Population-based prevalence study of common congenital malformations of the alimentary tract and abdominal wall in Thailand: a study using data from the National Health Security Office

Introduction

Congenital malformations of the gastrointestinal tract and the abdominal wall are birth defects that usually require immediate surgical management. Atresia and stenosis are relatively common anomalies that may occur in any part of the gastrointestinal tract from the esophagus and the duodenum to the rest of the small intestine.1–3 Apart from alimentary tract anomalies, developmental defects involving the abdominal wall and diaphragm are among the common surgical conditions during the neonatal period.4 5 The population-based prevalence of these congenital anomalies varies among geographical areas and periods of study.6–8 Most of the reported prevalences of these diseases have been from studies in the Western world, which might not always represent the sizes of the same problem in other countries.

As an example of (ESO), a large population-based registry in European countries reported varying prevalences during 1987–2006, from 1.28/10 000 births to 4.55 births,9 while a population-based registry in Nagano, Japan, reported a prevalence of 1.97/10 000 births.10 The incidence of congenital duodenal obstruction (CDO) in the UK was estimated at 1.22 cases per 10 000 live births,11 while another study from Japan reported a prevalence of 0.83.10 An interesting disparity in reported prevalences was found in the case of gastroschisis (GAS), of which the prevalence reported in North America was around 3.4–4.18/10 000 births,5 12 while the figure was 2.69 in Brazil13 and 1.59 in Liaoning, China.14 Population-based prevalences of these common pediatric surgical problems have never been reported from Southeast Asian countries, and the subject is worth exploring, as this type of epidemiological data can be meaningful in resource planning.

In Thailand, although there is no official registry for congenital anomalies nor rare diseases, reimbursement data from the National Health Security Office (NHSO) covered nearly all cases treated in the public sector. This study aimed to assess the number of neonatal cases diagnosed with congenital anomalies of the alimentary tract, abdominal wall defects and congenital diaphragmatic hernia (CDH) to estimate the prevalences of these anomalies in the Thai population. In addition, crude mortality rates were calculated for each group of anomalies.

MATERIALS & METHODS

Results

A total of 2539 matched ICD-10 codes of interest were found in 2376 individuals, divided into 741, 583, 530, and 522 records in 2017–2020, respectively. The diagnostic group with the highest number of patients was HSCR (4.69 per 10 000 live births), followed by ARM (2.57 per 10 000 live births) and ESO (0.88 per 10 000 live births). The crude prevalences of the various diagnostic groups are shown in table 1. The overall male-to-female ratio was 1.55, with the highest proportion in HSCR at 1.76 (table 2). The distribution of cases among the 13 health regions in Thailand is shown in online supplemental figure S1. Note that although the number of study-relevant births in Bangkok during the 4-year analysis was 12.5% of the whole country, 23.6% of all cases were treated in the Bangkok region, followed by health region 12 (10.7%), which covers the southernmost part of the country.

There were 216 records that matched the ESO group; 88 were recorded as Q391 (esophageal atresia without fistula); 139 were Q392 (esophageal atresia with tracheoesophageal fistula); and 78 were the remaining types. CDO had 133 matched records, while 108 cases matched the code for atresia/stenosis of other small intestinal parts. Associated DS was found in 45.9% of the duodenal atresia/stenosis cases and 4.6% in cases of atresia/stenosis of other small intestinal parts.

HSCR (Q43.1) had the highest incidence among the anomalies in this study. The incidence rate of DS in HSCR was 5.1%, while the associated cardiac anomaly rate was 3.7%. Considering records of primary procedure of the individual patients, 94 cases had an International Classification of Diseases-9 (ICD-9) record of procedure codes related to a temporary ostomy (codes 4610, 4611, 4620, 4621, and 4652); 204 cases had the record of ‘4841: Soave submucosal resection of rectum’, 104 cases with ‘4849: other pull-through resection of rectum’, 42 cases with ‘4865: Duhamel resection of rectum’, and 30 cases with ‘4824: biopsy of rectum’

During the 4 years, there were 631 records that matched the set of ARM-related codes, which included 76 cases of Q42.0: congenital absence, atresia, and stenosis of the rectum with fistula; 35 cases of Q42.1: congenital absence, atresia, and stenosis of the rectum without fistula; 184 cases of Q42.2: congenital absence, atresia, and stenosis of the anus with fistula; 413 cases of Q42.3: congenital absence, atresia, and stenosis of anus without fistula; 13 cases of Q42.8: congenital absence, atresia, and stenosis of other parts of the large intestine; and 34 cases of Q42.9: congenital absence, atresia, and stenosis of the large intestine, part unspecified.

On analysis of inter-relations among the eight anomalies studied, high coincidences were found between esophageal atresia and ARM (10.65%) and between GAS and intestinal atresia (6.48%) (table 3). Interestingly, the study found a high incidence of ESO (7.25%) recorded in cases of duodenal atresia (table 3) and a relatively high incidence of HSCR in cases of INTES (3.70%). On Kaplan-Meier survival analysis, the survival probability at 1 year after the first admission was 93.52% (95% CI=92.42% to 94.44%) and the survival probability at 4 years was 91.85% (95% CI=90.54% to 92.99%). Patients with associated CHD had significantly poorer survival than those without CHD (online supplemental figure S2 B). In the mortality cases, half of all deaths occurred within 83 days after the first treatment, and 75% occurred within 237 days. The average length of hospital stay at the first admission was 12.1 days.

When analyzing the presence of Down syndrome and the presence of CHD with survival, both factors were risks of mortality at hazard ratios (HRs) of 2.54 (95% confidence interval (CI)=1.72 to 3.76) and 4.93 (95% CI=3.60 to 6.76), respectively (table 4 and online supplemental figure S2A,B). However, by multivariable Cox regression, DS was found to be a covariate of CHD in survival outcomes with adjusted HRs of 1.34 (95% CI=0.87 to 2.04) and 4.52 (95% CI=3.21 to 6.37), respectively. On univariate analysis performed in each anomaly group, DS had a significant association with the mortality status of patients with HSCR (p<0.001) and CDH (p=0.006), while CHD had significant associations with death in ESO (p=0.004), CDO (p=0.004), HSCR (p<0.001), ARM (p<0.001), OMP (p=0.026), and CDH (p<0.001) (table 4). Interestingly, in cases of HSCR, DS (adjusted hazard ratio (HR) 5.55, 95% CI=2.63 to 11.75) was found to outperform congenital cardiac anomalies (adjusted HR 2.11, 95% CI=0.88 to 5.10) in survival function (online supplemental figure S2C,D), while both factors had an independent association with survival in CDH (adjusted HR 19.90, 95% CI=1.66 to 239.23) for DS and 13.65 (95% CI=2.75 to 87.08) in cardiac anomalies.

Discussion

Recent progress in surgery and perioperative support has tremendously reduced mortality from surgical conditions in newborn patients, especially in high-resource countries.15–17 In Japan and the USA, overall mortality in surgical newborns has been reduced to less than 10% in previous decades.15 16 One of the key success factors in surgical quality improvement in these countries was the establishment of a national case-registry program that drove gradual but sustainable changes in treatment outcomes. In some situations, with limitations of neonatal surgical facilities, a number of cases are transferred to receive surgical care in other centers, and hospital-based case series might not reflect the picture of the whole country. In Thailand, such a registry for surgical neonates has not been established. However, as nearly all neonatal surgical patients are taken care of in a provincial referral center or a university hospital and the reimbursement of hospital expenses is submitted to the universal health coverage (UHC) program managed by the NHSO, mining of the discharge summary data can provide a crude picture of disease prevalence that can be used in health policy planning.18 Currently, the only data available through the NHSO are from the fiscal years 2017–2020.

According to our analysis of the 4 years of NHSO data, the number of neonates with common surgical conditions in Thailand ranged from 550 to 750 cases per year or around 9.5–11.5 cases per 10 000 births, with no significant trend of change over the period. Our prevalence of ESO at 0.9/10 000 births was much lower than those reported in Europe (1987–2006) at 2.43/10 000 births9 or in Japan (1993–2011) at 1.97/10 000 births,10 which might be explained by under-recording or miscoding. Lower-than-expected prevalence figures were also found in other anomalies in our study, except for HSCR (4.69/10 000 births) and ARM (2.54/10 000 births).19 20 We can explain the high prevalence of HSCR in the first year of the study by its requirement for multiple admissions, and patients born in the prior year would also be counted when they were admitted in the study period. This phenomenon occurred in a German study of ARM, in which they compared diagnostic-related groups and the ICD-10.6 A study from Australia found that when comparing hospital discharge data with a registered cohort, coverage accuracy varied among the types of anomalies. In that study, while abdominal wall defects had nearly complete concordance in coding, digestive system anomalies had an approximately 93% concordance rate, and there was a tendency of hospital records to overcount these anomalies.21 Comparing the number of abdominal wall defects with other registry-based reports,22–24 our number of cases was estimated to capture approximately 60%–80% of the true prevalence. As of 2021, the number of certified pediatric surgeons registered with the Thai Medical Council was 214.25 Based on the prevalence figures, each surgeon would have an opportunity to take care of fewer than five new cases of surgical neonates with major gastrointestinal anomalies per year, which indicates that neonatal surgical care in Thailand should be centralized to regional centers rather than disseminating to every general hospital, considering the fact that the case volume is essential for maintaining technical skills.

Our general mortality rate at 7.1% was comparable with other reports, and it was not surprising that relatively high mortality was found in ESO, CDO, and OMP. The mortality of diaphragmatic hernia in our study (8.5%) was relatively lower than those reported in the literature (10%–30%).26 A possible explanation was that our study did not include stillbirths and neonatal deaths at the local hospital before referral, without accurate diagnostic codes. The study also found that associated cardiac anomalies were the main factors associated with poorer survival outcomes in these birth defects and some other anomalies, including HSCR, ARM, and CDH. Overall, cardiac defects had a greater negative association with survival than DS, with the exception of HSCR. The strong association between DS and mortality in Hirschsprung disease might be explained by a higher incidence of enterocolitis and poorer continence in Hirschsprung’s patients with DS.27 The evidence emphasizes the significance of concomitant management of associated anomalies in a neonate with congenital anomalies involving the gastrointestinal tract, which might hamper a good outcome. Analyzing associations among groups of anomalies was also feasible with our derived data. Apart from a Vertebral anomalies, Anorectal malformations, Cardiac defects, TrecheoEsophageal anomalies, Renal anomalies, and Limb anomalies (VACTERL) association that was expected, associations between esophageal atresia and CDO and between INTES and GAS were also found.

The notable strength of our study was that the data were taken from a nationwide public reimbursement program (UHC) that covers at least 72% of the healthcare services in the country. Additionally, the data were not affected by referral status, and mortality data were not distorted by publication bias from institutes with better academic resources. However, as data from the UHC program are not a predefined registration program, the accuracy of recorded diagnosis was a limitation of our study. Induced abortion by fetal indications was not legal in Thailand until 2021, and the practice might be performed in cases of fetal anomalies under maternal health concerns. Although an exact figure of pregnancy termination after a prenatal diagnosis of congenital anomalies was not known, a survey by the Ministry of Public Health in 2004 reported an estimated induced abortion ratio of 19.5 per 1000 live births, which influences the prevalence of some congenital anomalies.28 29 The lower induced abortion rate in region 12, in which Muslim Thai predominates, might explain the higher prevalence of certain congenital anomalies in this region.

In conclusion, the study analyzed a nationwide database for the prevalence of common congenital gastrointestinal and abdominal wall anomalies. Although the study found lower-than-expected prevalences when compared with population-based registries from other countries, the mortality figures were comparable and were significantly affected by the presence of associated cardiac defects.