Related Experiment Video
Updated: Oct 20, 2025
Implementation of a Real-Time Psychosis Risk Detection and Alerting System Based on Electronic Health Records using CogStack
Published on: May 15, 2020
Prediction models in gynaecology: Transparent reporting needed for clinical application
Malou E Gelderblom1, Kelly Y R Stevens2, Saskia Houterman3
1Department of Obstetrics and Gynaecology, Catharina Hospital, Eindhoven, the Netherlands; Radboud Institute for Health Sciences, Department of Obstetrics and Gynaecology, Radboud University Medical Centre, Nijmegen, the Netherlands.
This review assessed prediction models in benign gynaecology against the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) criteria. Most studies showed good model performance but lacked comprehensive reporting and external validation, hindering clinical use.
Area of Science:
- Obstetrics and Gynaecology
- Medical Informatics
- Biostatistics
Background:
- Clinical use of prediction models is rising in obstetrics and gynaecology for personalized patient care.
- Accurate and transparent reporting of these models is crucial to prevent misuse.
- The Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) statement provides guidelines for model development and reporting.
Purpose of the Study:
- To review published prediction models for benign gynaecological conditions.
- To evaluate the adherence of these studies to the TRIPOD criteria.
Main Methods:
- A literature search was conducted in PubMed, Embase, and Cochrane Library up to August 2020.
- Twenty-two studies developing multivariable prediction models for benign gynaecology were included.
- A checklist derived from TRIPOD criteria was used to assess study quality.
Main Results:
- Most studies (91%) used univariable analysis before multivariable model development.
- Missing data handling was inadequate in many studies (9%).
- While 68% of models had a C-index ≥ 0.7, only 9% performed external validation, and 50% did not measure calibration.
Conclusions:
- Published prediction models in benign gynaecology often lack complete adherence to TRIPOD criteria, particularly regarding external validation and calibration reporting.
- Improved adherence to TRIPOD guidelines is essential for reliable clinical application of prediction models.
- Critical review and potential refusal of publication for studies not meeting criteria are recommended.
