Back

Protocol for a scoping review for risk prediction models for identifying late-onset cardiotoxic effects in cancer survivors for use in a primary care setting

Brady Bates, O.; Grecu, A. N.; Emery, J.; Stokes, D.; Cullen, W.; Gallagher, J.

2025-08-15 oncology
10.1101/2025.08.13.25333626 medRxiv
Show abstract

RationaleAfter completing treatment for cancer, survivors may experience late cardiotoxic effects: consequences of treatment that persist or arise after a latent period. Primary care providers are well situated to manage these effects. Risk prediction models may assist with this. PurposeTo identify and describe all models that predict the risk of late effects and could be used in an ambulatory care setting. Data sourcesSearches will be conducted across MEDLINE and EMBASE supplemented by manual reference list reviews and targeted grey literature searches using Google. Study SelectionStudies to be included are those describing models that 1) predicted the absolute risk of a late cardiotoxic effect present at least one year post-treatment, across all cancer treatment modalities, and 2) could be used in an ambulatory care clinical setting, Data Extraction & SynthesisData will be extracted as per the CHARMS checklist, an assessment of bias will be conducted as per the McGinn Checklist. DiscussionThis scoping review will comprehensively map existing tools for risk prediction of the late cardiotoxic effects of cancer therapy in cancer survivors, applicable to an ambulatory care setting. RegistrationThis protocol was registered on 21.6.25 to Zenodo : 10.5281/zenodo.15708760 Key MessagesO_LICancer survivors face a significant long-term risk of late-onset cardiotoxicity, often undetected until advanced stages. C_LIO_LIThis review will identify and map validated multivariable risk prediction models that could support the early identification of cardiotoxicity in ambulatory and primary care settings. C_LIO_LIIntegrating prediction tools into primary care may help address growing oncology workforce gaps while enabling timely, risk-adapted cardiovascular follow-up. C_LI

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.