Is it possible to predict appendicitis caused by Enterobius vermicularis before surgery?

Abstract

Objectives: This study aimed to develop statistical models based on hemogram data to predict the presence of the nematode parasite Enterobius vermicularis (EV) in the appendix before surgery. Patients and methods: The retrospective case-control study was generated from histopathological data of appendectomy 9,605 patients between January 1, 2007, and August 1, 2023. Enterobius vermicularis was detected in 32 patients (24 pediatric and 8 adult patients). Six patients were excluded due to missing hemograms, and 26 participants (pediatric patients: 10 males, 9 females; mean age: 11.1±4.4 years; range, 2 to 17 years & adult patients: 1 male, 6 females; mean age: 39.9±14.7 years; range, 22 to 68 years) were evaluated. The control (non-EV) group was composed of 99 randomly selected patients (42 males, 57 females; mean age: 18.3±13.5 years; range, 2 to 68 years) of nonparasitic acute appendicitis with available preoperative hemogram data. Univariate analysis was conducted on hemogram parameters to compare the groups, followed by predictive modeling using binomial logistic regression. Results: Enterobius vermicularis was present in 0.33% of all appendicitis patients and in 0.54% of pediatric patients. Histopathological diagnosis of appendicitis was present in 46.8% of EV patients, with a higher rate among pediatric patients (50%) compared to adult patients (37.5%). Patients with EV exhibited significantly lower counts of neutrophils and white blood cells in comparison to the non-EV group (p-values 0.031 and 0.046, respectively). The most effective EV prediction model (area under the curve: 0.685 [0.528-0.770]) ultimately included platelet distribution width and neutrophil count after evaluating all parameters (with corresponding p-values of 0.022 and 0.042, respectively). Conclusion: It is difficult to predict the presence of EV based on hemogram data prior to appendectomy. Studies that collect large amounts of data from multiple centers and different populations could provide better predictive models.