# Factors influencing therapeutic efficacy of non-lactational mastitis based on hematological inflammatory markers and establishment of a predictive model

## TL;DR

This study identifies blood markers that predict treatment success in non-lactational mastitis and builds a model to help guide therapy.

## Contribution

A novel predictive model for NLM therapeutic efficacy using hematological markers SII and CRP is established.

## Key findings

- SII and CRP were found to significantly influence treatment outcomes in non-lactational mastitis.
- The predictive model showed strong performance with an AUC of 0.94 in the modeling group and 0.85 in the validation group.
- The model demonstrated good calibration and clinical utility based on DCA analysis.

## Abstract

To explore factors influencing the therapeutic efficacy of non-lactational mastitis (NLM) based on hematological inflammatory markers and to establish a predictive model.

Two hundred and sixty-four cases of NLM patients admitted to The Fourth People’s Hospital of Zhenjiang and Changzhou Maternal and Child Health Care Hospital from January 2019 to December 2022 were retrospectively selected. The patients were divided into a modeling group (n = 185) and a validation group (n = 79) in a 7:3 ratio with a random number table method. In the modeling group, patients were further divided based on therapeutic efficacy into a good efficacy group (n = 133) and a poor efficacy group (n = 52).

The results of binary logistics regression analysis showed that SII and CRP were influencing factors of therapeutic efficacy of NLM (p < 0.05). The formula for the model expression was Logit(P) = −74.457 + (0.823X1) + (0.589X2). The calibration curve slope of the model in the modeling group and validation group was close to a straight line, indicating good consistency between the predicted risk and actual risk. The ROC analysis results showed that the area under the curve (AUC) of the model in the modeling group was 0.94 (95% CI: 0.823–0.966), the area under the curve of the model in the validation group was 0.85 (95% CI: 0.794–0.893), drawing DCA curves to evaluate the clinical utility of the model in predicting efficacy showed clear positive net benefits, indicating good clinical utility.

SII and CRP are influencing factors of therapeutic efficacy of NLM. This study successfully established a prediction model based on hematological inflammatory markers.