aceCombining Statistical Modeling and Machine Learning for Prognostic Feature Selection in Cervical Cancer: A Retrospective Study Based on SEER 2004 to 2015 Data
Abstract: Objectives Cervical cancer is a leading female malignancy with high global morbidity/mortality, and remains high recurrence risk after standard treatment. Accurate prognostic feature identification is critical for personalized therapy and patient survival improvement, while traditional indicators and single biomarkers lack sufficient accuracy in prognostic prediction. Methods In this population-based retrospective study, we analyzed 5392 patients with cervical squamous cell carcinoma from the surveillance, epide…
202306 Cervical Cancer Screening by DataBase Center for Life Science (DBCLS). CC BY 4.0 · https://creativecommons.org/licenses/by/4.0
Researchers conducted a population-based retrospective analysis of 5392 patients with cervical squamous cell carcinoma in the SEER database from 2004 to 2015, using multivariable logistic regression and machine learning to evaluate sociodemographic and clinical predictors of overall survival. They found marital status, median household income, tumor grade, disease stage and tumor size were collectively related to prognosis, and built a model including age and race that achieved an AUC of 0.70.
The work matters because it shows that combining routinely collected social determinants with clinical features can improve risk stratification for a cancer with high recurrence risk after standard treatment, potentially informing personalized therapy. Uncertainty remains about generalizability beyond SEER squamous cell cases from 2004-2015, prospective validation, and whether the 0.70 discrimination is sufficient for clinical deployment.
- Study analyzed 5392 patients with cervical squamous cell carcinoma from SEER 2004 to 2015 using multivariable logistic regression and machine learning models.
- Identified prognostic factors included marital status, median household income, tumor grade (Grade Recode 2017), disease stage, and tumor size, along with age and race.
- Patients with lower socioeconomic status and residing in nonmetropolitan areas were found to have worse survival outcomes.
A model combining clinical and sociodemographic variables predicted overall survival in cervical squamous cell carcinoma with acceptable discrimination.
The rundown
The authors used SEER data from 2004 to 2015 covering 5392 patients with cervical squamous cell carcinoma and applied multivariable logistic regression alongside machine learning models to test associations between sociodemographic and clinical variables and overall survival.
The resulting model that included marital status, median household income, tumor grade, disease stage, tumor size, age and race achieved an AUC of 0.70, which the authors described as acceptable, and was proposed as a basis for stratified risk prediction tools.
Sources
- Peer-reviewedAmerican Journal of Clinical Oncology2026-08-24
How should this claim be treated?
The debate