Development and internal validation of a 12-month recurrence nomogram for non-puerperal mastitis: a two-center retrospective cohort study

Abstract To construct and internally validate a recurrence risk prediction model for non-puerperal mastitis (NPM) by integrating sociodemographic, clinical, pathological, laboratory, imaging, and treatment data. A retrospective study was conducted on 517 patients with first-diagnosed NPM admitted to two hospitals from January 2020 to August 2024. Of the 764 initially assessed patients, 83 were excluded for insufficient follow-up, leaving 681 patients who completed the 12-month follow-up. Among these 681 patients, 164 were further excluded due to predefined inclusion/exclusion criteria or incomplete data, resulting in 517 patients in the final analysis. The index date was the documented resolution of clinical symptoms and active imaging abnormalities after initial treatment. LASSO regression was used to select predictors, and multivariate Logistic regression was applied to construct a nomogram model. Model performance was evaluated using the receiver operating characteristic curve (ROC curve), calibration curve, and decision curve analysis. Internal validation was performed using the Bootstrap method. During the 12-month follow-up, 56 patients (10.8%) experienced recurrence. LASSO regression identified five predictors: age, white blood cell count, monocyte proportion, axillary lymphadenopathy, and surgical procedure. Multivariate Logistic regression showed that elevated white blood cell count (OR = 1.449), axillary lymphadenopathy (OR = 4.513), and minimally invasive surgery (OR = 14.390) were independent risk factors for recurrence, while increasing age (OR = 0.941) and elevated monocyte proportion (OR = 0.716) were protective factors. The apparent AUC was 0.921 (95%CI: 0.894–0.948); the optimism-corrected AUC after 5,000 bootstrap resamples was 0.897. At the Youden-optimal threshold, sensitivity was 0.82, specificity 0.87, positive predictive value 0.43, and negative predictive value 0.98. The calibration curve indicated good agreement between predicted and actual probabilities (Hosmer-Lemeshow test, P = 0.147). Decision curve analysis demonstrated clinical net benefit within the threshold probability range of 0.05–0.80. This exploratory nomogram shows good apparent discrimination in the development cohort, but the surgical-procedure coefficient should be interpreted as a predictive marker rather than a causal effect due to confounding by indication. External and center-specific validation are required before clinical use.

Authors

Institutions

Publication Details

Journal
Scientific Reports
Published
2026-11-16
DOI
https://doi.org/10.1038/s41598-026-79745-y
Primary Topic
Breast Lesions and Carcinomas
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Development and internal validation of a 12-month recurrence nomogram for non-puerperal mastitis: a two-center retrospective cohort study

Jian Qin, Ning Wang, Xing-song Tian, Ting-ting Ma et al.
3 links
Scientific Reports
Breast Lesions and Carcinomas
article

Development and internal validation of a 12-month recurrence nomogram for non-puerperal mastitis: a two-center retrospective cohort study

Jian Qin, Ning Wang, Xing-song Tian, Ting-ting Ma, Gao-chun Wu, Rui-feng Zhang
article en

Abstract

Abstract To construct and internally validate a recurrence risk prediction model for non-puerperal mastitis (NPM) by integrating sociodemographic, clinical, pathological, laboratory, imaging, and treatment data. A retrospective study was conducted on 517 patients with first-diagnosed NPM admitted to two hospitals from January 2020 to August 2024. Of the 764 initially assessed patients, 83 were excluded for insufficient follow-up, leaving 681 patients who completed the 12-month follow-up. Among these 681 patients, 164 were further excluded due to predefined inclusion/exclusion criteria or incomplete data, resulting in 517 patients in the final analysis. The index date was the documented resolution of clinical symptoms and active imaging abnormalities after initial treatment. LASSO regression was used to select predictors, and multivariate Logistic regression was applied to construct a nomogram model. Model performance was evaluated using the receiver operating characteristic curve (ROC curve), calibration curve, and decision curve analysis. Internal validation was performed using the Bootstrap method. During the 12-month follow-up, 56 patients (10.8%) experienced recurrence. LASSO regression identified five predictors: age, white blood cell count, monocyte proportion, axillary lymphadenopathy, and surgical procedure. Multivariate Logistic regression showed that elevated white blood cell count (OR = 1.449), axillary lymphadenopathy (OR = 4.513), and minimally invasive surgery (OR = 14.390) were independent risk factors for recurrence, while increasing age (OR = 0.941) and elevated monocyte proportion (OR = 0.716) were protective factors. The apparent AUC was 0.921 (95%CI: 0.894–0.948); the optimism-corrected AUC after 5,000 bootstrap resamples was 0.897. At the Youden-optimal threshold, sensitivity was 0.82, specificity 0.87, positive predictive value 0.43, and negative predictive value 0.98. The calibration curve indicated good agreement between predicted and actual probabilities (Hosmer-Lemeshow test, P = 0.147). Decision curve analysis demonstrated clinical net benefit within the threshold probability range of 0.05–0.80. This exploratory nomogram shows good apparent discrimination in the development cohort, but the surgical-procedure coefficient should be interpreted as a predictive marker rather than a causal effect due to confounding by indication. External and center-specific validation are required before clinical use.

Scientific Reports
Shandong University (CN), Shandong Provincial Hospital (CN), Jining First People's Hospital (CN), Inner Mongolia Maternal and Child Health (CN), Jining Medical University (CN)
Good health and well-being
Breast Lesions and Carcinomas
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon