The prevalence and survival of children with congenital septal defects in the UK using CPRD

Onuwe, Toluwalope Yetunde (2022) The prevalence and survival of children with congenital septal defects in the UK using CPRD. MPhil thesis, University of Nottingham.

PDF (Thesis - as examined) - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Available under Licence Creative Commons Attribution.
Download (3MB) | Preview



Congenital heart disease (CHD) is a major contributor to infant mortality worldwide, accounting for a third of total congenital defects and a fifth of total global early neonatal mortality. Congenital septal defects (SD) are the most common CHD. This study assesses the prevalence and survival of children diagnosed with congenital septal defects in the UK.


All children who survived to at least the first 30 days after birth and registered with a GP within their first year of birth diagnosed with any CHD at any time in England between January 1998 and June 2017 were extracted from Clinical Practice Research Datalink (CPRD). Similar records for children diagnosed with other congenital anomalies were also obtained as well as children without any congenital anomaly constituting as the control group. These records were matched with the Office for National Statistics (ONS) death registration data, and the Index of Multiple Deprivation (IMD) data to allow assessment of death records and deprivation records, respectively. Prevalence risk ratios were calculated according to specific sociodemographic factors such as gender, year of birth, ethnicity, socioeconomic status, and geographical location for SD and its variants. Mortality risk ratios were calculated by gender, ethnicity, and socioeconomic status for SD and its variants. Also, in the absence of Hospital Episode Statistics (HES) data, I assessed the recording of CHD-related procedures/interventions in children’s GP records as a measure of disease severity within the CPRD-based birth cohort used in this thesis.


The prevalence of any CHD was 86.1 per 10,000 live GP-registered births (95% CI, 83.7 – 88.4). SD accounted for 53.4% of the total CHD within this population and the overall prevalence of SD was 48.64 per 10,000 live GP-registered births (95% CI 46.90-50.44). Of all SD diagnosis identified, 59.1% had a VSD diagnosis (35.50 per 10,000 live GP-registered births, 95% CI 34.02-37.04), 28.8% ASD (17.31 per 10,000 live GP-registered births, 95% CI 16.29-18.40), 6.5% for ToF (3.91 per 10,000 live GP-registered births, 95% CI 3.44-4.45), 0.3% for AoSD (0.17 per 10,000 live GP-registered births, 95% CI 0.09-0.31), 3.2% for AVSD (1.93 per 10.000 live GP-registered births, 95% CI 1.61-2.32), and 2.1% for unspecified SD (1.28 per 10.000 live GP-registered births, 95% CI 1.02-1.60). For any SD, gender was significantly associated with occurrence, with the prevalence in females being 15% more than males (PR (Male-Female) 1.15, 95% CI 1.07-1.24), however, there was considerable variation across specific SD variants. Whilst there was no evidence of an association with the year of birth, ethnicity, or socioeconomic status on the prevalence of overall SD, specific sub-analysis of the SD variants showed some associations. The prevalence of SD and its variants varied by geographical regions in England with the highest prevalence of any SD observed in the East Midlands.

The total number of deaths due to CHD from 30 days after birth to 19 years between January 1998 and June 2017 was 166 and the mortality rate was 5.6 (95% CI, 4.8-6.5) per 1,000 person-years. By SD variants, children with Aortopulmonary septal defect (54.9; 95% CI 13.7-219.6), Atrioventricular septal defect (7.2; 95% CI, 3.0-17.3), and Tetralogy of Fallot (8.1; 95% 4.6-14.3) had the highest mortality rates compared to other defects, although small numbers of children resulted in large confidence limits for some of these estimations.

The hazard of death for children with any septal defect was over 16-fold higher than for children without a CA (HR 16.2; 95% CI, 12.8-20.6); aHR 16.4, 95% CI, 13.0-20.9, p<0.001). Of this, children with Aortopulmonary septal defect had the highest hazard of death with a 160-fold increase compared to children without a CA (HR, 164.6, 95% CI, (41.1-659.5); aHR 159.3, 95% CI, 39.7-641.4, p<0.001) and was lowest for children with VSD (HR, 15.5, 95% CI, (11.7-20.5); aHR 14.9, 95% CI, 12.0-21.0, p<0.001). For overall septal defects, whilst there was no difference in the hazard of death or mortality rate by ethnicity, there was a significant association with socioeconomic status largely driven by ASD. Mortality rates by age groups between 0 and 9 years were also considered; whilst there was substantial variation within the various defects and age groups, there was a 12-fold increased risk of post neonatal death for all SD compared to children with no CA across age groups.

Of the total SD cases in this dataset, 5% were recorded to have had an intervention within the first year of life. Of the SD sub variants, 4% of each of the total VSD and ASD had an intervention recorded in population. The proportion of interventions was higher for the other SD sub variants with ToF, AVSD and AoSD at 29%, 36% and 40% respectively. Of the 381 CHD patients who had a recording of at least one CHD-related intervention, only eight (8) patients died. The subsequent analysis of the severity of SD variants in relation to mortality was therefore precluded as for all variants the number of deaths recorded was less than 5.


Using a study population derived through data linkage between CPRD, ONS, and IMD in England, this study reinforces broadly that AoSD, ToF, AVSD whilst being the least prevalent SD, constitute the highest mortality variants of SD whilst VSD and ASD (the more prevalent forms) are not as lethal. Regardless, all SD variants considered had higher mortality than children without SD. Considerable effort is required to further improve the survival rates of children with these defects and ultimately improve the prognosis for children diagnosed with these defects. This study further raises the need for the consideration and development of region-specific and ethnicity-specific strategies to improving diagnosis and disease outcomes for children with a septal defect in England. As the dataset was not linked to HES, the ability to adequately assess the severity of septal defects in the study population was impaired. Using the limited codes for recording of CHD-related procedures within the GP-based dataset as a measure of severity unmasked the likely under-reporting of hospital events or procedures as children who should have received interventions based on their diagnosis did not always have this recorded in their GP record. Further studies incorporating the available information from HES are required to effectively determine the severity of septal defects and its variants especially with regards to how it affects mortality and mortality

Item Type: Thesis (University of Nottingham only) (MPhil)
Supervisors: Tata, L.J.
Gibson, J.E.
Keywords: Congenital Heart Defects, Congenital Septal Defects, CPRD
Subjects: W Medicine and related subjects (NLM Classification) > WG Cardiocascular system
Faculties/Schools: UK Campuses > Faculty of Medicine and Health Sciences > School of Medicine
Item ID: 69436
Depositing User: Onuwe, Toluwalope
Date Deposited: 01 Aug 2022 04:40
Last Modified: 01 Aug 2022 04:40

Actions (Archive Staff Only)

Edit View Edit View