and core outcomes.

Supporting information

S1 File. Reoperation etiological fractions.

https://doi.org/10.1371/journal.pone.0204643.s001

(PDF)

S2 File. Freedom from reoperations inputs.

https://doi.org/10.1371/journal.pone.0204643.s002

(PDF)

S5 File. Variables distributions and parameters used for the uncertainty analysis.

https://doi.org/10.1371/journal.pone.0204643.s005

(PDF)

Acknowledgments

We would like to express our sincere gratitude to Anastasia Pochuprina for the digitizing of the data from plots and graphs from published literature, and to Emily McIlwaine for proofreading the final version of the article.

References

  1. 1. Hoffman J. The global burden of congenital heart disease. Cardiovasc J Afr. 2013;24(4):141–5. pmid:24217047.
  2. 2. Marelli AJ, Ionescu-Ittu R, Mackie AS, Guo L, Dendukuri N, Kaouache M. Lifetime prevalence of congenital heart disease in the general population from 2000 to 2010. Circulation. 2014;130(9):749–56. pmid:24944314.
  3. 3. van der Linde D, Konings EE, Slager MA, Witsenburg M, Helbing WA, Takkenberg JJ, et al. Birth prevalence of congenital heart disease worldwide: a systematic review and meta-analysis. J Am Coll Cardiol. 2011;58(21):2241–7. pmid:22078432.
  4. 4. Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, et al. Heart disease and stroke statistics—2015 update: a report from the American Heart Association. Circulation. 2015;131(4):e29–322. pmid:25520374.
  5. 5. Peterson C, Grosse SD, Oster ME, Olney RS, Cassell CH. Cost-effectiveness of routine screening for critical congenital heart disease in US newborns. Pediatrics. 2013;132(3):e595–603. pmid:23918890.
  6. 6. Dean PN, Hillman DG, McHugh KE, Gutgesell HP. Inpatient costs and charges for surgical treatment of hypoplastic left heart syndrome. Pediatrics. 2011;128(5):e1181–6. pmid:21987703.
  7. 7. Lakshmanan R, Krishnan UM, Sethuraman S. Living cardiac patch: the elixir for cardiac regeneration. Expert Opin Biol Ther. 2012;12(12):1623–40. pmid:22954059.
  8. 8. Mirensky TL, Breuer CK. The development of tissue-engineered grafts for reconstructive cardiothoracic surgical applications. Pediatr Res. 2008;63(5):559–68. pmid:18427302.
  9. 9. Odim J, Laks H, Allada V, Child J, Wilson S, Gjertson D. Results of aortic valve-sparing and restoration with autologous pericardial leaflet extensions in congenital heart disease. The Annals of thoracic surgery. 2005;80(2):647–53; discussion 53–4. pmid:16039221.
  10. 10. Pok S, Jacot JG. Biomaterials advances in patches for congenital heart defect repair. J Cardiovasc Transl Res. 2011;4(5):646–54. pmid:21647794.
  11. 11. Schoof PH, Hazekamp MG, van Ulzen K, Bartelings MM, Bruyn JA, Helbing W, et al. Autologous pericardium for ventricular septal defect closure. J Heart Valve Dis. 1998;7(4):407–9. pmid:9697062.
  12. 12. Us MH, Sungun M, Sanioglu S, Pocan S, Cebeci BS, Ogus T, et al. A retrospective comparison of bovine pericardium and polytetrafluoroethylene patch for closure of ventricular septal defects. J Int Med Res. 2004;32(2):218–21. pmid:15080027.
  13. 13. Neethling WM, Strange G, Firth L, Smit FE. Evaluation of a tissue-engineered bovine pericardial patch in paediatric patients with congenital cardiac anomalies: initial experience with the ADAPT-treated CardioCel(R) patch. Interact Cardiovasc Thorac Surg. 2013;17(4):698–702. pmid:23832918.
  14. 14. Walther T, Tsang VT, Deanfield JE, de Leval MR. Closure of recurrent VSD due to dehiscence of calcified patch. Eur J Cardiothorac Surg. 2003;23(2):246–7. pmid:12559356.
  15. 15. Strange G, Brizard C, Karl TR, Neethling L. An evaluation of Admedus' tissue engineering process-treated (ADAPT) bovine pericardium patch (CardioCel) for the repair of cardiac and vascular defects. Expert Rev Med Devices. 2015;12(2):135–41. pmid:25431988.
  16. 16. Sobieraj M, Cudak E, Mrowczynski W, Nalecz TK, Westerski P, Wojtalik M. Application of the CardioCel bovine pericardial patch—a preliminary report. Kardiochir Torakochirurgia Pol. 2016;13(3):210–2. pmid:27785133.
  17. 17. LONG-TERM CARDIOCEL STUDY SHOWS NO EVIDENCE OF CALCIFICATION AFTER 8 YEARS. 2016.
  18. 18. Nordmeyer S, Cho MY, Nordmeyer J, Ovroutski S, Miera O, Musci M, et al. First Experience of a Novel Decellularized Patch Material (CardioCel) for Different Applications in Congenital Heart Surgery. Thorac cardiovasc Surg. 2015;63(S 01):OP158.
  19. 19. Prabhu S, Armes JE, Bell D, Justo R, Venugopal P, Karl T, et al. Histologic Evaluation of Explanted Tissue-Engineered Bovine Pericardium (CardioCel). Seminars in Thoracic and Cardiovascular Surgery. 2017. https://doi.org/10.1053/j.semtcvs.2017.05.017.
  20. 20. Siebert U. When Should Decision-Analytic Modeling Be Used in the Economic Evaluation of Health Care? The European Journal of Health Economics. 2003;4(3):143–50.
  21. 21. Siebert U, Alagoz O, Bayoumi AM, Jahn B, Owens DK, Cohen DJ, et al. State-Transition Modeling. Medical Decision Making. 2012;32(5):690–700. pmid:22990084
  22. 22. Caro JJ, Briggs AH, Siebert U, Kuntz KM. Modeling good research practices—overview: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force-1. Med Decis Making. 2012;32(5):667–77. Epub 2012/09/20. pmid:22990082.
  23. 23. Husereau D, Drummond M, Petrou S, Carswell C, Moher D, Greenberg D, et al. Consolidated Health Economic Evaluation Reporting Standards (CHEERS) statement. BMJ: British Medical Journal. 2013;346.
  24. 24. Hospital Episode Statistics. Centre H& SC information; 2015.
  25. 25. Monro JL, Alexiou C, Salmon AP, Keeton BR. Reoperations and survival after primary repair of congenital heart defects in children. J Thorac Cardiovasc Surg. 2003;126(2):511–20. pmid:12928652.
  26. 26. Sakurai T, Stickley J, Stumper O, Khan N, Jones TJ, Barron DJ, et al. Repair of isolated aortic coarctation over two decades: impact of surgical approach and associated arch hypoplasia. Interact Cardiovasc Thorac Surg. 2012;15(5):865–70. pmid:22833510.
  27. 27. National life tables: United Kingdom. In: Statistics OfN, editor. Newport, South Wales 2017.
  28. 28. Nieminen HP, Jokinen EV, Sairanen HI. Causes of late deaths after pediatric cardiac surgery: a population-based study. J Am Coll Cardiol. 2007;50(13):1263–71. pmid:17888844.
  29. 29. Alexiou C, Chen Q, Langley SM, Salmon AP, Keeton BR, Haw MP, et al. Is there still a place for open surgical valvotomy in the management of aortic stenosis in children? The view from Southampton. European Journal of Cardio-Thoracic Surgery. 2001;20(2):239–46. pmid:11463538
  30. 30. Ginde S, Lam J, Hill GD, Cohen S, Woods RK, Mitchell ME, et al. Long-term outcomes after surgical repair of complete atrioventricular septal defect. The Journal of thoracic and cardiovascular surgery. 2015;150(2):369–74. pmid:26048271.
  31. 31. Latimer N. NICE DSU TECHNICAL SUPPORT DOCUMENT 14: SURVIVAL ANALYSIS FOR ECONOMIC EVALUATIONS ALONGSIDE CLINICAL TRIALS—EXTRAPOLATION WITH PATIENT-LEVEL DATA. Decision Support Unit, ScHARR, University of Sheffield, 2011.
  32. 32. Pagel C, Brown KL, Crowe S, Utley M, Cunningham D, Tsang VT. A mortality risk model to adjust for case mix in UK paediatric cardiac surgery. Health Services and Delivery Research. Southampton (UK)2013.
  33. 33. Majeed A, Baird CW, Borisuk MJ, Sanders S, Padera R. Histology of Pericardial Tissue Substitutes Used in Congenital Heart Surgery. Pediatr Dev Pathol. 2015. pmid:26492092.
  34. 34. Mistry H, Gardiner HM. The Cost-Effectiveness of Prenatal Detection for Congenital Heart Disease Using Telemedicine Screening. Journal of Telemedicine and Telecare. 2013;19(4):190–6. pmid:23576807
  35. 35. Knowles R, Griebsch I, Dezateux C, Brown J, Bull C, Wren C. Newborn screening for congenital heart defects: a systematic review and cost-effectiveness analysis. Health technology assessment (Winchester, England). 2005;9(44):1–152, iii–iv. pmid:16297355.
  36. 36. Brown KL, Ridout DA, Hoskote A, Verhulst L, Ricci M, Bull C. Delayed diagnosis of congenital heart disease worsens preoperative condition and outcome of surgery in neonates. Heart (British Cardiac Society). 2006;92(9):1298–302. pmid:16449514.
  37. 37. Kirsch J, McGuire A. Establishing health state valuations for disease specific states: an example from heart disease. Health Econ. 2000;9(2):149–58. pmid:10721016.
  38. 38. Yount LE, Mahle WT. Economic analysis of palivizumab in infants with congenital heart disease. Pediatrics. 2004;114(6):1606–11. pmid:15574622.
  39. 39. Caviness AC, Cantor SB, Allen CH, Ward MA. A cost-effectiveness analysis of bacterial endocarditis prophylaxis for febrile children who have cardiac lesions and undergo urinary catheterization in the emergency department. Pediatrics. 2004;113(5):1291–6. pmid:15121944.
  40. 40. Orlando R, Pennant M, Rooney S, Khogali S, Bayliss S, Hassan A, et al. Cost-effectiveness of transcatheter aortic valve implantation (TAVI) for aortic stenosis in patients who are high risk or contraindicated for surgery: a model-based economic evaluation. Health technology assessment (Winchester, England). 2013;17(33):1–86. pmid:23948359.
  41. 41. Lacour-Gayet F, Clarke D, Jacobs J, Comas J, Daebritz S, Daenen W, et al. The Aristotle score: a complexity-adjusted method to evaluate surgical results. Eur J Cardiothorac Surg. 2004;25(6):911–24. pmid:15144988.
  42. 42. HRG4+ 2015/16 Reference Costs Grouper Health and Social Care Information Centre; 2016 [cited 2016]. http://www.hscic.gov.uk/casemix/costing.
  43. 43. Mangham LJ, Petrou S, Doyle LW, Draper ES, Marlow N. The cost of preterm birth throughout childhood in England and Wales. Pediatrics. 2009;123(2):e312–27. pmid:19171583.
  44. 44. British National Formulary 2016. http://www.bnf.org/products/books/.
  45. 45. NHS Reference Costs: National Health Service 2016. http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_062884.
  46. 46. The Green Book: Appraisal and Evaluation in Central Government HM Treasury. http://www.hm-treasury.gov.uk/d/green_book_complete.pdf.
  47. 47. Guide to the methods of technology appraisal 2013: NICE; 2013 [cited 2016]. http://publications.nice.org.uk/pmg9.
  48. 48. Roche SL, Silversides CK. Hypertension, obesity, and coronary artery disease in the survivors of congenital heart disease. Can J Cardiol. 2013;29(7):841–8. pmid:23688771.
  49. 49. Tutarel O, Kempny A, Alonso-Gonzalez R, Jabbour R, Li W, Uebing A, et al. Congenital heart disease beyond the age of 60: emergence of a new population with high resource utilization, high morbidity, and high mortality. Eur Heart J. 2014;35(11):725–32. pmid:23882067.
  50. 50. Puvimanasinghe JP, Takkenberg JJ, Edwards MB, Eijkemans MJ, Steyerberg EW, Van Herwerden LA, et al. Comparison of outcomes after aortic valve replacement with a mechanical valve or a bioprosthesis using microsimulation. Heart (British Cardiac Society). 2004;90(10):1172–8. Epub 2004/09/16. pmid:15367517.
  51. 51. Puvimanasinghe JPA, Takkenberg JJM, Edwards MB, Eijkemans MJC, Steyerberg EW, van Herwerden LA, et al. Comparison of outcomes after aortic valve replacement with a mechanical valve or a bioprosthesis using microsimulation. Heart (British Cardiac Society). 2004;90(10):1172.
  52. 52. van Geldorp MWA, Jamieson WRE, Kappetein AP, Puvimanasinghe JPA, Eijkemans MJC, Grunkemeier GL, et al. Usefulness of microsimulation to translate valve performance into patient outcome: Patient prognosis after aortic valve replacement with the Carpentier–Edwards supra-annular valve. The Journal of Thoracic and Cardiovascular Surgery. 2007;134(3):702–9.e1. https://doi.org/10.1016/j.jtcvs.2007.03.051. pmid:17723821
  53. 53. Fraser CD Jr. A clinical commentary on the article "biomaterials advances in patches for congenital heart defect repair": cardiac bioengineering for congenital heart disease: time for progress. J Cardiovasc Transl Res. 2011;4(5):655–7. pmid:21845479.
  54. 54. Halpern DG, Steigner ML, Prabhu SP, Valente AM, Sanders SP. Cardiac Calcifications in Adults with Congenital Heart Defects. Congenit Heart Dis. 2015;10(5):396–402. pmid:25564755.