Back

Comparison of Laboratory-Based and Combined Clinical-Laboratory Models for Predicting Complicated Appendicitis: Incremental Value of the Prognostic Nutritional Index

Miura, D.

2026-08-02 gastroenterology
10.64898/2026.07.30.26359322 medRxiv
Show abstract

Background: Early identification of complicated appendicitis is essential for appropriate management, yet accurate preoperative risk stratification remains challenging. Whether objective laboratory-based models can achieve predictive performance comparable to models incorporating clinical findings remains unclear. Objective: To compare clinical, laboratory-based, and combined clinical-laboratory prediction models for complicated appendicitis and to evaluate the incremental value of the Prognostic Nutritional Index (PNI). Methods: This retrospective study included 1,132 consecutive patients with acute appendicitis. Cases were classified as uncomplicated or complicated appendicitis based on pathological findings and imaging. Three multivariable logistic regression models were developed: a clinical model (Model A), a laboratory-based model (Model B), and a combined clinical-laboratory model (Model C). Model discrimination was assessed using the area under the receiver operating characteristic curve (AUC), and AUCs were compared using DeLong's test. Results: Complicated appendicitis was identified in 341 patients (30.1%). Higher C-reactive protein, higher neutrophil percentage, and lower PNI were independent predictors of complicated appendicitis. The AUCs were 0.791 for Model A, 0.847 for Model B, and 0.857 for Model C. Model B significantly outperformed Model A (P = 0.0020), whereas no significant difference was observed between Models B and C (P = 0.0816). PNI remained an independent predictor after adjustment for conventional inflammatory markers. Conclusion: A laboratory-based prediction model incorporating PNI demonstrated discrimination comparable to that of a combined clinical-laboratory model while relying exclusively on routinely available objective laboratory parameters. These findings support laboratory-based risk stratification as a practical adjunct to clinical assessment for the early identification of complicated appendicitis.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.