Artikel
The prognostic quality of risk prediction models to assess the individual breast cancer risk in women: an overview of reviews
Suche in Medline nach
Autoren
-
Sarah Wolf
- HTA Austria – Austrian Institute for Health Technology Assessment GmbH, Österreich
-
Ingrid Zechmeister-Koss
- HTA Austria – Austrian Institute for Health Technology Assessment GmbH, Österreich
-
Irmgard Fruehwirth
- HTA Austria – Austrian Institute for Health Technology Assessment GmbH, Österreich
| Veröffentlicht: | 21. März 2023 |
|---|
Gliederung
Text
Background/research question: Breast cancer is the most common cancer among women globally, with an incidence of approximately two million cases in 2018. Organised age-based breast cancer screening programs were established worldwide to detect breast cancer earlier and to reduce mortality. Today, great hopes are set on risk-based screening programs, considering various risk factors next to age. The present study investigated the discriminatory accuracy of breast cancer risk prediction models and whether they suitable for risk-based breast cancer screening programs.
Methods: Following the PICO scheme, we conducted an overview of systematic reviews and performed a systematic search in four databases (the Ovid MEDLINE, EMBASE, the Cochrane Library and CRD databases). One author performed the literature selection, data extraction and synthesis, while a second author controlled all steps. We assessed the quality of the selected systematic reviews following the AMSTAR 2 tool.
Results: Based on the pre-defined inclusion criteria, we included eight systematic reviews with 107 primary studies out of 833 hits. Three of the eight systematic reviews were assessed as having a high risk of bias, while the others were rated with a moderate or low risk of bias. The identified breast cancer risk prediction models, e.g. the Gail, the Rosner-Colditz and the Tyrer-Cuzick model, show a low prognostic quality. They, therefore, are not suitable for predicting the short- and long-term breast cancer risk for individual women with sufficient accuracy. Adding breast density and genetic information as risk factors to the prediction models improved their discriminatory accuracy, but only moderately.
Conclusion: All breast cancer risk prediction models published to date show a limited ability to predict the individual breast cancer risk in women. Hence, it is too early to implement them in national breast cancer screening programs. Although relevant randomised controlled trials about the benefit-harm ratio of risk assessments in breast cancer screening are underway, results are not expected soon. After further evidence is available, a re-evaluation of prediction models in risk-based breast cancer screening programs is needed.
Competing interests: The authors have no relevant financial or non-financial interests to disclose.
