Hyper-realistic image of a clinical setting where advanced AI systems assist doctors in prognosis and diagnosis. The background features medical monitors and data screens filled with predictive models. In the foreground, healthcare professionals analyze reports alongside AI-driven tools. Vibrant colors emphasize the futuristic theme.
The TRIPOD+AI statement provides updated guidance for reporting clinical prediction models developed with both regression and machine learning methods. These models have become vital tools in healthcare, aiding in prognosis and diagnosis. The original TRIPOD statement, published in 2015, mainly focused on regression-based models, but with the advancements in artificial intelligence (AI), especially in machine learning (ML), a revision was deemed necessary.
The TRIPOD+AI aims to standardize the reporting process, ensuring that models developed with machine learning, a rapidly growing area in medical decision-making, are reported in a transparent and complete manner. Gary Collins, one of the lead authors, emphasized, “The new guidelines are crucial in promoting the accurate reporting of AI-powered models, which are increasingly used in clinical settings.”
The authors propose a 27-item checklist that includes clear instructions for reporting on different aspects of model development and evaluation. The checklist is intended to help researchers, peer reviewers, and clinicians alike in evaluating the quality of the studies. “The goal is to facilitate model implementation in clinical practice by improving the reliability and usability of the models,” said Karel Moons, a co-author.
Prediction models play a critical role in clinical decision-making, guiding healthcare professionals in risk assessment, diagnostics, and even treatment plans. Popular examples of such models include the Framingham risk score for cardiovascular disease and the Gail model for breast cancer. With the integration of AI and machine learning, models can now process large datasets, offering even more precision. However, incomplete or inaccurate reporting can impair their reliability.
Poor reporting can obscure potential biases in model design, leading to risks when models are applied in patient care. This is why, according to the authors, accurate reporting is an ethical obligation. By following the TRIPOD+AI recommendations, researchers will ensure that models are more transparent, reducing the likelihood of implementing flawed models.
To further support the guidelines, the TRIPOD+AI checklist also addresses the development of protocols, abstracts, and systematic reviews for model-based studies. For instance, it highlights the importance of reporting sufficient details in abstracts to allow readers a quick yet comprehensive understanding of the model’s performance.
TRIPOD+AI is an essential step forward in ensuring that the development of machine learning models is ethically sound, methodologically robust, and applicable in real-world healthcare settings. The authors stress that the checklist should be used for all prediction model studies, irrespective of whether regression or machine learning techniques were used.
Citation:
Collins GS, Moons KGM, Dhiman P, Riley RD, Beam AL, Van Calster B, et al. TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods. BMJ. 2024;385. DOI: 10.1136/bmj-2023-078378
License:
This content is generated from the original article original article and it’s under Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial.
