Risk factors and predictive models for early recurrent intussusception in children: a retrospective cohort study

Introduction

Intussusception is one of the most common abdominal emergencies in children. Recurrent intussusception refers to the recurrence of intussusception in children who have been successfully reset by enema therapy or surgery. Recurrence of intussusception is fairly common regardless of the method of reduction (1), with incidence ranging from 2% to 20% (2-5).

Intussusception recurrence can be divided into early recurrent intussusception (ERI) and late recurrence intussusception (LRI). Currently, there is no clear time limit for what early recurrence is. Some surveys found that nearly 50% of recurrent intussusception occurred within the first week after the first reduction (3), known as ERI, while others considered recurrence within 24 hours to be an early recurrence, on the other hand, some researchers still defined early recurrence as 48 or 72 hours (6-9). Research by Wang et al. has shown that 92.1% of early relapses occur within 48 hours (10).

Some cases of intussusception can recur several times in a short period, which seriously affects the physical and mental health of the patient. Previous studies have failed to identify consistent risk factors for intussusception recurrence, and risk factors for ERI in children have rarely been reported. Therefore, we attempted to investigate the risk factors for ERI in children and build predictive models to help with the individualized judgment of clinicians. We present this article in accordance with the TRIPOD reporting checklist (available at https://tp.amegroups.com/article/view/10.21037/tp-23-269/rc).

Methods

Study design and population

The purpose of this study was to retrospectively investigate the data on intussusception in children admitted to the First Affiliated Hospital of Shantou University Medical College, China. This retrospective study has been registered with the Chinese Clinical Trial Registry (No. ChiCTR2200063862). This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The research protocol was approved by the Ethics Committee of the First Affiliated Hospital of Shantou University Medical College (No. B-2022-163). The informed consent was waived for this retrospective study.

Our study cohort was comprised of 983 children admitted to the First Affiliated Hospital of Shantou University Medical College from January 2011 to December 2021 due to pediatric intussusception. The inclusion criteria were as follows: (I) a previous diagnosis of intussusception that had been successfully reduced by enema therapy or surgery; (II) a confirmed diagnosis of recurrent intussusception based on the patient’s clinical manifestations, physical examination, color ultrasound, and other tests, with other causes excluded; and (III) the recurrence having occurred within 48 hours of the first occurrence. The exclusion criteria were as follows: incomplete clinical data. After applying all the inclusion criteria and exclusion criteria, the final study cohort consisted of 869 cases.

Data collection

The data collected included 11 factors before treatment, including the child’s sex, age, onset quarter, duration of symptoms, clinical manifestations (abdominal pain or crying, vomiting, bloody stool, fever), white blood cells (WBCs), lymphocyte ratio, and monocyte ratio.

Definitions

Existing articles have different definitions of early recurrence time. Moreover, children treated for intussusception in this study were generally discharged after 48 hours of observation following reduction. Thus, this study defined recurrence within 48 hours as early recurrence. Referring to the “Reference intervals of blood cell analysis for children” (11), we define WBCs, lymphocyte ratio, and monocyte ratio as high, normal, and low respectively.

Statistical analysis

Clinical data from the ERI group and the no relapse intussusception (NRI) group were statistically analyzed, and categorical variables were expressed in terms of frequency and percentage. The χ² test or Fisher’s exact test was used for univariate analysis of related variables. Statistically significant risk factors in the univariate analysis were then subjected to logistic multifactorial analysis of independent risk factors for ERI (12). Patients were randomly divided into a training set and a validation set in a 7:3 ratio, which was used for model development and validation, respectively. Based on the results of logistic multivariate analysis, we constructed a nomogram to predict ERI. We plotted a calibration curve on this diagram. Finally, we used the receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA) to detect the performance of the predictive model. P<0.05 was considered statistically significant.

All analyses were performed with the statistical software packages R (https://www.r-project.org/; The R Project for Statistical Computing, Vienna, Austria) and Free Statistics software version 1.3.

Results

General characteristics

The characteristics of patients in each study group are shown in Figure 1. Among the children who did not relapse, 104 (13.21%) underwent surgical treatment, 83 (10.55%) underwent simple manual reduction, and 9 children (1.14%) had intestinal necrosis; among the children with early recurrence, 2 (2.44%) underwent surgical treatment, 1 (1.22%) underwent simple manual reduction, and 0 children had intestinal necrosis. Children with early recurrence had a smaller probability of undergoing simple manual reduction or experiencing intestinal necrosis. Table 1 shows the age distribution of the 869 children with intussusception included in this study. Intussusception and NRI were more likely to occur in children <1 year old of age, and the incidence of both decreased with age. ERI tended to occur in children aged ≥2 and <3 years old. The incidence of ERI increased with age in children aged <3 years old and decreased in children ≥3 years old.

Figure 1 Specific groups and treatment methods for children with intussusception.

Univariate analysis

Of the 869 children with intussusception, 82 had an ERI, and 787 had NRI. Univariate analysis of 11 factors showed that age, abdominal pain or crying, vomiting, bloody stools, and monocyte ratio were potential risk factors for ERI (P<0.05) (Table 2). Compared with the NRI group, children with ERI tended to be older than 1 year old, and had more abdominal pain or crying symptoms, less vomiting and bloody stool, and higher monocytes.

Multivariate analysis

Multivariate analysis showed that age [odds ratio (OR), 7.67; 95% confidence interval (CI): 2.27–25.89; P=0.001], vomiting (OR, 0.17; 95% CI: 0.09–0.32; P<0.001), bloody stool (OR, 0.14; 95% CI: 0.03–0.63; P=0.01), and monocyte ratio (OR, 9.52; 95% CI: 2.14–42.25; P=0.003) were independently associated with the clinical endpoints. Age, vomiting, bloody stool, and monocyte ratio were independent risk factors for ERI in children (Table 3). The results of the multivariate analysis are presented in a logistic regression forest plot (Figure 2).

Figure 2 Logistic regression forest plot. The blue line represents OR >1 and the red line represents the OR <1. **, 0.001≤P<0.05; ***, P<0.001. OR, odds ratio.

Establishment and validation of predictive models

A nomogram (Figure 3) was constructed based on four independent predictors of ERI: age, vomiting, bloody stool, and monocyte ratio. By adding the scores for each selected variable in the nomogram we could estimate the likelihood of ERI in a single patient.

Figure 3 Nomogram predicting the occurrence of ERI in children. ERI, early recurrent intussusception.

Predictive accuracy and net benefit of the nomogram

The calibration curve showed the relationship between the risk of ERI in children and the actual risk of occurrence. In our cohort, the calibration curve was close to the ideal diagonal (Figure 4). Based on the diagnostic score of the disease, the performance of the predictive model was detected using a ROC curve with area under the curve (AUC) value of 0.883 (95% CI: 0.846–0.920) (Figure 5). In addition, DCA showed significant net benefits for the predictive model (Figure 6). These data show that our nomogram has important potential in clinical decision-making.

Figure 4 Calibration curve showing the consistency of the risk of ERI in children with the actual risk of occurrence. ERI, early recurrent intussusception.

Figure 5 ROC curve of the established predictive model. The AUC of this model was 0.883 (95% CI: 0.846–0.920). AUC, area under the curve; ROC, receiver operating characteristic.

Figure 6 DCA in the prediction of ERI in children. DCA, decision curve analysis; ERI, early recurrent intussusception.

Discussion

In previous studies, the incidence of ERI in children ranged from 5% to 12.5% (10,13,14), in agreement with this range, the incidence of pediatric ERI in this study was 9.44%. The mortality rate of intussusception depends heavily on the availability of appropriate treatment in a timely manner, ranging from less than 1% in developed countries to as high as 10% in low-income countries (15). ERI in children recurs within a very short period of time. Early discharge followed by readmission can delay treatment and increase the hassle of traveling back and forth, causing problems for doctors and parents. Early prediction of ERI is therefore important in determining the length of hospital stay and reducing mortality from intestinal obstruction. Although modern medicine has made great progress, there is still insufficient attention to the recurrence of intussusception. There has been no consistent report on the recurrence of intussusception in previous studies, and the characteristics of ERI are rarely reported.

Our study retrospectively analyzed 11 factors—including sex, age, onset quarter, duration of symptoms, clinical manifestations (abdominal pain or crying, vomiting, bloody stool, fever), WBCs, lymphocyte ratio, and monocyte ratio before treatment in 82 children with ERI and 787 children with NRI. The results of this study show that age, vomiting, bloody stool, and monocyte ratio are independent risk factors for relapse of intussusception in children.

Age >1 year has been shown to be a risk factor for early recurrence (16,17), and our findings are consistent with this. In this study, we found that ERI incidence in children aged 0–3 years increased with age, but then decreased in children older than 3 years. Research by Stewart et al. has shown that the development of the pediatric gut microbiota goes through three distinct stages: the developmental stage before 14 months of age, the transitional stage of 15 to 30 months of age, and the stable stage of 31 to 46 months of age (18). During this process, the composition of the microbiome changes rapidly until about the age of 3 years when the diversity and complexity of the microbiome resemble that of adulthood (19-21). This coincidental timing begs the question of whether the occurrence of pediatric ERI is related to the transformation of the intestinal flora. A study has shown that the gut microbiota is essential for the development of gut-associated lymphoid tissue, including Peyer’s plaque and mesenteric lymph nodes, and contributes to the local and systemic immune systems (22). In intussusception reduction surgery, enlarged Peyer’s plaques and mesenteric lymph nodes are common (10). Therefore, we speculate that, from the age of 0–3 years, when the pediatric intestinal flora is in a period of dramatic change, the development of Peyer’s plaque and mesenteric lymph nodes can be easily perturbed, causing Peyer’s spots and mesenteric lymph nodes to enlarge and affecting the systemic immune system. Enlargement of these structures can disrupt intestinal peristalsis patterns, leading to ERI. After the age of 3 years, the intestinal flora tends to stabilize, and the incidence of pediatric ERI gradually decreases.

Vomiting and bloody stool are common clinical symptoms of intussusception in children, often caused by intestinal obstruction, ischemia, and necrosis. In previous studies of ERI, children with early intussusception were found to have a significantly lower rate of vomiting and bloody stool than children who did not have a recurrence (14,23). In the current study, vomiting and bloody stool were protective factors for pediatric ERI, and children with ERI exhibited less vomiting and fewer bloody stools. We hold the opinion that vomiting and bloody stools are unlikely to manifest in the early stage of intussusception onset when the irregular activity of the intestine is more intense and more likely to lead to ERI.

Our results show that the monocyte ratio is an independent risk factor for pediatric ERI and that children with a high monocyte ratio are more likely to relapse early. Monocytes are circulating WBCs that are essential for immune defense against viral, bacterial, and fungal infections (24,25). Since there is no clear starting point for intussusception, the etiology of primary intussusception remains inconclusive. At present, the main causes associated with primary intussusception in children are anatomical factors, viral infections, gastrointestinal dysfunction, and weather factors (26-28). Viral infection causes mesenteric lymphadenopathy and an elevated monocyte ratio, which may explain the elevated monocyte ratio in pediatric ERI patients. The intestinal flora is closely related to the immune status of the gut. In a study of type 1 diabetes (T1D), a significant T1D-specific inflammatory state was found, mainly manifested by increased monocyte/macrophage lineage infiltration. The expression of T1D inflammation-specific genes was associated with the abundance of specific bacteria in the duodenum (29). Do changes in the microbiome in the gut affect the proportion of monocytes? Increased intestinal mucosal macrophage infiltration may damage the intestinal mucosal barrier (29). The intestinal microbiota can translocate through the damaged intestinal mucosa to Peyer’s plaque and mesenteric lymph nodes, causing enlargement of Peyer’s plaque and mesenteric lymph nodes (30). Enlarged Peyer’s spots and mesenteric lymph nodes could also be seen in children with intussusception who required a surgical reduction in this study. Therefore, we propose that the intestinal flora of children with ERI changes, causing an increase in intestinal monocytes and macrophages, resulting in damage to the intestinal mucosal barrier. The gut microbiome affects the systemic immune system through the damaged intestinal mucosal barrier so that monocytes further increase or reach Peyer’s plaque and mesenteric lymph nodes, causing Peyer’s spots and mesenteric lymph nodes to enlarge, and eventually cause ERI. This is consistent with our inference above regarding the development of the gut microbiota as a function of age, but the specific mechanisms require further study.

Based on the independent risk factors for ERI found in this study, we established a predictive model for pediatric ERI. Age >1 year is positively correlated with the onset of pediatric ERI, a high proportion of monocytes is the most significant independent risk factor predicting ERI risk in children, and vomiting and blood are protective factors for pediatric ERI. We plotted a calibration curve that shows that the actual probability is substantially consistent with the predicted probability. At the same time, we used the ROC curve to detect the performance of the predictive model with an AUC value of 0.883 (95% CI: 0.846–0.920), and the calibration curve was close to the ideal diagonal line. In addition, the DCA showed that the model had significant net benefits. These results show that our predictive models have good accuracy and stability.

However, our study had several limitations. First, in this retrospective study, potentially meaningful predictors such as “ultrasonography (USG), C-reactive protein, birth pattern, birth weight, feeding pattern, antibiotic use, etc.” were not evaluated due to a lack of data. In further studies, we will evaluate more potential indicators and build more accurate prediction models for pediatric ERI based on clinical features. Second, because the study is retrospective, there is inevitably some degree of internal bias. Third, there is no clear cut-off point for what constitutes an early relapse. Research by Wang et al. suggested that 92.1% of early recurrences occur within 48 hours (10), and in this study children with intestinal obstruction were observed for 48 hours after successful reduction and then discharged, so a recurrence within 48 hours was defined as an early recurrence. Large studies are still needed to define ERI and LRI in intussusception. Fourth, only one central set of data was collected for analysis, so we will seek external validation assessments in multicenter studies.

Conclusions

We found that age, vomiting, bloody stool, and monocyte ratio are independent risk factors for the occurrence of pediatric ERI. Children older than 1 year, who do not exhibit vomiting or bloody stool, and who do exhibit an increased ratio of monocytes, are at a higher risk for ERI. Our predictive model has the potential to provide individualized risk estimates for pediatric ERI for children with intussusception. And it can provide a theoretical basis for the length of stay of the child.

Acknowledgments

We would like to thank all of the staff of the First Affiliated Hospital of Shantou University Medical College of Pediatric Surgery for their support.

Funding: This work was supported by grants from the National Natural Science Foundation of China (No. 81801432), the Natural Science Foundation of Guangdong Province of China (Nos. 2022A1515010407, 2020A1515010135, and 2018A030307045), the Guangdong Provincial Science and Technology Fund for High-Level Hospital Construction (No. STKJ2021119), the “Dengfeng Project” for the Construction of High-Level Hospitals in Guangdong Province-the First Affiliated Hospital of Shantou University Medical College Supporting Funding (No. [2019]70), and the Major Science and Technology Project of Nanshan District of Shenzhen City (No. NSZD2023022).

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tp.amegroups.com/article/view/10.21037/tp-23-269/rc

Data Sharing Statement: Available at https://tp.amegroups.com/article/view/10.21037/tp-23-269/dss

Peer Review File: Available at https://tp.amegroups.com/article/view/10.21037/tp-23-269/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tp.amegroups.com/article/view/10.21037/tp-23-269/coif). All authors report that receiving funding from the National Natural Science Foundation of China (No. 81801432), the Natural Science Foundation of Guangdong Province of China (Nos. 2022A1515010407, 2020A1515010135, and 2018A030307045), the Guangdong Provincial Science and Technology Fund for High-Level Hospital Construction (No. STKJ2021119), the “Dengfeng Project” for the Construction of High-Level Hospitals in Guangdong Province-the First Affiliated Hospital of Shantou University Medical College Supporting Funding (No. [2019]70), and the Major Science and Technology Project of Nanshan District of Shenzhen City (No. NSZD2023022). The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The research protocol was approved by the Ethics Committee of the First Affiliated Hospital of Shantou University Medical College (No. B-2022-163). The informed consent was waived for this retrospective study.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Esmaeili-Dooki MR, Moslemi L, Hadipoor A, et al. Pediatric Intussusception in Northern Iran: Comparison of Recurrent With Non-Recurrent Cases. Iran J Pediatr 2016;26:e3898. [Crossref] [PubMed]
2. Binu V, Nicholson C, Granger J, et al. Ultrasound guided hydrostatic enema reduction of ileocolic intussusception: a safe and effective technique. ANZ J Surg 2023;93:1993-8. [Crossref] [PubMed]
3. Efrati Y, Klin B, Kozer E, et al. The role of dexamethasone in decreasing early recurrence of acute intussusception in children: A retrospective study. J Pediatr Surg 2017;52:1141-3. [Crossref] [PubMed]
4. Zhu D, Xu X, Zhang M, et al. Significance of MCP-1 in predicting the short-term recurrence of primary intussusception in children: An observational study. Medicine (Baltimore) 2022;101:e30743. [Crossref] [PubMed]
5. Ting X, Xufei D, Jiangbin L, et al. Development and Validation of a Nomogram for Predicting Pathological Intussusceptions in Children Prior to Surgical Intervention. Front Pediatr 2022;10:877358. [Crossref] [PubMed]
6. Yu F, Chen H, Cao X, et al. A Technique to Reduce the Early Recurrence of Intussusception in Ultrasound-Guided Hydrostatic Reduction. J Ultrasound Med 2022;41:2467-73. [Crossref] [PubMed]
7. Özcan R, Hüseynov M, Emre Ş, et al. A review of intussusception cases involving failed pneumatic reduction and re-intussusception. Ulus Travma Acil Cerrahi Derg 2016;22:259-64. [Crossref] [PubMed]
8. Talabi AO, Famurewa OC, Bamigbola KT, et al. Sonographic guided hydrostatic saline enema reduction of childhood intussusception: a prospective study. BMC Emerg Med 2018;18:46. [Crossref] [PubMed]
9. Chen SC, Wang JD, Hsu HY, et al. Epidemiology of childhood intussusception and determinants of recurrence and operation: analysis of national health insurance data between 1998 and 2007 in Taiwan. Pediatr Neonatol 2010;51:285-91. [Crossref] [PubMed]
10. Wang Z, He QM, Zhang H, et al. Intussusception patients older than 1 year tend to have early recurrence after pneumatic enema reduction. Pediatr Surg Int 2015;31:855-8. [Crossref] [PubMed]
11. National Health Commission of the People's Republic of China. WS/T 779-2021 Reference intervals of blood cell analysis for children. 2021. Available online: http://www.nhc.gov.cn/wjw/s9492/202105/19e3455a5dfe47e7a84756dd3968e475/files/f206930de56a400b9a8a33e5e59f6c05.pdf
12. Duan S, Zhang X, Jiang X, et al. Risk factors and predictive model for abdominal wound dehiscence in neonates: a retrospective cohort study. Ann Med 2021;53:900-7. [Crossref] [PubMed]
13. Gilmore AW, Reed M, Tenenbein M. Management of childhood intussusception after reduction by enema. Am J Emerg Med 2011;29:1136-40. [Crossref] [PubMed]
14. Kim JH, Lee JS, Ryu JM, et al. Risk Factors for Recurrent Intussusception After Fluoroscopy-Guided Air Enema. Pediatr Emerg Care 2018;34:484-7. [Crossref] [PubMed]
15. Jiang J, Jiang B, Parashar U, et al. Childhood intussusception: a literature review. PLoS One 2013;8:e68482. [Crossref] [PubMed]
16. Shen G, Zhang C, Li J, et al. Risk factors for short-term recurrent intussusception and reduction failure after ultrasound-guided saline enema. Pediatr Surg Int 2018;34:1225-31. [Crossref] [PubMed]
17. Xie X, Wu Y, Wang Q, et al. Risk factors for recurrence of intussusception in pediatric patients: A retrospective study. J Pediatr Surg 2018;53:2307-11. [Crossref] [PubMed]
18. Stewart CJ, Ajami NJ, O'Brien JL, et al. Temporal development of the gut microbiome in early childhood from the TEDDY study. Nature 2018;562:583-8. [Crossref] [PubMed]
19. Chong CYL, Bloomfield FH, O'Sullivan JM. Factors Affecting Gastrointestinal Microbiome Development in Neonates. Nutrients 2018;10:274. [Crossref] [PubMed]
20. Brooks B, Firek BA, Miller CS, et al. Microbes in the neonatal intensive care unit resemble those found in the gut of premature infants. Microbiome 2014;2:1. [Crossref] [PubMed]
21. Matamoros S, Gras-Leguen C, Le Vacon F, et al. Development of intestinal microbiota in infants and its impact on health. Trends Microbiol 2013;21:167-73. [Crossref] [PubMed]
22. Hebbandi Nanjundappa R, Sokke Umeshappa C, Geuking MB. The impact of the gut microbiota on T cell ontogeny in the thymus. Cell Mol Life Sci 2022;79:221. [Crossref] [PubMed]
23. Ksia A, Mosbahi S, Brahim MB, et al. Recurrent intussusception in children and infants. Afr J Paediatr Surg 2013;10:299-301. [Crossref] [PubMed]
24. Kratofil RM, Kubes P, Deniset JF. Monocyte Conversion During Inflammation and Injury. Arterioscler Thromb Vasc Biol 2017;37:35-42. [Crossref] [PubMed]
25. Shi C, Pamer EG. Monocyte recruitment during infection and inflammation. Nat Rev Immunol 2011;11:762-74. [Crossref] [PubMed]
26. Lee YW, Yang SI, Kim JM, et al. Clinical features and role of viral isolates from stool samples of intussuception in children. Pediatr Gastroenterol Hepatol Nutr 2013;16:162-70. [Crossref] [PubMed]
27. Tseng WY, Chao HC, Chen CC, et al. Adenovirus infection is a risk factor for recurrent intussusception in pediatric patients. Pediatr Neonatol 2023;64:428-34. [Crossref] [PubMed]
28. Feng Y, Zhou H, Shu Q, et al. Association of meteorological factors with paediatric intussusception in Hangzhou: an 8-year retrospective cohort study. BMJ Open 2022;12:e064967. [Crossref] [PubMed]
29. Pellegrini S, Sordi V, Bolla AM, et al. Duodenal Mucosa of Patients With Type 1 Diabetes Shows Distinctive Inflammatory Profile and Microbiota. J Clin Endocrinol Metab 2017;102:1468-77. [Crossref] [PubMed]
30. Gencay C, Kilicoglu SS, Kismet K, et al. Effect of honey on bacterial translocation and intestinal morphology in obstructive jaundice. World J Gastroenterol 2008;14:3410-5. [Crossref] [PubMed]