A nationwide comparative study of pediatric and adult intussusception in emergency departments during the COVID-19

Abstract

Intussusception is a common cause of acute abdominal pain among children presenting to emergency departments, but it is rare in adults, with distinct clinical and epidemiological patterns. However, large-scale comparative data are scarce, and the impact of infectious disease remains poorly understood. We analyzed nationwide emergency department data from South Korea (2019–2023) to compare cross-age clinical characteristics and investigate temporal trends associated with infectious disease incidence, using COVID-19 social distancing as a natural experiment. Among 14,723 cases, 80.8% were pediatric patients, peaking at ages 1–4, while adult cases were evenly distributed. Children presented to higher-level centers, whereas adults required hospitalization with higher clinical acuity. Vomiting and hematochezia predominated in pediatric patients, while abdominal pain and nonspecific symptoms like asthenia and dyspnea were more common in adults. During COVID-19 distancing, pediatric intussusception declined significantly, while adult cases remained stable. Correlation analysis revealed stronger associations of pediatric intussusception with respiratory viruses, particularly adenovirus and rhinovirus, than with enteric pathogens. Causal mediation analysis confirmed that respiratory viruses accounted for 65.6% of the pandemic-associated reduction, with adenovirus (68.3%) and rhinovirus (62.8%) contributing most. These findings highlight age-dependent differences and suggest temporal associations between respiratory viral infections and pediatric intussusception.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-18175-8.

Introduction

Intussusception is a medical emergency characterized by the invagination of one segment of the intestine into an adjacent segment, leading to bowel obstruction and, without prompt treatment, potential ischemia or necrosis¹. It occurs most commonly in infants and young children and typically presents with sudden abdominal pain, vomiting, and bloody stools². In contrast, adult intussusception is rare and is often associated with identifiable pathological lead points, such as neoplasms^3,4. These differences in etiology and clinical presentation underscore the importance of applying distinct diagnostic and management approaches to pediatric and adult patients.

Although these clinical differences between children and adults have been recognized, the literature remains heavily weighted toward pediatric populations, leaving adult presentations relatively underexplored. Some retrospective studies have described age-specific differences in clinical manifestations, diagnostic approaches, and underlying etiologies, suggesting that pediatric and adult intussusception may represent distinct clinical entities^5–7. However, most studies have been limited by small sample sizes and single-center data, with large-scale, population-based comparisons between pediatric and adult intussusception still lacking.

Beyond these epidemiological differences, the underlying pathophysiology of intussusception, particularly in pediatric populations, may involve infectious triggers. Viral infections may contribute to the development of pediatric intussusception, possibly through mechanisms such as lymphoid hyperplasia and altered intestinal motility^8,9. However, the role of specific pathogens remains controversial. Few studies have evaluated the relationship between seasonal infection trends and intussusception incidence at a national level, particularly under conditions of widespread behavioral and epidemiological change, such as those observed during the COVID-19 pandemic^10,11.

In the current study, we analyzed nationwide emergency department (ED) data from 2019 to 2023 using the South Korean National Emergency Department Information System (NEDIS). This study pursued two complementary objectives: a cross-age comparison of the clinical and epidemiological characteristics of intussusception, and an investigation into temporal trends and their associations with community-level infectious disease activity, taking advantage of the natural experiment provided by COVID-19–related social distancing. This research provides a comparative analysis of pediatric and adult intussusception, clarifies the age-specific clinical patterns, and utilizes pandemic-related changes in infection patterns to elucidate previously difficult-to-study relationships between seasonal infections and intussusception.

Methods

Study design and data sources

This retrospective cross-sectional study utilized two primary data sources. Although these databases were not linked at the individual patient level, we analyzed them in parallel to explore temporal associations between population-level infectious disease activity and intussusception incidence. The first was the NEDIS, a nationwide database managed by the National Emergency Medical Center and operated by the Ministry of Health and Welfare in South Korea. Since its launch in 2003, NEDIS has collected real-time administrative and clinical data from emergency departments nationwide, including all 411 EDs as of 2023, with ≥ 99.8% coverage maintained since 2019 (Supplementary Table 1). The data collected include patient demographics, visit characteristics, diagnoses, procedures, and disposition outcomes. This database has been used in various epidemiological and health services studies^12,13. Access to NEDIS data requires formal application through the National Emergency Medical Center (www.nemc.or.kr).

The second data source was the Sample Surveillance Data for Infectious Diseases, provided by the Korea Disease Control and Prevention Agency (KDCA). This national surveillance system compiles weekly reports from designated sentinel medical institutions on the incidence of notifiable infectious diseases, including respiratory viruses detected from respiratory specimens (human adenovirus, human bocavirus, enterovirus, influenza virus, parainfluenza virus, respiratory syncytial virus, rhinovirus, human metapneumovirus, and human coronavirus [common cold coronaviruses, not severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)]), and enteric pathogens from stool specimens, including viruses (rotavirus, astrovirus, norovirus, and sapovirus) and bacteria (Salmonella spp., Vibrio spp., enterotoxigenic Escherichia coli [ETEC], enteroinvasive Escherichia coli [EIEC], enteropathogenic Escherichia coli [EPEC], Campylobacter spp., Clostridium spp., Staphylococcus aureus, Bacillus spp., Yersinia spp., and Listeria spp.). Data were downloaded from https://dportal.kdca.go.kr/pot/index.do on May 28, 2025. The surveillance data used in this study are provided as Supplementary Data to ensure reproducibility. These data were used to explore temporal associations between seasonal infectious disease trends and intussusception incidence.

Study population

The study included all patients who presented to EDs across South Korea between January 1, 2019, and December 31, 2023, and who were diagnosed with intussusception (ICD-10 code K56.1) either at ED discharge or following hospital admission.

For comparative analysis, patients were categorized into pediatric (< 20 years) and adult (≥ 20 years) groups. An age cutoff of 20 years was selected to reflect health data reporting practices in South Korea, which use 5-year age intervals and commonly classify individuals younger than 20 years as part of the pediatric and adolescent population.

Social distancing periods during the COVID-19 pandemic

To evaluate the impact of COVID-19-related public health measures on intussusception trends, we defined three distinct time periods based on national COVID-19 distancing measures: pre-distancing (before March 22, 2020), during distancing (March 22, 2020, to April 18, 2022), and post-distancing (after April 18, 2022)^14,15.

During the social distancing period, South Korea implemented a multi-tiered framework ranging from Level 1 (“distancing in daily life,” allowing daily economic activities while adhering to individual quarantine guidelines) to Level 2 (“rapid spread of local transmission, beginning of nationwide community transmission”) and Level 3 (“nationwide massive community transmission,” staying at home and restricting all activities other than essential economic activities), with region-specific adjustments based on local case trends. Policies included school closures, restrictions on private and public gatherings (ranging from 2 to 10 people), limited operating hours for businesses, capacity restrictions in large venues (e.g., religious facilities and sports events), and mandatory contact tracing. While most distancing measures were lifted on April 18, 2022, indoor mask mandates remained in place until January 30, 2023. These defined periods were used as reference intervals in the interrupted time series analysis and for evaluating temporal trends in intussusception incidence.

Variables and definitions

To facilitate the interpretation of institutional characteristics and clinical severity, we applied standardized operational definitions. In South Korea, EDs are classified into three types under the Emergency Medical Service Act. Level I EDs correspond to regional emergency medical centers that offer comprehensive critical care services and advanced diagnostic capabilities. Level II EDs are local emergency medical centers with intermediate levels of resources and staffing. Level III EDs are local emergency facilities that provide only basic emergency care with limited infrastructure.

Clinical severity at ED presentation was assessed using the Korean Triage and Acuity Scale (KTAS), a validated five-level emergency department triage system implemented nationwide in Korea since 2016¹⁶. KTAS is based on the Canadian Triage and Acuity Scale, with modifications for the Korean healthcare environment. The system categorizes patients into five levels of acuity, with level 1 representing the most urgent condition (Resuscitation) and level 5 the least urgent (Non-Urgent). KTAS assessment utilizes 167 predefined chief complaint categories with first-order modifiers (pain, hemorrhagic disease, hemodynamic state, body temperature, respiration, level of consciousness, and accident mechanism) and complaint-specific second-order modifiers. The system incorporates age-appropriate physiological parameters to ensure accurate triage across all age groups.

From the NEDIS database, we extracted information on age, sex, date of visit, ED type (I, II, or III), KTAS level, time and day of visit, length of ED stay (defined as the time from arrival to final disposition, including discharge, admission, or transfer), and final disposition (discharge, admission, or transfer).

Outcomes

The primary outcome of this study was a comparison of the clinical and epidemiological characteristics of pediatric and adult patients diagnosed with intussusception. Secondary outcomes included an analysis of temporal trends in intussusception incidence during the COVID-19 pandemic and an evaluation of correlations between weekly intussusception cases and the incidence of infectious diseases.

Statistical analysis

Descriptive statistics were used to summarize the characteristics of the study population. Categorical variables are reported as frequencies and percentages, while continuous variables are expressed as medians with interquartile ranges (IQRs). For group comparisons, the chi-square test or Fisher’s exact test was used for categorical variables, whereas the Wilcoxon rank-sum test was applied for continuous variables. Pearson correlation analysis was performed to examine the relationship between weekly intussusception incidence and the incidence of infectious diseases.

To quantify the causal impact of COVID-19 social distancing measures on intussusception incidence, we conducted interrupted time series analysis using segmented regression models. The interrupted time series analysis model was specified as: Y_t = β_0 + β_1 × time_t + β_2 × intervention_t + β_3 × time_since_intervention_t + β_4 × post_intervention_t + β_5 × time_since_post_t + ε_t, where β_1 represents the baseline temporal trend, β_2 the immediate level change at intervention onset, β_3 the slope change during the intervention, β_4 the level change at intervention cessation, and β_5 the slope change post-intervention. Model assumptions, including autocorrelation and heteroscedasticity, were assessed using the Durbin-Watson test and Breusch-Pagan test, respectively.

Causal mediation analysis was performed to further assess whether reductions in specific infectious pathogens mediated the observed changes in intussusception incidence during the pandemic. Using the mediation package in R, we estimated the total, direct, and indirect effects of the COVID-19 intervention on intussusception through both group-level and pathogen-specific models. Multiple mediator models evaluated the joint contribution of respiratory and enteric pathogens, while single mediator models were used to quantify the effect of each pathogen individually. Bootstrapped confidence intervals (1,000 simulations) were used to assess the statistical significance of indirect effects, and the proportion mediated was calculated for significant pathways. Analyses were conducted separately for pediatric and adult populations.

A p value of less than 0.05 was considered statistically significant. All statistical analyses were conducted using R software (version 4.4.3; R Foundation for Statistical Computing, Vienna, Austria).

Ethics statement

This study was approved by the Institutional Review Board of Chungnam National University Hospital (IRB No. CNUH 2024-09-027), which also waived the requirement for informed consent owing to the retrospective nature of the study. All procedures were conducted in accordance with relevant guidelines and regulations.

Results

Demographic characteristics of pediatric and adult intussusception patients

To compare the distinct characteristics of pediatric and adult intussusception, we categorized patients into pediatric (< 20 years) and adult (≥ 20 years) groups. To characterize the baseline distribution, we analyzed the demographic profiles of all patients diagnosed with intussusception who presented to an ED between 2019 and 2023 (Fig. 1; Table 1). Of the 14,723 patients included, 80.8% (n = 11,895) were in the pediatric group and 19.2% (n = 2,828) were in the adult group. Among pediatric patients, most cases (58.1%) occurred in children aged 1–4 years, followed by infants aged 0–1 year (15.1%). The frequency of cases markedly declined with increasing age beyond early childhood. In contrast, adult cases were more evenly distributed across successive age decades, without a distinct peak. A significant male predominance was observed, more pronounced in the pediatric group than in the adult group, with 7,414 of 11,895 pediatric patients (62.3%) and 1,483 of 2,828 adult patients (52.4%) being male (p < 0.0001).

Fig. 1.

Age distribution of intussusception cases in pediatric and adult groups. The dotted vertical line indicates the age cutoff (20 years) used to distinguish pediatric from adult cases. Pediatric cases show a distinct peak in the 1–4 years age group, while adult cases are more evenly distributed across age groups.

Table 1.

Age distribution of intussusception patients presenting to the emergency department.

Age (year)	N (percent %)
Pediatric
Total	11,895 (80.8%)
0–1	2,225 (15.11%)
1–4	8,549 (58.07%)
5–9	862 (5.85%)
10–14	125 (0.85%)
15–19	134 (0.91%)
Adult
Total	2,828 (19.2%)
20–29	378 (2.57%)
30–39	356 (2.42%)
40–49	363 (2.47%)
50–59	479 (3.25%)
60–69	462 (3.14%)
70–79	351 (2.38%)
Over 80	439 (2.98%)

To further explore the appropriateness of the 20-year pediatric cutoff, we examined the correlation between each pediatric age subgroup and the adult population. All pediatric subgroups, including adolescents aged 10–19 years, exhibited weak correlations with adult intussusception trends (all r < 0.1) (Supplementary Table 2). While this exploratory analysis does not provide a definitive threshold, no clear trend toward stronger similarity with adults was observed in older pediatric groups. Considering that South Korea’s health data reporting practices adopt an age cutoff of 20 years for pediatric classification, as already specified in the Methods, and the consistent weak correlations with adult trends across all pediatric age subgroups, these findings support the use of the 0–19 age range as a practical and clinically meaningful definition of pediatric cases, consistent with national and international classification systems.

Emergency care utilization in pediatric and adult intussusception

We examined differences in emergency care utilization between pediatric and adult patients with intussusception (Table 2). Pediatric patients were more likely to present to level I EDs (54.7% vs. 35.3%), while adults more commonly visited level II and III centers. Adults exhibited higher clinical acuity, with 78.1% triaged at KTAS levels 1–3, compared to 65.9% in pediatric patients. Additionally, prolonged ED stays (≥ 6 h) were more common among adults (33.8% vs. 25.1%). To contextualize these findings, we compared them with national emergency medical statistics¹⁷, which show that adults generally have longer ED stays than pediatric patients (27.3% vs. 16.6% for stays ≥ 6 h). However, both adult and pediatric intussusception patients exceeded their respective population baselines (adult patients: +6.5% points; pediatric patients: +8.5% points), suggesting that disease-specific factors contribute to prolonged evaluation times beyond general age-related differences.

Table 2.

Characteristics and outcomes of pediatric and adult intussusception patients presenting to the emergency department.

	Pediatric (n = 11,895)	Adult (n = 2,828)	P value
Type of ED			< 0.0001
Level I	6,510 (54.73%)	997 (35.25%)
Level II	5,098 (42.86%)	1,493 (52.79%)
Level III	287 (2.41%)	336 (11.88%)
KTAS			< 0.0001
1–3	7,738 (65.86%)	2,156 (78.12%)
4–5	4,012 (34.14%)	604 (21.88%)
Duration of ED stay, h			< 0.0001
< 6	8,914 (74.94%)	1,872 (66.20%)
≥ 6	2,981 (25.06%)	956 (33.80%)
Disposition			< 0.0001
Discharge	5,909 (49.79%)	472 (16.76%)
Admission to general ward	5,459 (45.99%)	1,862 (66.10%)
Admission to ICU	25 (0.21%)	231 (8.20%)
Transfer	476 (4.01%)	252 (8.95%)

ED emergency department, KTAS Korean Triage and Acuity Scale, ICU intensive care unit.

Disposition outcomes reflected these differences, with pediatric patients more often discharged directly from the ED (49.8% vs. 16.8%), while adults had higher rates of general ward admission (66.1% vs. 46.0%) and ICU admission (8.2% vs. 0.2%). These findings may reflect differences in the natural course of pediatric and adult intussusception, with pediatric cases often involving ileo-ileal types that are managed with non-surgical strategies such as reduction, enabling direct ED discharge, whereas adult cases are more frequently associated with pathological lead points that necessitate hospital admission and prolonged evaluation.

Distinct clinical presentations and co-diagnostic profiles in pediatric and adult intussusception

Symptom patterns differed considerably between the two groups (Table 3). In pediatric patients, abdominal pain was the most common symptom (52.1%), although it was less frequent than in adults (74.7%), where it often presented as the dominant complaint. Pediatric patients more frequently exhibited classic features such as nausea and vomiting (16.5%), hematochezia (10.2%), and irritability (3.4%). In contrast, systemic symptoms, including asthenia (1.7%) and dyspnea (1.2%), were more commonly reported in adults, albeit infrequently overall. Co-diagnoses showed age-specific patterns that paralleled these symptom differences, with pediatric cases primarily associated with infectious gastrointestinal conditions and adult cases exhibiting greater heterogeneity (Supplementary Table 3).

Table 3.

Presenting symptoms in pediatric and adult intussusception.

Symptom	Pediatric (n = 11,895)	Adult (n = 2,828)	P value
Abdominal pain	6,196 (52.09%)	2,112 (74.68%)	< 0.0001
Nausea and vomiting	1,964 (16.51%)	86 (3.04%)	< 0.0001
Hematochezia	1,211 (10.18%)	70 (2.48%)	< 0.0001
Irritability	406 (3.41%)	0 (0.00%)	< 0.0001
Diarrhea	300 (2.52%)	31 (1.10%)	< 0.0001
Fever	406 (3.41%)	94 (3.32%)	0.8625
Melena	97 (0.82%)	48 (1.70%)	< 0.0001
Hematemesis	25 (0.21%)	18 (0.64%)	0.0006
Asthenia	15 (0.13%)	49 (1.73%)	< 0.0001
Dyspnea	3 (0.03%)	33 (1.17%)	< 0.0001

Temporal trends in intussusception incidence during the COVID-19 pandemic

Given the known impact of social distancing on infectious disease transmission and the established association between infections and intussusception¹⁸, we analyzed temporal trends in intussusception over a 5-year period, stratified by age group and social distancing phases related to the COVID-19 pandemic (Fig. 2A and Table 4). Among pediatric patients, the median daily case count peaked during the pre-distancing period (9.91; IQR, 6), declined significantly during the distancing period (4.57; IQR, 3; p < 0.0001), and partially rebounded post-distancing (6.62; IQR, 4; p < 0.0001). In contrast, adult intussusception cases remained relatively stable, with median daily incidences of 2.01 (IQR, 2) in the pre-distancing period and 1.93 (IQR, 1) during the distancing period and a slight but significant increase to 2.20 in the post-distancing period (IQR, 2; p = 0.008).

Fig. 2.

Temporal trends of daily intussusception cases across COVID-19 social distancing phases in pediatric and adult patients. (A) Observed daily cases. (B) Predicted daily cases estimated using interrupted time series analysis. Daily cases in pediatric (blue) and adult (red) patients are shown separately. Vertical dashed lines indicate the start (March 22, 2020) and end (April 18, 2022) of national COVID-19 social distancing policies, dividing the timeline into pre-distancing, during distancing, and post-distancing periods.

Table 4.

Trends in the incidence of intussusception before and after the COVID-19 pandemic, with and without social distancing.

	Pediatric (median, IQR)	Adult (median, IQR)
Pre-distancing	9.91 (6)***	2.01 (2)
During distancing	4.57 (3)	1.93 (1)
Post-distancing	6.62 (4)***	2.20 (2) †

IQR, interquartile range; FDR, false discovery rate.

***Significantly higher than during the distancing period (p < 0.0001).

^†Significantly higher than during the distancing period (p = 0.008).

All p values were adjusted using the FDR-corrected Wilcoxon rank-sum test.

To further examine the impact of social distancing measures, we performed interrupted time series analysis (Fig. 2B; Table 5). In the pediatric group, social distancing measures led to an immediate level drop of − 3.64 cases per day (95% CI: − 4.39 to − 2.89, p < 0.001), followed by a gradual increase during the distancing period (+ 0.008/day², p < 0.001). After restrictions were lifted, there was a significant level rise (+ 2.68/day, 95% CI: 0.34 to 5.01, p = 0.025) and continued upward trend (+ 0.013/day², p < 0.001). The model explained a substantial portion of the variance (R² = 0.331). In adults, no immediate change was observed at the onset of distancing (+ 0.22/day, 95% CI: − 0.10 to 0.53, p = 0.176). However, small but statistically significant slope increases were detected both during (+ 0.001/day², p = 0.022) and after (+ 0.002/day², p < 0.001) distancing, along with a level increase of + 1.17/day post-distancing (p = 0.020). The model fit was weak (R² = 0.019), indicating limited explanatory power in adults. These findings suggest that pediatric intussusception is more sensitive to community-level changes in infectious disease activity, whereas adult cases are comparatively unaffected.

Table 5.

Interrupted time series analysis of intussusception incidence before and after COVID-19 pandemic.

	Pediatric		Adult
	Estimate (95% CI)	P value	Estimate (95% CI)	P value
Baseline trend	-0.008 (-0.010, -0.006)	< 0.001	-0.001 (-0.002, 0.000)	0.017
Immediate change	-3.64 (-4.39, -2.89)	< 0.001	0.22 (-0.10, 0.53)	0.176
Slope change (during)	0.008 (0.005, 0.010)	< 0.001	0.001 (0.000, 0.002)	0.022
Level change (post)	2.68 (0.34, 5.01)	0.025	1.17 (0.18, 2.16)	0.020
Slope change (post)	0.013 (0.011, 0.016)	< 0.001	0.002 (0.001, 0.003)	< 0.001
Model R²	0.331		0.019

Trend and slope change parameters are expressed as cases/day².

CI, confidence interval; R², coefficient of determination.

Respiratory viral infections as a likely contributor to pediatric intussusception trends

To further explore the potential infectious etiology underlying these temporal patterns, we analyzed weekly case trends in infectious diseases reported by the KDCA (Fig. 3A and B, Supplementary Table 4). Although we previously visualized daily intussusception trends, weekly aggregations were employed for correlation analyses to align with the reporting frequency of national infectious disease surveillance data. Among pediatric cases, strong correlations were observed with respiratory viral infections (r = 0.755, p < 0.0001), followed by enteric bacterial infections (r = 0.393, p < 0.0001) and enteric viral infections (r = 0.304, p < 0.0001). In contrast, adult intussusception exhibited weak correlations with respiratory viral infections (r = 0.089, p = 0.1976), enteric bacterial infections (r = 0.181, p = 0.0034), and enteric viral infections (r = 0.044, p = 0.4767). These findings suggest that pediatric intussusception incidence patterns are more strongly temporally associated with seasonal infectious disease activity than adult intussusception.

Fig. 3.

Temporal patterns and correlation between intussusception and infectious diseases. (A) Weekly case counts of pediatric and adult intussusception by infectious disease category from 2019 to 2023. (B) Correlation between intussusception and infectious disease category in pediatric and adult patients. (C) Weekly case counts of individual respiratory viral infections. (D) Correlation between pediatric intussusception and individual respiratory viral infections. For panels (B) and (D), the circle size represents the Pearson correlation coefficient (r) while the color intensity indicates statistical significance (–log₁₀P). Vertical lines indicate the start and end dates of national COVID-19 social distancing policies in South Korea (March 22, 2020, and April 18, 2022, respectively).

Because correlation analysis does not establish causality, we employed causal mediation analysis to further validate these findings. Mediation analysis revealed that respiratory viral infections explained 65.6% of the overall mediated effect in the pediatric population (Table 6; model R² = 0.902). In contrast, enteric bacterial and viral infections played smaller roles. Among adults, the total effect of the pandemic on intussusception incidence was negligible, making mediation estimates less meaningful (model R² = 0.152).

Table 6.

Causal mediation analysis of pediatric and adult intussusception incidence by pathogen group.

	Pediatric		Adult
	Indirect effect	Proportion mediated (%)	Indirect effect	Proportion mediated (%)
Respiratory viruses	− 24.63	65.6	1.98	–*
Enteric bacteria	− 8.66	23.1	− 1.86	–*
Enteric viruses	1.36	3.60	0.23	–*
Direct pathway	− 5.6	14.9	− 0.45	–*
Total effect	− 37.54	100	− 0.11	100
Model R2	0.902	–	0.152	–

*Not applicable due to a negligible total effect size in adults (–0.11 cases/week).

R², coefficient of determination.

Adenovirus and rhinovirus as potential mediators during the COVID-19 pandemic

To further identify candidate pathogens, we analyzed weekly trends in specific respiratory viral infections (Fig. 3C and D, Supplementary Table 5). Among these pathogens, adenovirus (r = 0.802) and rhinovirus (r = 0.750) showed the strongest temporal correlations, followed by enterovirus (r = 0.573), bocavirus (r = 0.527), and parainfluenza virus (r = 0.510). Moderate correlations were observed for human metapneumovirus (r = 0.426), while weaker but still significant associations were found for respiratory syncytial virus (r = 0.205), influenza virus (r = 0.199), and human coronavirus (r = 0.182). All temporal associations were statistically significant (p < 0.01). Analysis of enteric pathogen infections revealed weaker but significant correlations, with astrovirus infections (r = 0.501) showing the strongest association among enteric viruses and Bacillus infections (r = 0.399) showing the strongest association among enteric bacteria (Supplementary Fig. 1, Supplementary Table 6).

To confirm the specific respiratory viral pathways identified through correlation analysis, we performed pathogen-level mediation analyses (Table 7 and Supplementary Table 7). Adenovirus and rhinovirus emerged as the most prominent mediators, with 68.30% and 62.80% of the pandemic-associated reduction in pediatric intussusception incidence mediated through each virus, respectively. Adenovirus showed an indirect effect of − 11.89 (95% CI: −21.54 to − 3.64), while rhinovirus demonstrated an indirect effect of − 13.96 (95% CI: −21.54 to − 7.22). Additional significant indirect effects were observed for enterovirus (− 30.83; 17.90%), bocavirus (− 31.61; 15.80%), parainfluenza virus (− 31.81; 15.30%), and human metapneumovirus (− 32.67; 13.00%). Notably, human coronavirus (excluding SARS-CoV-2) showed a considerable mediated effect (− 44.58; 18.80%), despite having a relatively modest correlation coefficient.

Table 7.

Causal mediation analysis of respiratory viral infections on pediatric intussusception incidence during the COVID-19 pandemic.

Virus	Indirect effect (95% CI)	Direct effect (95% CI)	Total effect (95% CI)	Proportion mediated (%)
Adenovirus	− 11.89 (− 21.54, − 3.64)	− 25.65 (− 32.11, − 18.39)	− 37.54 (− 42.12, − 32.72)	68.30
Rhinovirus	− 13.96 (− 21.54, − 7.22)	− 23.58 (− 28.06, − 18.63)	− 37.54 (− 41.88, − 32.73)	62.80
Enterovirus	− 30.83 (− 36.15, − 24.51)	− 6.70 (− 9.49, − 4.47)	− 37.54 (− 42.07, − 32.32)	17.90
Bocavirus	− 31.61 (− 37.01, − 26.20)	− 5.93 (− 8.23, − 3.84)	− 37.54 (− 42.25, − 32.54)	15.80
Parainfluenza	− 31.81 (− 37.08, − 25.23)	− 5.73 (− 10.12, − 2.74)	− 37.54 (− 41.93, − 32.30)	15.30
Human metapneumovirus	− 32.67 (− 38.98, − 26.04)	− 4.86 (− 7.65, − 2.31)	− 37.54 (− 42.34, − 32.52)	13.00
Respiratory syncytial virus	− 39.72 (− 45.13, − 33.49)	2.19 (0.34, 4.20)	− 37.54 (− 42.47, − 32.32)	5.80
Influenza virus	− 39.90 (− 46.79, − 32.18)	2.37 (− 1.08, 5.49)	− 37.54 (− 42.38, − 32.60)	6.30
Human coronavirus *	− 44.58 (− 50.36, − 38.18)	7.04 (4.07, 10.22)	− 37.54 (− 42.81, − 32.65)	18.80

* Excluding SARS-CoV-2, CI, confidence interval.

Taken together, these results confirm and extend the correlation findings, providing statistical evidence that reductions in the circulation of specific respiratory viruses, particularly adenovirus and rhinovirus, contributed to the observed decline in pediatric intussusception during the pandemic period.

Discussion

Intussusception is a well-recognized cause of acute abdominal pain across all age groups, yet its clinical presentation, underlying etiology, and urgency of care differ markedly between pediatric and adult patients¹⁹. While pediatric intussusception has been extensively studied, adult cases remain comparatively underexplored, and large-scale, population-based comparisons across age groups are limited. Our study addresses this gap by analyzing nationwide ED data from over 14,000 patients, providing population-level insights into age-specific patterns.

We identified clear epidemiological and clinical differences between pediatric and adult intussusception. Pediatric cases were concentrated in early childhood and typically presented with the classic triad of abdominal pain, vomiting, and hematochezia, consistent with previous studies^20,21. In contrast, adult cases were more evenly distributed across age groups and frequently involved isolated abdominal pain or nonspecific systemic symptoms such as asthenia or dyspnea. Notably, a greater proportion of adult patients were triaged with higher acuity and required hospitalization, contradicting the widespread view that adult intussusception is generally mild³. Interestingly, while adults more often presented and were admitted to lower-tier EDs, pediatric patients typically visited tertiary centers and were discharged shortly after evaluation. One plausible explanation is that level I EDs are typically located in tertiary hospitals equipped with pediatric subspecialties and advanced imaging services. This discrepancy may reflect underlying differences not only in disease characteristics but also in age-specific healthcare utilization patterns and referral system dynamics.

Our findings reveal a notable decline in hospitalization rates for pediatric intussusception in Korea. While a previous Korean study from the early 2000s reported that 82.5% of pediatric intussusception cases required hospitalization with a median stay of 3.0 days²², our recent ED data (2019–2023) showed that 49.8% of pediatric patients were discharged directly from the ED. This substantial decrease in admission rates likely reflects evolving clinical approaches influenced by international evidence demonstrating equivalent safety outcomes between ED management and inpatient observation^23–26. Notably, the nationwide Korean study also reported acceptable recurrence rates of 10.6% post-discharge and 15.0% overall²⁷, supporting the feasibility of outpatient management. However, no comprehensive studies have systematically evaluated this trend in hospitalization patterns across Korean emergency departments, representing an important area for future research to ensure optimal patient safety and healthcare efficiency.

The pathogenesis of intussusception, particularly in children, remains incompletely understood. Numerous studies have suggested that viral infections may act as important triggers by inducing lymphoid hyperplasia in the intestinal wall, especially in young children^8,9. The rotavirus vaccine has long been suspected of being associated with intussusception, but this relationship remains controversial^28–30. In addition to rotavirus, the potential involvement of other viruses, such as adenovirus, enterovirus, and HHV-6, has been extensively investigated. However, the precise etiological roles of these viruses remain unconfirmed. Differences in study findings may stem from the heterogeneity of specimen types across studies: some focused exclusively on stool samples^18,31, while others examined multiple specimen types, including blood, respiratory samples, stool, and tissue²⁵. Differences in the types of specimens collected (e.g., stool vs. respiratory or tissue samples) can lead to variability in virus detection rates, which in turn complicates the interpretation and comparison of findings across studies^18,25,31.

To further explore this hypothesized viral contribution, we analyzed temporal trends in intussusception incidence during the COVID-19 pandemic. Our analysis revealed a marked decline in pediatric intussusception during periods of social distancing, consistent with several recent studies^11,26,32 that reported reductions ranging from 36 to 63% during pandemic restrictions. The consistent magnitude of the decreases across multiple continents suggests that viral infections may contribute more substantially to pediatric intussusception than previously recognized.

To identify which specific pathogens might be responsible for this viral association, we directly compared intussusception trends with infectious disease activity and found that, intriguingly, the correlations were stronger with respiratory viral infections than with enteric infections. Among respiratory viruses, adenovirus and rhinovirus exhibited the strongest associations with pediatric cases. This adenovirus association is corroborated by a Korean study showing that nonenteric adenoviruses, primarily respiratory pathogens, were significantly more prevalent in pediatric intussusception cases than enteric adenoviruses (40.5% vs. 1.6%) and exhibited strong seasonal correlations with disease incidence³³. Additional support comes from case-control studies demonstrating that non-enteric adenovirus was detected in 78.9% of intussusception cases, particularly adenovirus C, which showed significant associations with intussusception in children younger than 3 years^34,35. Although the association between rhinovirus and intussusception has been relatively underexplored, rhinovirus was detected in approximately 23% of intussusception cases in one study²⁵. Our findings suggest that rhinovirus may be an underrecognized contributor to pediatric intussusception pathogenesis, alongside the well-established adenovirus association.

These findings suggest that respiratory viruses may play a more prominent role in the development of pediatric intussusception than enteric viruses, consistent with previous reports^11,18. In contrast, adult cases remained relatively stable throughout the pandemic and showed no clear associations with infectious disease activity. This divergence supports the interpretation that pediatric and adult intussusception may be driven by distinct pathophysiological mechanisms. In children, viral or immune-mediated factors may be more relevant, whereas, in adults, structural or neoplastic causes are more likely predominant. Although adult cases showed higher rates of gastroenteritis co-diagnosis, the lack of a temporal association with infectious disease patterns during the COVID-19 period suggests that any infectious involvement in adults may be secondary to an underlying structural pathology rather than the primary etiological factor observed in pediatric cases. While these findings do not establish individual-level causality, they underscore the need for mechanistic and patient-level investigations to clarify the role of respiratory viruses in pediatric intussusception. These findings advance our understanding of age-specific etiologies and underscore the importance of incorporating infectious disease surveillance into pediatric intussusception research and prevention strategies.

Despite the strengths and clinical relevance of this large-scale, nationwide analysis, several limitations should be acknowledged. The use of administrative data restricted access to detailed clinical variables, such as laboratory findings, imaging results, and treatment courses, limiting the evaluation of disease severity and outcomes. Notably, information on the diagnostic modality used in each case was not available, which precluded differentiation between anatomical subtypes of intussusception, such as ileo-ileal versus ileo-colic. This limitation may partly explain the relatively high discharge rate, particularly among pediatric patients, because transient, self-resolving ileo-ileal cases may have been included.

Furthermore, our study did not examine the presence of viral pathogens in stool samples from intussusception patients, which hindered the assessment of direct pathogen–host interactions. The ecological design of the correlation analyses also precludes the inference of causality between infection patterns and intussusception incidence. Moreover, as the study population was limited to individuals presenting to EDs, our findings may not fully represent the broader clinical spectrum or less acute presentations of intussusception. While our time series correlation analyses revealed strong temporal associations between pediatric intussusception and respiratory viral infections, these ecological correlations should be interpreted cautiously, as they do not establish individual-level causal relationships.

Conclusions

Our study provides population-level insights into age-specific patterns of intussusception using nationwide ED data. We observed distinct epidemiological, clinical, and healthcare utilization differences between pediatric and adult cases. During COVID-19 social distancing, pediatric intussusception incidence declined significantly while showing stronger temporal associations with respiratory viral infections, particularly adenovirus and rhinovirus. Causal mediation analysis revealed that respiratory viruses accounted for 65.6% of the pandemic-associated reduction, suggesting that respiratory viruses may play a more prominent role in pediatric intussusception pathogenesis than enteric viruses. However, these ecological correlations should be interpreted with caution as they do not establish individual-level causal relationships. These findings indicate that pediatric and adult intussusception may involve different underlying mechanisms, warranting further investigation through individual-level studies to confirm these population-level associations.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

D.-U.K. and H.G. contributed equally to this work. D.-U.K. and H.G. conducted data analysis and interpretation. E.H.C. conceptualized and supervised the study. D.-U.K. drafted the manuscript, and all authors critically revised and approved the final version.

Data availability

NEDIS data require formal application through the National Emergency Medical Center (www.nemc.or.kr). KDCA surveillance data used in this study are provided as Supplementary Data. Current surveillance data are available at https://dportal.kdca.go.kr/pot/index.do, though historical data organization may vary. All data used in this study are referenced in this published article and its Supplementary Information files.

Declarations

Competing interests

The authors declare no competing interests.