Development and internal evaluation of models for screening kinesiophobia after hyperthermic intraperitoneal chemotherapy in patients with gastrointestinal cancer

Aims To examine characteristics associated with kinesiophobia and develop screening models for gastrointestinal cancer patients after surgery and hyperthermic intraperitoneal chemotherapy (HIPEC). Design A single-center cross-sectional study with internal model evaluation. Methods From October 2025 to May 2026, 521 patients were recruited by convenience sampling at a tertiary hospital in Liaoning Province, China. Candidate predictors and kinesiophobia were assessed on postoperative day 3 after clinical stabilization. A Tampa Scale for Kinesiophobia score ≥37 defined screening-positive status. LASSO and multivariable logistic regression were used for variable selection and association analysis. Logistic regression (LR), support vector machine (SVM), random forest (RF), and decision tree (DT) models were compared. A separate six-factor LR model was developed in 365 patients and evaluated in a same-center internal test set of 156 patients. Performance assessment included discrimination, classification, calibration, Brier scores, and decision curve analysis. Results Overall, 201 patients (38.6%) screened positive for kinesiophobia. Pain severity, anxiety, depression, general self-efficacy, postoperative catheter number, and comorbidities constituted the six-factor model. In the four-model comparison, test-set AUCs were 0.847 for RF, 0.842 for LR, 0.827 for SVM, and 0.730 for DT; corresponding Brier scores were 0.157, 0.153, 0.157, and 0.170. The separate six-factor model achieved a test-set AUC of 0.880 (95% CI, 0.825–0.934) and a Brier score of 0.144. At a probability threshold of 0.50, sensitivity was 76.92%, specificity was 78.85%, and the Youden index was 0.558. The calibration intercept-in-the-large and slope were −0.743 and 1.017, respectively; the Hosmer–Lemeshow test yielded χ 2 = 15.631 (df = 8; P = 0.048), indicating imperfect calibration. At a 20% assessment threshold, net benefit was 0.236, compared with 0.167 for assessing all patients. Conclusions The six-factor model demonstrated favorable discrimination for contemporaneous screening but overestimated screening-positive probability overall. Independent external validation, assessment of recalibration, and prospective clinical impact evaluation are required before routine implementation.

Authors

Institutions

Publication Details

Journal
Frontiers in Oncology
Published
2026-11-19
DOI
https://doi.org/10.3389/fonc.2026.1960163
Primary Topic
Cancer survivorship and care
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 evaluation of models for screening kinesiophobia after hyperthermic intraperitoneal chemotherapy in patients with gastrointestinal cancer

Yilong Song, Xinmin Sun, Yun Li, Jinxue Wang et al.
Frontiers in Oncology
Cancer survivorship and care
article

Development and internal evaluation of models for screening kinesiophobia after hyperthermic intraperitoneal chemotherapy in patients with gastrointestinal cancer

Yilong Song, Xinmin Sun, Yun Li, Jinxue Wang, Jing Li, Yanmei Ma
article en

Abstract

Aims To examine characteristics associated with kinesiophobia and develop screening models for gastrointestinal cancer patients after surgery and hyperthermic intraperitoneal chemotherapy (HIPEC). Design A single-center cross-sectional study with internal model evaluation. Methods From October 2025 to May 2026, 521 patients were recruited by convenience sampling at a tertiary hospital in Liaoning Province, China. Candidate predictors and kinesiophobia were assessed on postoperative day 3 after clinical stabilization. A Tampa Scale for Kinesiophobia score ≥37 defined screening-positive status. LASSO and multivariable logistic regression were used for variable selection and association analysis. Logistic regression (LR), support vector machine (SVM), random forest (RF), and decision tree (DT) models were compared. A separate six-factor LR model was developed in 365 patients and evaluated in a same-center internal test set of 156 patients. Performance assessment included discrimination, classification, calibration, Brier scores, and decision curve analysis. Results Overall, 201 patients (38.6%) screened positive for kinesiophobia. Pain severity, anxiety, depression, general self-efficacy, postoperative catheter number, and comorbidities constituted the six-factor model. In the four-model comparison, test-set AUCs were 0.847 for RF, 0.842 for LR, 0.827 for SVM, and 0.730 for DT; corresponding Brier scores were 0.157, 0.153, 0.157, and 0.170. The separate six-factor model achieved a test-set AUC of 0.880 (95% CI, 0.825–0.934) and a Brier score of 0.144. At a probability threshold of 0.50, sensitivity was 76.92%, specificity was 78.85%, and the Youden index was 0.558. The calibration intercept-in-the-large and slope were −0.743 and 1.017, respectively; the Hosmer–Lemeshow test yielded χ 2 = 15.631 (df = 8; P = 0.048), indicating imperfect calibration. At a 20% assessment threshold, net benefit was 0.236, compared with 0.167 for assessing all patients. Conclusions The six-factor model demonstrated favorable discrimination for contemporaneous screening but overestimated screening-positive probability overall. Independent external validation, assessment of recalibration, and prospective clinical impact evaluation are required before routine implementation.

Frontiers in OncologyVol. 16
Shanghai University of Traditional Chinese Medicine (CN), Jinzhou Medical University (CN)
Cancer survivorship and care
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