Abstract
Background
COVID-19 has caused significant worldwide morbidity and mortality. Congenital heart disease (CHD) is likely to increase vulnerability and understanding the predictors of adverse outcomes is key to optimising care.
Objective
Ascertain the impact of COVID-19 on people with CHD and define risk factors for adverse outcomes.
Methods
Multicentre UK study undertaken 1 March 2020–30 June 2021 during the COVID-19 pandemic. Data were collected on CHD diagnoses, clinical presentation and outcomes. Multivariable logistic regression with multiple imputation was performed to explore predictors of death and hospitalisation.
Results
There were 405 reported cases (127 paediatric/278 adult). In children (age <16 years), there were 5 (3.9%) deaths. Adjusted ORs (AORs) for hospitalisation in children were significantly lower with each ascending year of age (OR 0.85, 95% CI 0.75 to 0.96 (p<0.01)). In adults, there were 24 (8.6%) deaths (19 with comorbidities) and 74 (26.6%) hospital admissions. AORs for death in adults were significantly increased with each year of age (OR 1.05, 95% CI 1.01 to 1.10 (p<0.01)) and with pulmonary arterial hypertension (PAH; OR 5.99, 95% CI 1.34 to 26.91 (p=0.02)). AORs for hospitalisation in adults were significantly higher with each additional year of age (OR 1.03, 95% CI 1.00 to 1.05 (p=0.04)), additional comorbidities (OR 3.23, 95% CI 1.31 to 7.97 (p=0.01)) and genetic disease (OR 2.87, 95% CI 1.04 to 7.94 (p=0.04)).
Conclusions
Children were at low risk of death and hospitalisation secondary to COVID-19 even with severe CHD, but hospital admission rates were higher in younger children, independent of comorbidity. In adults, higher likelihood of death was associated with increasing age and PAH, and of hospitalisation with age, comorbidities and genetic disease. An individualised approach, based on age and comorbidities, should be taken to COVID-19 management in patients with CHD.
Keywords: COVID-19; EPIDEMIOLOGY; Heart Defects, Congenital; RISK FACTORS
WHAT IS ALREADY KNOWN ON THIS TOPIC
Infection with respiratory viruses can have serious consequences in people with congenital heart disease, however, understanding of the impact of COVID-19 in this population is limited.
WHAT THIS STUDY ADDS
New data pertaining to the impact of the first two waves of the COVID-19 pandemic in the UK on people with congenital heart disease, and identification of key risk factors for worse outcomes.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Identification of medical and non-medical factors associated with increased rates of death in children and adults with congenital heart disease could help inform future decision-making regarding shielding when rates of COVID-19 infection are high or there are new variants.
Introduction
The COVID-19 pandemic commenced in theUK in 2020 and by 23 March 2020 groups in the population considered high risk of severe disease were advised to ‘shield’ or severely limit social contacts. The British Congenital Cardiac Association (BCCA) was asked to help define which subgroups of patients with congenital heart disease (CHD) should receive tailored clinical advice based on professional consensus due to limited evidence beyond experience with other viruses. The BCCA commissioned the UK-wide COVID-19 in CHD (COaCHeD) study to address the knowledge vacuum for patients with CHD in the UK and to document the experience of COVID-19 illness in patients with CHD.
Knowledge of the effects of COVID-19 infection on people with CHD has been informed by studies within the adult population.1 2 Data from the paediatric population remain limited, although severe disease, hospitalisation and death are less frequently reported than for adults.3–5 We published data on early reported cases of COVID-19 from a review of the literature in 2020 to help guide this study.6 Paediatric inflammatory multisystem syndrome temporally associated with SARS-CoV2 (COVID-19) (PIMS-TS/MIS-C) has a clear cardiac phenotype but is not specifically associated with CHD.7 8
This study aimed to investigate the impact of COVID-19 infection on children and adults with CHD over the first and second waves of the pandemic in the UK, during which management evolved due to research, shielding and seasonal variation. We specifically investigated the risk factors predictive of survival or a medical decision to admit to hospital.
Methods
This multicentre study involved 23 congenital cardiac services in 18 UK centres (12 paediatric, 11 adult).It had ethical approval from the Health Research Authority, HRA/HRCW:20/HRA/2958 and the Royal Brompton and Harefield NHS Foundation Trust was the co-ordinating centre.
Participants were children (aged under 16 years) and adults (16 years and older) with CHD who were infected with COVID-19, defined as SARS-COV-2 confirmed by PCR testing or lateral flow testing (LFT), between 1 March 2020 and 30 June 2021.
Children were eligible for inclusion if they had a diagnosis of CHD, arrhythmic disease (confirmed electrophysiological diagnosis), inherited cardiac disease (such as hypertrophic cardiomyopathy) and/or previous cardiac transplantation. Children with PIMS-TS/MIS-C but without CHD were excluded. Adults were eligible for inclusion if they had a diagnosis of CHD. Adults with arrhythmia, inherited cardiac disease and previous cardiac transplantation were excluded.
Congenital cardiologists who were members of the BCCA were prompted to report cases by monthly updates in the BCCA newsletter and emails from the central study team. To improve ascertainment, adult patients, or parents of children, could self-refer to the study team. Several patient support charities disseminated information about the study including Little Hearts Matter, the Brompton Fountain, the Somerville Foundation, ECHO (Evelina Children’s Heart Organisation) and Leeds Congenital Hearts.
Clinicians completed standardised questionnaires using clinical case notes. As data were anonymised, individual consent was waived. Information was collected on demographics (age, area deprivation and ethnicity), cardiac diagnosis/interventions, oxygen saturations, comorbidities (defined as a confirmed diagnosis of an additional medical or psychological condition above that of the diagnosis of CHD), genetic disease (defined as a specific disease-causing mutation or syndrome) and outcomes following COVID-19 infection. For self-referred/parent-referred participants, standardised questionnaires were self-completed, and consent was sought to contact the local cardiologist to validate data. Duplicate reports were merged with the primary report.
Data analysis
Only the patient’s first episode of COVID-19 infection was included in analyses. ‘Severe disease’ was defined as an episode requiring hospital admission, including emergency department assessment and ‘hospice at home’ management. Analysis of hospital admissions in this study include cases who were admitted to hospital for other reasons and were subsequently found to have COVID-19. Death with COVID-19 was defined according to UK government criteria as death within 28 days of a positive COVID-19 test.9 Ethnicity was grouped using the Office for National Statistics Census 2011 framework into white, Asian, black, mixed and other. Participants were assigned to quintiles of deprivation according to area deprivation score reported by clinicians using English and Welsh indices of deprivation.10 11 Children’s weights were converted to z-scores for age and sex. Adult body mass index (BMI) was included as a binary variable (<30 kg/m2 or ≥30 kg/m2). Baseline oxygen saturations at diagnosis were categorised as <90% or ≥90%.
Factors related to mortality in COVID-19 (age, sex, obesity, deprivation and ethnicity) or to CHD outcomes (type of circulation, pulmonary arterial hypertension (PAH), comorbid conditions, genetic diagnosis and oxygen saturations) were investigated as potential predictors of hospital admission in both adults and children, and mortality in adults. Due to the low number of deaths, predictors of mortality in children were not explored. We explored individual predictors in univariable analyses then constructed a multivariable model using stepwise addition and evaluation of model fit. No interaction terms were significant. In a sensitivity analysis, multivariable models were adjusted for COVID-19 wave (wave 1 before 1 September 2020; wave 2 on and after 1 September 2020) to take account of potential changes in the approach to management. We initially used a complete-case approach then applied multiple imputation using chained equations to address missing data and minimise bias. The imputation model generated 25 imputed datasets and included all variables in the analysis model, as well as the outcome and one auxiliary variable (CHD diagnosis).
Analyses were performed using Stata V.17 (StataCorp. 2021). Data are presented as median and IQR or n (%) for continuous and categorical data, respectively. A p<0.05 was considered statistically significant.
Results
There were 405 eligible participants; median (IQR) age of participants was 26(11–38) years and 189 (46.7%) were female (table 1).
Table 1.
Demographic and characteristic data of participants in the study presented as all cases and as paediatric and adult subgroups
| All patients (n=405) | Children (<16) (n=127) | Adults (≥16) (n=278) | |
| Age in years at COVID-19 diagnosis Median (IQR) |
26 (11–38) | 5 (1–10) (52 (11–122) months) |
32 (25–46) |
| Age≤1 year, n (%) | 32 (25.2) | ||
| Age>50 years, n (%) | 41 (14.7) | ||
| Sex, n (%) | |||
| Female | 189 (46.7) | 52 (40.9) | 137 (49.3) |
| Male | 194 (47.9) | 75 (59.1) | 119 (42.8) |
| Missing | 22 (5.4) | 0 (0.0) | 22 (7.9) |
| Ethnicity, n (%) | |||
| White | 174 (43.0) | 59 (46.5) | 115 (41.4) |
| Asian/Asian British (including Chinese) | 41 (10.1) | 23 (18.1) | 18 (6.5) |
| Black/black British | 11 (2.7) | 6 (4.7) | 5 (1.8) |
| Mixed | 4 (1.0) | 3 (2.4) | 1 (0.4) |
| Other (including Arab) | 7 (1.7) | 6 (4.7) | 1 (0.4) |
| Missing | 168 (41.5) | 30 (23.6) | 138 (49.6) |
| Deprivation quintile, n (%) | |||
| 1 | 99 (24.4) | 28 (22.0) | 71 (25.5) |
| 2 | 75 (18.5) | 29 (22.8) | 46 (16.5) |
| 3 | 73 (18.0) | 17 (13.4) | 56 (20.1) |
| 4 | 61 (15.1) | 17 (13.4) | 44 (15.8) |
| 5 | 55 (13.6) | 15 (11.8) | 40 (14.4) |
| Missing | 42 (10.4) | 21 (16.5) | 21 (7.6) |
| No in household (excluding patient), n(%) | |||
| 0–3 | 52 (12.8) | 20 (15.7) | 32 (11.5) |
| 4–6 | 48 (11.8) | 32 (25.2) | 16 (5.7) |
| >11 (includes care homes, hospital, etc) | 2 (0.4) | 0 (0.0) | 2 (0.7) |
| Missing | 303 (74.8) | 75 (59.0) | 228 (82.0) |
| Anthropometry | |||
| Adult BMI Median (IQR) |
22.9 (18.3–27.1) | 16.7 (15.2–19.5) | 24.4 (22–29) |
| Adult BMI>30, n (%) | 37 (9.14) | ||
| Weight z-scores (children) Median (IQR |
−0.72 (−2.2 to 0.89) | ||
| IOTF category (children) n (%) | |||
| Grade 3 Thin | 30 (23.6) | ||
| Grade 2 Thin | 9 (7.1) | ||
| Grade 1 Thin | 12 (9.5) | ||
| Normal | 18 (14.2) | ||
| Overweight | 4 (3.2) | ||
| Obese | 1 (0.8) | ||
| Missing | 53 (41.7) | ||
| Circulation, n (%) | |||
| Biventricular | 334 (82.5) | 106 (83.5) | 228 (82.0) |
| Mixed | 6 (1.5) | 2 (1.6) | 4 (1.4) |
| Univentricular | 63 (15.6) | 17 (13.4) | 46 (16.6) |
| Missing | 2 (0.5) | 2 (1.6) | 0 (0.0) |
| Genetic diagnosis, n (%)† | |||
| Yes | 57 (14.1) | 29 (22.8) | 28 (10.1) |
| No | 279 (68.9) | 70 (55.1) | 209 (75.2) |
| Missing | 69 (17.0) | 28 (22.1) | 41 (14.8) |
| At least one comorbidity, n (%)* | |||
| Yes | 170 (42.0) | 54 (42.5) | 116 (41.7) |
| No | 164 (40.5) | 47 (37.0) | 117 (42.1) |
| Missing | 71 (17.5) | 26 (20.5) | 45 (16.2) |
*Comorbidities across multiple organ systems, including neurodisability, chronic lung disease, immune deficiency (myelodysplastic syndrome, asplenia), chronic kidney disease, diabetes and previous malignancy (online supplemental table 6 for complete list).
†Genetic diagnosis—16 different diagnoses including Trisomy 21, 22q11 microdeletion syndrome, Noonan syndrome, William syndrome, Ehlers Danlos syndrome, Nemaline Myopathy and Hurler syndrome (online supplemental table 5 for complete list).
BMI, body mass index; IOTF, International Obesity Task Force.
openhrt-2023-002356supp001.pdf (112.8KB, pdf)
A broad spectrum of CHD was represented (table 2; online supplemental table 1). Of these, 128 (31.6%) had an associated genetic diagnosis (online supplemental table 2) and 170 (42.0%) at least one comorbidity (online supplemental table 3). Seventy-four (18.3%) were asymptomatic and identified through hospital screening or the UK test and trace system.12 The case/hospitalisation ratio was 29.1% (118/405) and case fatality ratio was 7.2% (29/405).
Table 2.
Primary cardiac diagnosis of participants
| Congenital heart disease category | All (n=405) n (%) | Children (n=127) n (%) | Adults (n=278) n (%) |
| Hypoplastic left heart syndrome | 11 (2.7) | 9 (7.1) | 2 (0.7) |
| Functionally univentricular heart | 34 (8.4) | 7 (5.5) | 27 (9.7) |
| Common arterial trunk | 6 (1.5) | 5 (3.9) | 1 (0.4) |
| Transposition of the great arteries and VSD/double outlet right ventricle transposition type | 20 (4.9) | 11 (8.7) | 9 (3.2) |
| Interrupted aortic arch | 1 (0.3) | 0 (0.0) | 1 (0.4) |
| Transposition of the great arteries | 13 (3.2) | 3 (2.4) | 10 (3.6) |
| Pulmonary atresia with intact ventricular septum | 12 (3.0) | 2 (1.6) | 10 (3.6) |
| Pulmonary atresia with VSD | 23 (5.7) | 5 (3.9) | 18 (6.5) |
| Miscellaneous congenital heart disease* | 11 (2.7) | 1 (0.8) | 10 (3.6) |
| Atrioventricular septal defect | 33 (8.2) | 12 (9.5) | 21 (7.6) |
| Tetralogy of Fallot/double outlet right ventricle tetralogy type | 52 (12.8) | 17 (13.4) | 35 (12.6) |
| Aortic stenosis | 31 (7.7) | 6 (4.7) | 25 (9.0) |
| Tricuspid valve abnormality/Ebstein abnormality | 11 (2.7) | 2 (1.6) | 9 (3.2) |
| Mitral valve abnormality | 15 (3.7) | 5 (3.9) | 10 (3.6) |
| Total anomalous pulmonary venous connection | 3 (0.7) | 2 (1.6) | 1 (0.4) |
| Aortic arch obstruction | 30 (7.4) | 4 (3.2) | 26 (9.4) |
| Pulmonary stenosis | 15 (3.7) | 6 (4.7) | 9 (3.2) |
| Subaortic stenosis | 1 (0.3) | 0 (0.0) | 1 (0.4) |
| VSD±atrial septal defect±patent ductus arteriosus | 41 (10.1) | 11 (8.7) | 30 (10.8) |
| Atrial septal defect | 21 (5.2) | 4 (3.2) | 17 (6.1) |
| Patent ductus arteriosus | 1 (0.3) | 1 (0.8) | 0 (0.0) |
| Arrhythmia | 1 (0.3) | 1 (0.8) | 0 (0.0) |
| Miscellaneous terms† | 10 (2.7) | 5 (3.9) | 5 (1.8) |
| Cardiomyopathy (including post-transplant) | 8 (2.2) | 8 (1.3) | NA |
| Aortic regurgitation | 1 (0.3) | 0 (0.0) | 1 (0.3) |
Note: A single primary cardiac diagnosis was assigned to each case using a hierarchical system.18
*Miscellaneous congenital heart disease includes: congenitally corrected transposition of the great arteries±other morphological changes.
†Miscellaneous terms include: postcardiac transplant and aortic root dilatation/dissection.
VSD, ventricular septal defect.
Paediatric cases
There were 127 paediatric cases, of whom 32 (25.2%) were ≤1 year and 52 (40.9%) were female. Median (IQR) age at COVID-19 diagnosis was 52 (11–122) months. Twenty-nine (22.8%) had an underlying genetic diagnosis, of whom 12 had T21, and 54 at least one comorbidity. The most common presenting symptoms were fever (n=31 (24.4%)), coryza (n=26 (20.5%)) and respiratory distress (n=24 (18.9%)); 50 (39.4%) were asymptomatic.
Deaths
Five children died (four aged ≤12 months); in one case, death was due to COVID-19 pneumonitis. All children who died had significant CHD-related comorbidity which was stated as the primary cause of death on the death certificate, including sepsis, multiorgan failure and acute renal failure, and prolonged hospital admission before death (range 12–90 days). These children met UK criteria for a death involving COVID-19,9 however, in the view of their clinical team COVID-19 was not the primary cause of death.
Hospital admissions
There were 44 (34.6%) children admitted to hospital (34 in a cardiac specialist centre; 10 in a local district hospital); 24 were female, and median (IQR) age at diagnosis was 13.5 (2.8–72) months. Twenty-six hospitalised paediatric cases had at least one comorbidity, most often respiratory disease, preterm birth and neurological disease, 11 had a genetic diagnosis and 6 had univentricular circulations.
Approximately half of paediatric cases managed in hospital had respiratory distress prior to admission (n=21) and 27 were treated with antibiotic therapy. Ten children required oxygen therapy and 11 were intubated and ventilated (clinicians reported ventilating 5 children due to COVID-19 and 6 due to CHD). Eight were treated with inotropes (two prior to COVID-19 infection), of which seven were also ventilated. Ten cases were asymptomatic and identified as having COVID-19 while admitted to hospital for another non-related reason. Median (IQR) length of stay was 7 (4–30) days (table 3).
Table 3.
Clinical features and management of COVID-19 infection for all cases and hospitalised cases
| Variable | All cases (n=405) | Hospitalised patients | |
| Children (n=44) | Adults (n=74) | ||
| Symptoms prior to hospital admission, n (%) | |||
| Shortness of breath | 88 (21.7) | 21 (47.7) | 22 (29.7) |
| Coryza | 97 (24.0) | 15 (34.1) | 15 (20.3) |
| Anosmia | 46 (11.4) | 0 (0.0) | 3 (4.1) |
| Chest Pain | 16 (4.0) | 0 (0.0) | 3 (4.1) |
| Asymptomatic | 75 (18.5) | 10 (22.7) | 5 (6.8) |
| Missing | 0 (0.0) | 0 (0.0) | 0 (0.0) |
| Medications administered, n (%) | |||
| Intravenous antibiotics | 43 (10.6) | 27 (61.4) | 14 (18.9) |
| Antivirals | 9 (2.2) | 9 (20.5) | 0 (0.0) |
| Steroids | 19 (4.7) | 8 (18.2) | 10 (13.5) |
| Immunoglobulin | 2 (0.5) | 2 (4.5) | 0 (0.0) |
| Missing | 308 (76.0) | 2 (4.5) | 50 (6.8) |
| Respiratory support in hospital, n (%) | |||
| Oxygen | 28 (6.9) | 10 (22.7) | 8 (10.8) |
| High flow oxygen | 6 (1.5) | 5 (11.4) | 1 (1.4) |
| Non-invasive ventilation | 3 (0.7) | 3 (6.8) | 0 (0.0) |
| Intubation and ventilation | 12 (3.0) | 11 (25.0) | 1 (1.4) |
| Nitric oxide | 2 (0.5) | 2 (4.6) | 0 (0.0) |
| Missing | 276 (68.1) | 1 (2.3) | 50 (67.6) |
| Circulatory support in hospital, n (%) | |||
| Inotropes | 9 (2.2) | 8 (18.2) | 1 (1.4) |
| Extra-corporeal membrane oxygenation | 1 (0.3) | 1 (2.3) | 0 (0.0) |
| Missing | 312 (77.0) | 1 (2.3) | 55 (74.3) |
| Death within 28 days of COVID-19, n (%) | |||
| Yes | 29 (7.1) | 5 (11.4) | 17 (23.0) |
| No | 361 (88.9) | 39 (88.6) | 48 (64.9) |
| Missing | 16 (3.9) | 0 (0.0) | 9 (12.2) |
| Length of hospital stay (days), median (IQR) | |||
| 7 (4-30) | 8 (4-18) | ||
| Missing n (%) | 9 (20.5) | 41 (55.4) | |
Predictors of hospitalisation
In univariable analyses (table 4), the OR of hospitalisation decreased significantly with each additional year of age (OR 0.83, 95% CI 0.75 to 0.91 (p<0.001)), and was significantly increased for females (OR 2.32, 95% CI 1.10 to 4.88 (p=0.03)) and those presenting with respiratory distress (OR 24.04, 95% CI 6.58 to 87.85 (p<0.001)). Asymptomatic cases were at reduced risk of hospitalisation (OR 0.28, 95% CI 0.18 to 0.63 (p=0.002)).
Table 4.
Predictors of hospitalisation and death (in the adult population) in univariable logistic regression models
| Outcome | Children | Adults | |
| Severe disease | Death | Hospitalisation | |
| OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| Pulmonary arterial hypertension | 5.89 (2.30 to 15.07) p<0.001 | 3.74 (1.80 to 7.79) p<0.001 | |
| Body mass index >30 kg/m2 | 0.48 (0.11 to 2.21) p=0.35 | 1.93 (0.95 to 3.94) p=0.07 | |
| Body mass index | 0.94 (0.83 to 1.06) p=0.29 | 1.05 (0.98 to 1.12) p=0.17 | 1.07 (1.02 to 1.13) p=0.01 |
| z-score weight | 0.91 (0.77 to 1.09) p=0.31 | ||
| Univentricular circulation | 1.03 (0.35 to 3.01) p=0.95 | 0.41 (0.09 to 1.79) p=0.23 | 0.57 (0.26 to 1.25) p=0.16 |
| Prior use of anticoagulation | 0.99 (0.41 to 2.40) p=0.99 | 1.22 (0.40 to 3.72) p=0.73 | 1.46 (0.72 to2.98) p=0.30 |
| At least one comorbidity | 1.54 (0.70 to 3.39) p=0.28 | 2.13 (0.86 to 5.25) p=0.10 | 6.85 (3.39 to13.87) p<0.001 |
| Genetic disease | 0.72 (0.30 to 1.74) p=0.47 | 4.89 (1.76 to 13.64) p=0.002 | 7.78 (3.18 to 19.04) p<0.001 |
| Baseline oxygen saturations <90% | 1.62 (0.61 to 4.25) p=0.33 | 2.51 (0.98 to 6.46) p=0.06 | 2.24 (1.15 to 4.39) p<0.001 |
| Baseline oxygen saturations | 1.19 (0.72 to 1.72) p=0.61 | 1.97 (1.29 to 2.99) p=0.002 | 1.51 (1.11 to 2.05) p=0.008 |
| Female sex | 2.32 (1.10 to 4.88) p=0.03 | 0.75 (0.32 to 1.76) p=0.50 | 1.07 (0.62 to 1.87) p=0.81 |
| Age (years) | 0.83 (0.75 to 0.91) p<0.001 | 1.05 (1.02 to 1.08) p<0.001 | 1.04 (1.02 to 1.06) p<0.001 |
| Age >50 years | 3.79 (1.48 to 9.72) p=0.006 | 3.56 (1.73 to 7.33) p=0.001 | |
| Age≤1 year | 6.00 (2.56 to 14.04) p<0.001 | ||
| Ethnicity (reference: white) | |||
| Asian | 1.79 (0.67 to 4.76) p=0.25 | 0.98 (0.20 to 4.76) p=0.98 | 1.01 (0.35 to 2.91) p=0.98 |
| Black | 3.90 (0.66 to 23.15) p=0.13 | 1.96 (0.20 to 18.91) p=0.56 | 1.35 (0.22 to 8.43) p=0.75 |
| Mixed or other | 0.61 (0.24 to 1.52) p=0.29 | 0.54 (0.22 to 1.36) p=0.19 | 0.56 (0.33 to 0.73) p=0.05 |
| Asymptomatic | 0.28 (0.18 to 0.63) p=0.002 | 0.63 (0.08 to 5.08) p=0.665 | 0.84 (0.30 to 2.39) p=0.745 |
| Respiratory distress | 24.04 (6.58 to 87.85) p<0.001 | 1.31 (0.52 to 3.31) p=0.57 | 1.57 (0.86 to 2.88) p=0.15 |
‘Severe disease’ defined as an episode requiring hospital admission, including emergency department assessment and ‘hospice at home’ management).
In a multivariable logistic regression analysis using imputed data, the OR for hospitalisation decreased significantly with each additional year of age (OR 0.85, 95% CI 0.75 to 0.96 (p<0.01)), after adjustment for ethnicity, sex, deprivation, number of comorbidities, circulation type, baseline oxygen saturations and weight z-score (table 5). No other independent predictors of hospitalisation were identified. In sensitivity analyses including COVID-19 wave, hospitalisation risk was significantly higher in the first wave, while younger age remained an independent predictor (online supplemental table 4).
Table 5.
Predictors of hospitalisation in children using multivariable logistic regression with the multiple imputation model (n=127)
| Outcome | OR | P value | 95% CI |
| Ethnicity (Reference: white) | |||
| Asian | 1.12 | 0.87 | 0.30 to 4.13 |
| Black | 2.46 | 0.38 | 0.33 to 18.40 |
| Mixed/other | 0.54 | 0.28 | 0.17 to 1.66 |
| Univentricular circulation | 0.32 | 0.23 | 0.05 to 2.05 |
| Age (years) | 0.85 | <0.01 | 0.75 to 0.96 |
| Female sex | 2.57 | 0.05 | 0.98 to 6.73 |
| Deprivation quintile (Reference: 1=most deprived) | |||
| 2 | 0.55 | 0.35 | 0.15 to 1.96 |
| 3 | 2.12 | 0.33 | 0.47 to 9.52 |
| 4 | 0.73 | 0.70 | 0.14 to 3.71 |
| 5 | 0.88 | 0.89 | 0.15 to 5.21 |
| No of comorbidities | 1.64 | 0.32 | 0.62 to 4.33 |
| Genetic disease | 0.56 | 0.34 | 0.18 to 1.83 |
| Oxygen saturations <90% | 2.29 | 0.31 | 0.46 to 11.44 |
| Z-score weight | 0.91 | 0.38 | 0.74 to 1.12 |
Adult cases
There were 278 adult cases; 137 (49.3%) were female, 28 (10.1%) had a genetic diagnosis and 116 (41.7%) at least one comorbidity. Median (IQR) age at diagnosis of COVID-19 was 32 (25–46) years. The most common presenting symptoms were fever (31 (11.2%)), coryza (26 (9.4%)) and respiratory distress (24 (8.6%)); 25 (9.0%) were asymptomatic.
Deaths
Death occurred in 24 (8.6%), all adult congenital heart disease (ACHD) cases met the UK criteria for death involving COVID-19.9 Of these, 19 had comorbidities, 8 had associated genetic disease (T21 was most common), 2 had univentricular circulation and 9 had PAH.
Predictors of death
Univariable analysis showed a significant increased risk of death in adults with PAH (OR 5.89, 95% CI 2.30 to 15.07 (p<0.001)), genetic disease (OR 4.89, 95% CI 1.76 to 13.64 (p=0.0020) and lower baseline oxygen saturations (OR 1.97, 95% CI 1.29 to 2.99 (p=0.002)) while the likelihood of hospitalisation was significantly greater with each increased year of age (OR 1.05, 95% CI 1.02 to 1.08 (p<0.001)).
In multivariable logistic regression analyses using imputed data, the OR for death increased significantly with each year of age (OR 1.05, 95% CI 1.01 to 1.10 (p<0.01)) and with the presence of PAH (OR 5.99, 95% CI 1.34 to 26.91 (p=0.02)) after adjustment for ethnicity, sex, quintile of deprivation, number of comorbidities, type of circulation, genetic disease, baseline oxygen saturations and BMI (table 6). In sensitivity analyses, COVID-19 wave was not an independent predictor of mortality (online supplemental table 5).
Table 6.
Predictors of mortality in adults using multivariable logistic regression with the multiple imputation model (n=278)
| Outcome | OR | P value | 95% CI |
| Ethnicity (Reference: white) | |||
| Asian | 1.19 | 0.47 | 0.31 to 12.63 |
| Black | 1.19 | 0.91 | 0.05 to 28.66 |
| Mixed/other | 0.73 | 0.57 | 0.25 to 2.15 |
| Univentricular circulation | 0.70 | 0.67 | 0.13 to 3.79 |
| Age (years) | 1.05 | <0.01 | 1.01 to 1.10 |
| Female sex | 0.47 | 0.18 | 0.16 to 1.41 |
| Deprivation quintile (Reference: 1=most deprived) | |||
| 2 | 0.26 | 0.14 | 0.04 to 1.54 |
| 3 | 0.32 | 0.19 | 0.06 to 1.76 |
| 4 | 0.32 | 0.21 | 0.05 to 1.92 |
| 5 | 1.54 | 0.53 | 0.39 to 6.07 |
| No of comorbidities | 0.30 | 0.11 | 0.07 to 1.31 |
| Genetic disease | 4.63 | 0.05 | 0.99 to 21.61 |
| Pulmonary arterial hypertension | 5.99 | 0.02 | 1.34 to 26.91 |
| Oxygen saturations <90% | 2.41 | 0.25 | 0.54 to 10.82 |
| BMI ≥30 | 2.02 | 0.40 | 0.39 to 10.50 |
Hospital admissions
There were 74 (26.6%) cases admitted to hospital; 38 were female and their median (IQR) age was 38 (28–52) years. Comorbidity was present in 50, univentricular circulation in 9, genetic disease in 18 and PAH in 19 patients. There was a wide range of cardiac diagnoses (table 2). Five cases were asymptomatic and admitted to hospital for another non-related reason. Median (IQR) length of hospital admission was 8 (4–18) days (table 3).
Predictors of hospitalisation
Univariable analysis showed that severe disease (defined as an episode requiring hospital admission, including emergency department assessment and ‘hospice at home’ management) was significantly associated with PAH (OR 3.74, 95% CI 1.80 to 7.79 (p<0.001)), greater BMI (OR 1.07, 95% CI 1.02 to 1.13 (p=0.01)), additional comorbidities (OR 6.85, 95% CI 3.39 to 13.87 (p<0.001)), genetic disease (OR 7.78, 95% CI 3.18 to 19.04 (p<0.001) and lower baseline oxygen saturation (OR 1.51, 95% CI 1.11 to 2.05 (p=0.008)). The OR for hospitalisation was significantly greater with each increased year of age in adults (OR 1.04, 95% CI 1.02 to 1.06 (p<0.001)).
Multivariable logistic regression analysis using imputed data demonstrated that the OR for hospitalisation increased significantly with each year of age (OR 1.03, 95% CI 1.00 to 1.05 (p=0.04)), greater number of comorbidities (OR 3.23, 95% CI 1.31 to 7.97 (p=0.01)) and genetic disease (OR 2.87, 95% CI 1.04 to 7.94 (p=0.04)) after adjustment for ethnicity, sex, quintile of deprivation, baseline oxygen saturations, PAH and BMI (table 7). The OR for hospitalisation was significantly lower in those with univentricular or mixed circulation compared with biventricular (OR 0.27, 95% CI 0.09 to 0.81 (p=0.02)). In sensitivity analyses including COVID-19 wave, hospitalisation risk was significantly higher in the first wave; older age remained an independent predictor while the influence of comorbidities and genetic disease were attenuated (online supplemental table 6).
Table 7.
Predictors of hospitalisation in adults using multivariable logistic regression with the multiple imputation model (n=278)
| Outcome | OR | P value | 95% CI |
| Ethnicity (Reference: white) | |||
| Asian | 1.52 | 0.50 | 0.45 to 5.14 |
| Black | 0.42 | 0.43 | 0.05 to 3.66 |
| Mixed/other | 0.49 | 0.05 | 0.24 to 1.01 |
| Univentricular circulation | 0.27 | 0.02 | 0.09 to 0.81 |
| Age (years) | 1.03 | 0.04 | 1.00 to 1.05 |
| Female sex | 0.90 | 0.78 | 0.45 to 1.82 |
| Deprivation quintile (Reference: 1=most deprived) | |||
| 2 | 0.79 | 0.63 | 0.29 to 2.10 |
| 3 | 0.61 | 0.33 | 0.23 to 1.65 |
| 4 | 0.44 | 0.13 | 0.15 to 1.28 |
| 5 | 0.42 | 0.13 | 0.13 to 1.29 |
| No of comorbidities | 3.23 | 0.01 | 1.31 to 7.97 |
| Genetic disease | 2.87 | 0.04 | 1.04 to 7.94 |
| Pulmonary arterial hypertension | 1.14 | 0.80 | 0.41 to 3.19 |
| Oxygen saturations <90% | 2.73 | 0.06 | 0.94 to 7.90 |
| BMI ≥30 | 1.58 | 0.36 | 0.59 to 4.27 |
BMI, body mass index.
Discussion
This large UK-wide study describes paediatric and adult CHD populations in the UK diagnosed with COVID-19 during the first two waves of the pandemic, a period of changing infection rates attributable to seasonal variation, shielding and evolution of disease variants. Key findings are that 7% of patients died (fulfilling the UK criteria for death involving COVID-199) and 29% were admitted to hospital with COVID-19 (who were either symptomatic with the disease or asymptomatic and admitted for another reason). Independent predictors of hospitalisation with COVID-19 infection were younger age in children, and older age, comorbidities and genetic disease in adults. Older age and PAH (Eisenmenger syndrome) were independent predictors of death for adults with CHD and COVID-19.
Case-fatality rate in the paediatric population was low with only one death directly attributed to COVID-19 pneumonitis. Low fatalities secondary to COVID-19 ranging from 0% to 6% were also seen in other paediatric cardiology studies.3–5 13 One Indian study reported 13 (27%) deaths in children with CHD, however, most were young and unoperated.14 Within the paediatric CHD population, hospital admission rates with COVID-19 were also low.3–5
Previous reports are consistent with our finding of greater risk of severe disease in younger patients13 14 and highlights the potential vulnerabilities of neonates/infants who are unoperated14 or undergoing cardiac surgery. Given these vulnerabilities, consideration of limiting social contacts in these groups prior to a planned hospital admission for intervention may help protect against disease transmission at this critical time.
Hospitalised patients with CHD in this study represent a fraction of all UK reported COVID-19 admissions (n=113/474 304; 0.02%)15 and deaths (n=21/128 297; 0.02%)9 between 19 March 2020 and 30 June 2021. The BCCA published online guidelines for vulnerable groups with CHD on its website related to UK government advice.16 Given the lack of published evidence of risk within the CHD population, stricter social distancing measures were advised for patients with PAH and infants under 1 year with unrepaired CHD, both of which are groups confirmed to have higher OR for adverse outcomes within our study, as well as for patients with univentricular circulation and cardiomyopathies, who we have not found to be at higher risk. In contrast, we found that adults with univentricular circulation were less likely to be admitted to hospital than those with biventricular circulation, one hypothesis for this could be that those with univentricular circulation may have adopted a more cautious approach in terms of preventative measures against COVID-19. Consistent with previous reports from the general adult population, we identified an increased risk of hospitalisation in older patients. However, the median age of patients with ACHD admitted to hospital was lower than for the UK adult population with COVID-19; this has also been demonstrated in similar studies of the ACHD population.1 2 12
Of deaths within the ACHD population, 10 (3.6%) were directly attributed to COVID-19, which is consistent with a large international study of patients with ACHD where 24 (2.3%) deaths were directly attributed to COVID-19.1 Other studies reported higher rates, including a US study with 41 (10.5%) deaths, however, this study was unable to differentiate cases that had COVID-19 as a primary or secondary cause.13 Risk of death in patients with ACHD with PAH was elevated in this study in keeping with the published literature.1 2 Such patients represent a high-risk population in terms of complications and mortality due to reduced cardiorespiratory reserve and chronic cyanosis. Acute COVID-19 exacerbates hypoxaemia and thrombotic risk associated with PAH. The higher complication risk within this group supports the principle that each patient with CHD should receive individualised care.
People with T21 may have comorbidities in different organ systems impacting on their ability to cope with infection; these include CHD, increased risk of PAH, obesity and respiratory disease. Some studies have shown more severe outcomes for those with T21 and COVID-19, but importantly also for those with additional comorbidities.17 Although we observed higher likelihood of hospitalisation and mortality for people with genetic disease, particularly T21, in our complete case models, this was attenuated after imputation of missing data. Nevertheless, the data suggest that people with T21 are a high-risk group and should have priority for vaccine roll-out regardless of CHD involvement.
Many paediatric cases (50/127 (39.4%)) were asymptomatic, including ten managed in hospital who were admitted for another medical reason. Our findings highlight the importance of testing asymptomatic children with CHD who may act as a pool for transmission, especially if unvaccinated against COVID-19.
Although inequalities have been highlighted as important to COVID-19 infection risk, we did not find that demographic factors such as sex, deprivation and ethnicity, were associated with increased risk of hospitalisation or death. However, it may be that our study was underpowered to detect such differences as, for example, data on ethnicity was missing in 42%.
This study was representative of adult and paediatric patients with CHD admitted to specialist cardiac centres with COVID-19 throughout the first and second waves of the pandemic, and our findings are generalisable to this population. However, some adult centres did not participate so case ascertainment was incomplete. As all paediatric specialist cardiac centres in Great Britain took part, there was potential to ascertain all cases admitted to paediatric intensive care with COVID-19, however children with asymptomatic, mild and self-limiting illness or those managed at home, in emergency departments or district hospitals were less likely to have been reported to the study. In contrast, clinical thresholds to admit patients may have been lower, particularly during the first pandemic wave, reflecting medical caution at a time when understanding of disease COVID-19 progression in patients with CHD was lacking. Overall, disease burden is likely to have been higher than we ascertained and case-hospitalisation and case-fatality ratios lower.
Reliance on data collection from medical records also presented some limitations, such as missing data, and we were unable to ascertain longer-term outcomes, therefore, late complications and deaths may not have been captured. Although we recorded ventilation in the paediatric population, we were unable to use this as a measure of COVID-19 severity as clinicians reported that this was precipitated by COVID-19 in only half of cases, and by underlying CHD or other comorbidity in the remainder.
Future research
The evolution of vaccination, new variants and release from lockdown have had an impact on COVID-19 distribution, with subsequent waves affecting larger numbers of young patients. Further studies are required to explore the impact of these changes on populations affected by CHD.
Conclusions
Overall, in the paediatric population there was a low risk of death or severe disease secondary to COVID-19, but neonates and infants experienced higher rates of hospitalisation. In adults with CHD, risk of COVID-19 death and severe disease was related to older age, particularly in those with associated PAH, multiple comorbidities and genetic disease. Individual clinical management plans should be formulated based on these risk factors.
Footnotes
Contributors: PEFD, SC, AC, SVB-N, JS conceptualised and designed the study; SC, AC, HN, DW, MP, JA, PJ, BK, AH, UM, TR, JS, ZR, LR, OP and MNL collected the data; SC and RK analysed the data; PEFD, SC, AC, SVB-N and JS interpreted the data; SC wrote the manuscript; all authors reviewed and approved the manuscript for submission; PEFD was the guarantor of the study.
Funding: Open access costs for publication were funded by the British Congenital Cardiac Association.
Competing interests: None declared.
Provenance and peer review: Not commissioned; externally peer reviewed.
Supplemental material: This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.
Data availability statement
Data are available on reasonable request.
Ethics statements
Patient consent for publication
Not applicable.
Ethics approval
This study involves human participants and had ethical approval from the Health Research Authority, HRA/HRCW:20/HRA/2958. The Royal Brompton and Harefield NHS Foundation Trust was the co-ordinating centre. Approved without participant consent by the Health Research Authority due to it being a COVID-19 study using anonymised data.
References
- 1.Broberg CS, Kovacs AH, Sadeghi S, et al. COVID-19 in adults with congenital heart disease. J Am Coll Cardiol 2021;77:1644–55. 10.1016/j.jacc.2021.02.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Schwerzmann M, Ruperti-Repilado FJ, Baumgartner H, et al. Clinical outcome of COVID-19 in patients with adult congenital heart disease. Heart 2021;107:1226–32. 10.1136/heartjnl-2020-318467 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Fisher JM, Badran S, Li JT, et al. Characteristics and outcomes of acute COVID-19 infection in paediatric and young adult patients with underlying cardiac disease. Cardiol Young 2021;32:1261–7. 10.1017/S1047951121004029 [DOI] [PubMed] [Google Scholar]
- 4.Lewis MJ, Anderson BR, Fremed M, et al. Impact of Coronavirus disease 2019 (COVID-19) on patients with congenital heart disease across the lifespan: the experience of an academic congenital heart disease center in New York city. J Am Heart Assoc 2020;9:e017580. 10.1161/JAHA.120.017580 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Sabatino J, Ferrero P, Chessa M, et al. COVID-19 and congenital heart disease: results from a nationwide survey. J Clin Med 2020;9:1774. 10.3390/jcm9061774 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Cleary A, Chivers S, Daubeney PE, et al. Impact of COVID-19 on patients with congenital heart disease. Cardiol Young 2021;31:163–5. 10.1017/S1047951120004345 [DOI] [PubMed] [Google Scholar]
- 7.Davies P, Evans C, Kanthimathinathan HK, et al. Intensive care admissions of children with paediatric inflammatory multisystem syndrome temporally associated with SARS-Cov-2 (PIMS-TS) in the UK: a multicentre observational study. Lancet Child Adolesc Health 2020;4:669–77. 10.1016/S2352-4642(20)30215-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Ramcharan T, Nolan O, Lai CY, et al. Paediatric inflammatory multisystem syndrome: temporally associated with SARS-Cov-2 (PIMS-TS): cardiac features, management and short-term outcomes at a UK tertiary paediatric hospital. Pediatr Cardiol 2020;41:1391–401. 10.1007/s00246-020-02391-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Government of UK . Daily deaths with COVID-19 on the death certificate by date of death. 2022. Available: https://coronavirus.data.gov.uk/details/deaths [Accessed 02 May 2022].
- 10.Parsons A. UK 2020 composite index of multiple deprivation. 2021. Available: https://github.com/mysociety/composite_uk_imd [Accessed 02 May 2022].
- 11.Abel GA, Barclay ME, Payne RA. Adjusted indices of multiple deprivation to enable comparisons within and between constituent countries of the UK including an illustration using mortality rates. BMJ Open 2016;6:e012750. 10.1136/bmjopen-2016-012750 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Government of UK . NHS test and trace: what to do if you are contacted. 2022. Available: https://www.gov.uk/guidance/nhs-test-and-trace-how-it-works
- 13.Strah DD, Kowalek KA, Weinberger K, et al. Worse hospital outcomes for children and adults with COVID-19 and congenital heart disease. Pediatr Cardiol 2022;43:541–6. 10.1007/s00246-021-02751-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sachdeva S, Ramakrishnan S, Choubey M, et al. Outcome of COVID-19-positive children with heart disease and grown-UPS with congenital heart disease: a multicentric study from India. Ann Pediatr Cardiol 2021;14:269–77. 10.4103/apc.apc_134_21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Government of UK . Patients admitted to hospital. 2021. Available: https://coronavirus.data.gov.uk/details/healthcare [Accessed 17 Nov 2021].
- 16.British Congenital Cardiac Association . BCCA COVID-19 guidance for vulnerable groups with congenital heart disease. 2020. Available: https://www.bcca-uk.org/pages/news_box.asp?NewsID=19495710 [Accessed 02 May 2022].
- 17.Krishnan US, Krishnan SS, Jain S, et al. SARS-Cov-2 infection in patients with down syndrome, congenital heart disease, and pulmonary hypertension: is down syndrome a risk factor? J Pediatr 2020;225:246–8. 10.1016/j.jpeds.2020.06.076 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mackey K, Ayers CK, Kondo KK, et al. Racial and ethnic disparities in COVID-19-related infections, hospitalizations, and deaths: a systematic review. Ann Intern Med 2021;174:362–73. 10.7326/M20-6306 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
openhrt-2023-002356supp001.pdf (112.8KB, pdf)
Data Availability Statement
Data are available on reasonable request.