Clinical prediction of complicated appendicitis: A case-control study utilizing logistic regression.

Area of Science:

• Gastroenterology
• Surgical Pathology
• Medical Diagnostics

Background:

• Increasing evidence supports conservative antibiotic treatment for acute appendicitis.
• Accurate differentiation between simple appendicitis (SA) and complicated appendicitis (CA) is essential for appropriate patient management.
• Previous studies identified several risk factors for CA but were limited by univariate analyses, necessitating multivariate approaches.

Purpose of the Study:

• To identify and evaluate non-laboratory and laboratory predictive factors for complicated appendicitis (CA) using logistic regression analysis.
• To develop a model for differentiating CA from SA based on clinical and laboratory data.

Main Methods:

• A retrospective case-control study involving 236 patients (198 SA, 38 CA) with diagnoses confirmed by computed tomography.
• Comparison of demographic, clinical, historical, vital sign, physical finding, and laboratory data between SA and CA groups.
• Logistic regression analyses were performed using non-laboratory factors alone and then including both non-laboratory and laboratory factors to predict CA.

Main Results:

• The CA group was older and had a longer onset-to-visit interval compared to the SA group.
• Elevated heart rate, body temperature, and serum C-reactive protein (CRP) levels were significantly higher in the CA group.
• Logistic regression Model 2, incorporating serum CRP, hyponatremia, and glomerular filtration rate, demonstrated that elevated CRP levels were a strong predictor of CA (AUC=0.87).

Conclusions:

• Serum C-reactive protein (CRP) level is a significant and dose-dependent predictor for differentiating complicated appendicitis (CA) from simple appendicitis (SA).
• Multivariate logistic regression analysis, particularly including laboratory markers like CRP, provides a robust method for CA prediction.