Inpatient hospitalizations in adults with congenital heart disease and down syndrome: A national perspective

Abstract

Background

Adults with Down syndrome (DS) have a greater likelihood of being diagnosed with congenital heart disease (CHD) and may face unique cardiovascular, respiratory, and immunologic health burdens during hospitalization. While CHD survival has improved in recent decades, national data on hospitalization outcomes of patients diagnosed with DS remain limited.

Objectives

The present study aimed to evaluate trends in hospital admissions, inpatient management, and perioperative outcomes stratified by the presence of DS among CHD patients.

Methods

This was a retrospective cohort of the 2016-2022 National Inpatient Sample. All adult (18-64 years) hospitalizations with a diagnosis of congenital heart disease, were tabulated using previously validated International Classification of Diseases diagnosis codes. Our Primary outcome of interest was in-hospital mortality. Secondary endpoints included temporal trends in admissions, utilization of various cardiac procedures, and length of stay.

Results

Of an estimated 54,410 CHD patients hospitalized, 3745 (6.9%) had diagnosis of DS. During the study period, annual CHD admissions increased (nptrend<0.05), while the proportion of those with DS remained stable (6.6% to 6.2%, 2016–2022; nptrend = 0.23). Patients with DS more commonly presented as having hypothyroidism, dementia, chronic obstructive pulmonary disease (all P < 0.05) and less frequently underwent cardiac procedures (7.6 vs 23.2%, P < 0.001). Following risk-adjustment, the presence of DS remained independently associated with greater odds of mortality among all hospitalizations (AOR 2.00, 95%CI 1.40–2.88, P < 0.001).

Conclusions

Adults with DS and CHD represent a clinically vulnerable population with higher in-hospital mortality and lower procedural utilization. These findings underscore the need for dedicated care pathways to improve outcomes during medical admissions.

Highlights

• •Adults with Down syndrome and CHD form a stable share of adult CHD admissions with distinct indications and comorbidities.
• •Cardiac surgery and catheterization are performed less often in adults with Down syndrome than in non-DS counterparts.
• •Down syndrome nearly doubled in-hospital mortality odds, driven by non-procedural admissions despite similar resource use.
• •Down syndrome is an independent risk marker in adults with CHD, supporting multidisciplinary care beyond cardiac centers.

Abbreviations

AOR

Adjusted Odds Ratio

AUROC:

Area Under the Receiver Operating Characteristic Curve

CHD

Congenital Heart Disease

CI

Confidence Interval

DS

Down Syndrome

HCUP

Healthcare Cost and Utilization Project

ICD-10

International Classification of Diseases, Tenth Revision

IQR

Interquartile Range

NIS

National Inpatient Sample

PDA

Patent Ductus Arteriosus

ToF

Tetralogy of Fallot

US:

United States

AVSD

Atrioventricular Septal Defect

VSD

Ventricular Septal Defect

STAT

Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery

1. Introduction

Down syndrome (DS) is the most common and viable human trisomy, present in approximately 16 per 10,000 live births in the United States [1]. Individuals with DS are at increased risk of associated anomalies, particularly congenital heart defects (CHD) and its lifelong sequalae [2]. Importantly, progress in the surgical and medical management of CHD over the past several decades has markedly improved survival, resulting in a growing adult population with DS [3]. Consequently, general practitioners and hospitalists may be increasingly tasked with managing these patients during non-specialized care visits [3].

Despite advances which have improved survival and reshaped care for individuals with DS, significant differences in survival persist [4]. A 2012 study by Baraona et al. found adults with DS and CHD to experience higher in-hospital mortality and less commonly undergo cardiac procedures compared to non-DS counterparts [5]. More recently, a national Swedish study by Engsner et al. reported individuals with both DS and CHD to face a 2.3-fold higher mortality risk compared to those with CHD alone, despite improved survival in the modern surgical era [6].

We aimed to delineate contemporary national trends in hospitalizations and use of cardiac procedures among adults with CHD/DS. We hypothesized DS to be independently associated with an increased risk of in-hospital mortality, as well as greater resource utilization, in the contemporary era.

2. Methods

This was a retrospective cohort of the 2016-2022 National Inpatient Sample (NIS), the largest publicly available all-payer inpatient database in the US, which provides accurate estimates for ∼97% of annual hospitalizations. The NIS is maintained as part of the Healthcare Cost and Utilization Project (HCUP), a network of databases and software developed through federal-state partnerships and sponsored by the Agency for Healthcare Research and Quality [7].

All adult (18-64 years) hospitalizations with a diagnosis of congenital heart disease, were tabulated using previously reported International Classification of Diseases Tenth Revision (ICD-10) diagnosis codes [8]. In accordance with prior work [5], and given the high risk of mortality in patients with DS ≥ 65 years of age, this cohort was not included. Patients with a concomitant diagnosis of Down syndrome comprised the DS cohort. The non-DS cohort comprised all remaining adult CHD hospitalizations without a recorded diagnosis of DS; this group was not further restricted by the presence or absence of other genetic, syndromic, or neurological diagnoses. Consistent with known definitions [2,5], CHD diagnosis included atrioventricular septal defect (AVSD), ventricular septal defect (VSD), tetralogy of Fallot (ToF), and patent ductus arteriosus (PDA). Records missing key variables such as age, sex, and primary payer, were excluded from further analysis (2.9%; Fig. 1).

Fig. 1.

Study CONSORT diagram. AVSD = Atrioventricular septal defect; VSD = Ventricular septal defect; ToF = Tetralogy of Fallot; PDA = Patent ductus arteriosus.

The HCUP data dictionary was used to define relevant patient and hospital characteristics, including age, sex, income quartile, primary payer, and hospital teaching status [9]. Reasons for admission were ascertained using primary diagnosis categories, according to the general classifications defined by the Centers for Medicare and Medicaid Services [10]. The Elixhauser Comorbidity Index was used to quantify the burden of chronic conditions [11]. Previously described ICD-10 diagnosis codes were used to identify specific comorbidities, which included arrhythmia, coronary artery disease, heart failure, pulmonary hypertension, syncope, and cyanosis/secondary polycythemia, as reported elsewhere [5,12,13]. Cardiac surgical complexity was quantified using the Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery (STAT) mortality score, incorporating procedure-specific risk estimates for ventricular septal defect repair, patent ductus arteriosus ligation, tetralogy of Fallot repair, and endocardial cushion defect repair.

The primary outcome of interest was in-hospital mortality. Secondary outcomes included the temporal trends in DS admissions among those with CHD, the prevalence of cardiac comorbidities, along with markers of resource utilization such as inpatient costs and length of stay. A subgroup analysis was conducted by stratifying admissions into procedural and non-procedural cohorts. Consistent with definitions set forth by Baraona et al., procedures encompassed cardiac operations requiring cardiopulmonary bypass, electrophysiology studies, pacemaker insertion or revision, percutaneous ablation procedures, percutaneous transluminal coronary angioplasty, and right- and/or left-sided cardiac catheterization [5,[14], [15], [16]].

3. Statistical analysis

Categorical variables are presented as percentages (%), while continuous variables are reported as medians with interquartile range (IQR). Mann-Whitney U and Pearson's χ² tests were used for bivariate comparisons of continuous and categorical variables, respectively. Cuzick's nonparametric test was used to examine the significance of temporal trends [17]. Multivariable linear and logistic regression models were constructed to adjust for patient and hospital differences and characterize the association between DS with study outcomes. Covariate selection was guided by elastic net regularization, an automated technique which minimizes overfitting and collinearity via a penalized least-squares methodology [18]. Final model covariates included age, race, sex, income, Elixhauser comorbidity index, as well as primary indication for admission. Model discrimination and goodness of fit were assessed using receiver operating characteristics, as appropriate.

Significance was considered at α = 0.05. Stata 18.1 (StataCorp, College Station, TX) was used for all statistical analyses. This study was deemed exempt from full review by the Institutional Review Board at the University of California, Los Angeles, due to the fully de-identified nature of the data.

4. Results

Of an estimated 54,410 hospitalizations involving adults with CHD, 3745 (6.9%) also had a diagnosis of DS. Over the study period, the total number of CHD admissions increased from 7150 in 2013 to 8100 in 2022 (Fig. 2). Yet, the proportion of DS cases decreased from 6.6% in 2016 to 6.2% in 2022 (nptrend = 0.23), though this trend did not reach statistical significance.

Fig. 2.

Annual trends in congenital heart disease and the proportion of patients with a concomitant diagnosis of Down syndrome. ∗nptrend <0.05.

Among patients with DS, the most common primary admission diagnoses were respiratory (20.0%), infectious (19.8%), and cardiovascular (19.5%) conditions (P < 0.001 for all) (Fig. 3). In contrast, non-DS CHD patients were most often admitted for cardiovascular issues (35.2%), followed by pregnancy-related conditions (13.4%) and congenital anomalies (8.0%) (P < 0.001 for all).

Fig. 3.

Reason for admissions, grouped by relevant ICD 10 diagnoses for Down Syndrome.

Compared to others, DS had a similar distribution of age (42 [28–52] vs 40 years [29–54], P = 0.32) and sex (female: 53.2 vs 53.9%, P = 0.69). However, DS patients were more frequently White (72.4 vs 61.3%, P < 0.001) and presented within similar income quartiles compared to non-DS (p = 0.52). Comorbidity burden, as assessed by the Elixhauser Comorbidity Index, was comparable between DS and non-DS cohorts (3 [2–4] vs. 3 [2–5], P = 0.13). Specifically, DS exhibited greater rates of hypothyroidism (38.1 vs 7.1%, P < 0.001), dementia (10.4 vs 0.4%, P < 0.001), and chronic obstructive pulmonary disease (26.0 vs 18.2%, P < 0.001), but lower rates of hypertension (15.2 vs 38.5%, P < 0.001), chronic liver disease (4.9 vs 7.4%, P < 0.001), and diabetes (8.9 vs 17.2%, P = 0.01). Furthermore, DS patients were less likely to report a history of smoking, alcohol use disorder, or substance use disorder compared to non-DS (Table 1). When examining specific CHD diagnoses between groups, DS more commonly presented with AVSD (25.5 vs 5.2%, P < 0.001), but were less frequently diagnosed with ToF (11.3 vs 16.7%, P < 0.001) and PDA (8.8 vs 17.2%, P < 0.001), while VSD rates were similar (64.5 vs 63.4%, P = 0.57). While the overall prevalence of valvular disease was similar between groups (DS 22.7% vs non-DS 23.6%, P = 0.60), the distribution of subtypes differed. Mitral valve disease was more than twice as prevalent among DS patients (8.1 vs 3.6%, P < 0.001), while aortic valve disease was less common (2.5 vs 6.0%, P < 0.001). Rates of tricuspid, pulmonary, prosthetic, and multi-valve disease were otherwise similar between groups (Supplemental Table 1) (see Table 2).

Table 1.

Demographic, clinical, and hospital characteristics during initial admission.

	Non-DS (n = 50,665)	DS (n = 3745)	P-value
Age (years [IQR])	40 [29-54]	42 [28-52]	0.27
Female (%)	53.5	51.8	0.39
Elixhauser Comorbidity Index	3 [2–5]	3 [2-4]	0.13
Race/Ethnicity			<0.001
White	61.3	71.4
Black	16.2	11.6
Hispanic	14.2	11.7
Asian/Pacific Islander	4.1	1.8
Other	4.2	3.5
Primary Type of Insurance			<0.001
Private	43.8	15.2
Medicare	18.2	58.3
Medicaid	28.9	22.7
Uninsured	5.5	1.2
Other	3.6	2.5
Income Quartile			0.52
76th – 100th	20.2	20.3
51st – 75th	24.1	27.1
26th – 50th	26.2	25.2
0–25th	29.5	27.4
Congenital Cardiac Defect
Atrioventricular canal defect	5.2	25.5	<0.001
Ventricular septal defect	63.4	64.5	0.57
Tetralogy of Fallot	16.7	11.3	<0.001
Patent ductus arteriosus	17.2	8.8	<0.001
Comorbidities
COPD	18.2	26.0	<0.001
Hypertension	38.5	15.2	<0.001
Coagulopathy	11.7	15.5	<0.001
Chronic liver disease	7.4	4.9	0.01
Chronic kidney disease	10.6	10.7	0.94
Diabetes	17.2	8.9	<0.001
Hypothyroidism	7.5	38.1	<0.001
Smoking history	20.8	0.1	<0.001
Alcohol use disorder	3.3	0.1	<0.001
Substance use disorder	6.6	0.1	<0.001
Psychiatrica	12.6	8.4	<0.001
Dementia	0.4	10.4	<0.001
Hospital Teaching Status			0.01
Metropolitan teaching	84.9	82.1
Metropolitan non-teaching	11.3	12.4
Non-metropolitan	3.8	5.5
Bed Size			<0.001
Large	62.4	55.4
Medium	23.4	26.0
Small	14.2	18.6
Hospital Region			<0.001
Northeast	18.8	24.3
Midwest	21.5	26.3
South	39.4	31.1
West	20.4	18.3

Reported as proportions unless otherwise noted. Statistical significance was set at α = 0.05.^aIQR, interquartile range.

^aDefined as psychosis or depression.

Table 2.

Cardiac comorbidities and underlying disease for those with and without Down Syndrome (DS).

	Non-DS (n = 67,205)	DS (n = 5330)	P-value
Cardiac Comorbidities
Coronary artery disease	16.3	4.9	<0.001
Congestive heart failure	34.0	32.7	0.47
Arrhythmia	33.2	20.6	<0.001
Syncope	1.4	1.7	0.41
Valvular disease	23.6	22.7	59.9
Cyanosis	3.3	9.1	<0.001
Pulmonary hypertension	17.2	34.6	<0.001

Considering cardiac comorbidities, patients with DS exhibited higher rates of pulmonary hypertension (34.6 vs 17.2%, P < 0.001) and cyanosis (9.7 vs 3.3%, P < 0.001), compared to non-DS. Conversely, coronary artery disease (4.9 vs 16.3%, P < 0.001) and arrhythmias (20.6 vs 33.2%, P < 0.001) were less prevalent among DS patients. The incidence of congestive heart failure (32.7 vs 34.0%, P = 0.47), valvular disease (22.7 vs 23.6%, P = 0.60), and syncope (1.7 vs 1.4%, P = 0.41) were comparable between groups.

Compared to others, DS patients more frequently had Medicare coverage (58.3 vs 18.2%, P < 0.001) and less often private insurance (15.2 vs 43.8%, P < 0.001), while also being less commonly treated at metropolitan teaching hospitals (82.1 vs 84.9%, P = 0.01). DS admissions were more common in the Northeast (24.3 vs 18.8%, P < 0.001) and Midwest (26.3 vs 21.5%, P < 0.001), and less frequent in the South (31.1 vs 39.4%, P < 0.001). In addition, cardiac procedures were performed less frequently among admissions with DS compared to those without (7.6 vs 23.2%, P < 0.001). Specifically, cardiac surgery was performed in 3.9% of DS admissions versus 11.7% of non-DS admissions (P < 0.001), and catheterization was also less common (0.9 vs 4.8%, P < 0.001). Rates of percutaneous coronary intervention, electrophysiology studies, ablation, and pacemaker insertion were otherwise similar between groups (Fig. 4). Among patients undergoing cardiac surgery, the mean STAT mortality category was 1.62 (±0.86) for DS admissions compared with 1.27 (±0.55) for non-DS admissions (P < 0.001).

Fig. 4.

Frequency of cardiac procedures during hospitalization of congenital heart disease, stratified by diagnostic status of down syndrome. Cardiac procedures were less commonly performed during down syndrome admissions. ∗P < 0.05.

On unadjusted analysis, in-hospital mortality was significantly higher among DS patients (6.5 vs 2.9%, P < 0.001). However, among a subgroup of those undergoing cardiac procedures, DS did not face increased risk of death (1.7% vs 3.4%, P = 0.48). In contrast, among patients who did not undergo procedures, DS admissions were associated with significantly higher mortality (6.9% vs 2.7%, P < 0.001). Moreover, the total cost of hospitalization was comparable between DS and non-DS admissions ($13,200 [7100–28,100] vs $13,700 [6800–35,000], P = 0.21), although DS experienced increased length of stay (5 [2–9] vs 4 [2–7] days, P = 0.001).

Following adequate multivariable risk adjustment for demographics (age, sex, race), year of admission, socioeconomic and healthcare access factors (income quartile, insurance payer, hospital teaching status, bed size, and hospital procedural volume), admission indication, type of cardiac procedure performed, and comorbidity burden (Elixhauser Comorbidity Index, coronary artery disease, congestive heart failure, cardiac arrhythmias, valvular disease, cyanosis, hypertension, diabetes, dementia, and smoking history (AUROC = 0.81), DS remained independently associated with increased odds of mortality (AOR 2.00, 95% CI 1.40–2.88, P < 0.001). Such risk-adjusted mortality rate persisted over the course of the study period as shown in Fig. 5. Among patients undergoing cardiac procedures, DS did not alter the odds of in-hospital mortality (AOR 0.57, 95% CI 0.07–4.45, P = 0.60). However, for non-procedural admissions, DS was associated with greater odds of in-hospital mortality (AOR 2.10, 95% CI 1.43–3.07, P < 0.001). Regarding resource utilization, DS status was not associated with increased costs (β: +$2130; 95% CI -$7520 to +$11,780) or length of stay (AOR +0.14, 95% CI -0.76 to +1.05, P = 0.71).

Fig. 5.

Marginal trends in risk adjusted odds of mortality, stratified by diagnosis of Down syndrome.

5. Discussion

This national analysis highlights a complex landscape for adults with DS and congenital heart disease, characterized by lower procedural utilization and higher in-hospital mortality compared to their non-DS counterparts. Despite an overall increase in total CHD admissions, the proportion of DS patients within this cohort remained stable from 2016 to 2022. Additionally, our findings reveal that CHD admissions for DS patients are associated with distinct clinical profiles and care patterns. Considering that general practitioners and hospitalists are increasingly responsible for the routine care of this complex group, several of our findings warrant additional discussion.

In our study from 2016 to 2022, the proportion of DS patients among CHD hospitalizations decreased, though this trend did not reach statistical significance, despite an increase in the overall volume of CHD admissions. This finding contrasts with a national study by Baraona et al. (2013), who reported a declining proportion of DS-associated CHD in years prior. Improvements in prenatal detection, childhood medical care, and longitudinal management may in part explain the rise noted in the present work [5]. Consistent with this premise, Wright et al. showed refinements in surgical technique and perioperative care to have progressively narrowed the survival gap at the time of repair between patients with and without DS, while a systematic review and meta-analysis by Glinianaia et al. documented significant gains in long-term survival for children with DS and CHD [4,19]. Specifically, enhancements in surgical techniques for AVSD, such as modified single-patch and double-patch repairs, have reduced mortality for these patients [20]. It is possible, therefore, that dedicated transitional and adult‐congenital heart disease care pathways may be needed to ensure that the early gains in childhood survival translate into adulthood for individuals with DS.

Notably, while the aggregate prevalence of valvular disease was comparable between groups, subtype analysis revealed important differences. DS patients demonstrated a significantly higher rate of mitral valve disease, a finding consistent with residual atrioventricular valve dysfunction following AVSD repair, which remains one of the most common long-term sequelae in this population [2,20]. The lower prevalence of aortic valve disease among DS patients likely reflects their lower burden of hypertension and other traditional cardiovascular risk factors. These patterns highlight the importance of ongoing valve surveillance in adults with DS, particularly those with prior AVSD repair.

Although adults with DS and congenital heart disease carry a substantial burden of additional comorbidities, we importantly did not observe longer hospitalizations or greater costs in this cohort. This finding may reflect the combined influence of payer incentives, case-mix, and modern care pathways. Moreover, DS patients in our analysis less commonly underwent high-cost cardiac surgery or catheter-based interventions, reducing the procedure-driven expenditures that may inflate cost among the non-DS CHD populations. Concurrently, the widespread adoption of standardized perioperative pathways and heart-failure management protocols over the past decade has reduced variability in resource use, mitigating the financial impact of additional comorbidities in DS [21]. Together, these observations suggest that payer mix and contemporary practice patterns, rather than intrinsic disease complexity alone, may modulate inpatient resource utilization among adults with DS and CHD.

Importantly, our analysis also revealed adults with DS to experience nearly twice the odds of in-hospital mortality compared to their non-DS counterparts. These findings accord with those of Engsner et al., who demonstrated a twofold higher long-term mortality risk among individuals with CHD and DS compared to those with CHD alone (HR 2.12; 95% CI 1.94–2.31) in a nationwide Swedish registry spanning patients born between 1970 and 2017. Both studies represent comprehensive national analyses leveraging population-based administrative data, yet differ meaningfully in design and scope [6]. Engsner et al. employed a registry-based cohort with sex- and birth-year-matched controls, enabling longitudinal survival analysis across nearly five decades. The present study instead characterizes discrete all-cause inpatient encounters from 2016 to 2022 within the heterogeneous, multi-payer healthcare landscape of the United States. Despite these methodological and systemic distinctions, the convergence of findings across two independent national cohorts substantially reinforces the conclusion that individuals with CHD and DS face persistently elevated mortality risk relative to non-DS counterparts, irrespective of advances in congenital cardiac care.

While the causes for such mortality cannot be definitively elucidated in the present retrospective study, these findings carry two important clinical implications. First, when appropriately selected for intervention, DS patients appear to achieve surgical outcomes comparable to non-DS individuals, consistent with the multicenter experience reported by Goldberg and colleagues [22]. Conversely, for medical admissions for respiratory or infectious conditions, patients with DS continue to face comparatively higher risks of in-hospital mortality. Emerging evidence suggests this vulnerability may be mitigated through structured multidisciplinary care. A cross-sectional study by Santoro and colleagues suggests that dedicated adult DS clinics, integrating primary care with cardiology, endocrinology, neurology, and allied specialties, may streamline care transitions and reduce preventable complications [23]. Notably, however, DS patients frequently present for routine or urgent care outside of specialized cardiac centers, general practitioners and hospitalists are increasingly responsible for their management [24]. Protocolized implementation of such pathway driven multidisciplinary programs beyond tertiary cardiac centers may be necessary to improve outcomes for this vulnerable population during medical admissions.

There are several important limitations to this study inherent due to its retrospective nature and the administrative coding of the NIS database. The NIS is reliant on ICD coding and is therefore subject to variable billing practices across hospitals and divisions. Clinical characteristics, such as echocardiogram findings, patient-specific laboratory values and disease severity, could not be ascertained. Factors like the complexity of CHD, the degree of debilitation from DS, and the management of comorbid conditions were similarly unavailable, potentially affecting both admission profiles and outcomes. Furthermore, the NIS does not provide detailed cause-of-death data, limiting the ability to differentiate between cardiac-related and non-cardiac-related mortality, which may be particularly relevant given the higher prevalence of respiratory and infectious admissions in DS patients. Additionally, the non-DS comparison cohort was defined by the absence of a concomitant ICD-10 diagnosis of DS and may therefore include patients with other genetic, syndromic, or neurological conditions; as the NIS does not reliably capture the etiology or severity of neurological impairment, meaningful stratification on this basis was not feasible, which may have influenced outcome ascertainment in this group. Lastly, we are unable to comment on outpatient or subsequent encounters due to the inherent data structure of the NIS.

In conclusion, our national analysis reveals a stable proportion of DS patients among congenital heart disease hospitalizations, coupled with lower procedural utilization and higher in-hospital mortality compared to non-DS counterparts. Although DS patients demonstrate comparable outcomes to non-DS patients when undergoing cardiac procedures, they frequently present with distinct comorbidities, which likely contribute to their increased vulnerability. These findings emphasize the need for proactive multidisciplinary management for adults with DS and CHD, translating early survival gains into durable, high-quality adult outcomes.

CRediT authorship contribution statement

Kevin Tabibian: Writing – review & editing, Writing – original draft, Visualization, Methodology, Formal analysis, Data curation, Conceptualization. Arjun Chaturvedi: Writing – review & editing, Writing – original draft, Validation, Data curation. Amulya Vadlakonda: Writing – review & editing, Validation, Data curation, Conceptualization. Jeffrey Balian: Writing – review & editing, Writing – original draft, Visualization. Sona Mahrokhi: Writing – review & editing, Writing – original draft. Daniel Tabibian: Investigation, Writing – original draft, Writing – review & editing. Oh Jin Kwon: Writing – review & editing, Writing – original draft, Validation, Conceptualization. Esteban Aguayo: Writing – review & editing, Conceptualization. Peyman Benharash: Writing – review & editing, Validation, Supervision, Data curation, Conceptualization.

Disclosure

Dr. Peyman Benharash received fees from AtriCure as a surgical proctor. This manuscript does not discuss any related products or services.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Peyman Benharash reports a relationship with AtriCure Inc that includes: consulting or advisory. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article: