Frontiers in Surgery (Jun 2022)

Development and Validation of a Clinical Prediction Model for Complicated Appendicitis in the Elderly

  • Hui Feng,
  • Hui Feng,
  • Qingsheng Yu,
  • Qingsheng Yu,
  • Jingxing Wang,
  • Jingxing Wang,
  • Yiyang Yuan,
  • Yiyang Yuan,
  • Shushan Yu,
  • Shushan Yu,
  • Feisheng Wei,
  • Feisheng Wei,
  • Zhou Zheng,
  • Zhou Zheng,
  • Hui Peng,
  • Hui Peng,
  • Wanzong Zhang,
  • Wanzong Zhang

DOI
https://doi.org/10.3389/fsurg.2022.905075
Journal volume & issue
Vol. 9

Abstract

Read online

BackgroundFor elderly patients with mild clinical symptoms of uncomplicated appendicitis(UA), non-surgical treatment has been shown to be feasible, whereas emergency surgical treatment is recommended in elderly patients with complicated appendicitis(CA), but it is still challenging to accurately distinguish CA and UA before treatment. This study aimed to develop a predictive model to assist clinicians to quickly determine the type of acute appendicitis.MethodsWe retrospectively studied the clinical data of elderly patients with acute appendicitis who visited the First Affiliated Hospital of Anhui University of Traditional Chinese Medicine from January 2012 to January 2022. The patients were divided into UA group and CA group, and the general conditions, medical history, physical examination, laboratory examination and imaging examination were compared between the two groups, and SPSS 26.0 and R 4.0.2 software were used to establish CA clinic. Predict the model, and validate it internally.ResultsThe clinical data of 441 elderly patients with acute appendicitis were collected, 119 patients were excluded due to incomplete clinical data or other diseases. Finally, 332 patients were included in the study and divided into UA group (n = 229) and CA group (n = 103). By analyzing the clinical data of the two groups of patients, the duration of abdominal pain [OR = 1.094, 95% CI (1.056–1.134)], peritonitis [OR = 8.486, 95% CI (2.017–35.703))] and total bilirubin [OR = 1.987, 95% CI (1.627–2.426)] were independent predictors of CA (all p < 0.01). The model's Area Under Curve(AUC) = 0.985 (95% CI, 0.975–0.994). After internal verification by Bootstrap method, the model still has high discriminative ability (AUC = 0.983), and its predicted CA curve is still in good agreement with the actual clinical CA curve.ConclusionWe found that a clinical prediction model based on abdominal pain duration, peritonitis, and total bilirubin can help clinicians quickly and effectively identify UA or CA before treatment of acute appendicitis in the elderly, so as to make more scientific clinical decisions.

Keywords