| Applicability Criteria |
Patients diagnosed with rectal cancer |
| Applicability Criteria |
Patients (to be) treated with radiation therapy |
| Name |
Model Card metadata |
| Based on template |
https://repo.metadatacenter.org/templates/8f15be13-9e24-487a-8844-f5b70bcc7cf0 |
| Creation date |
2024-08-19 |
| Email |
j.vansoest@maastrichtuniversity.nl |
| Ethical considerations - risks and harms |
Risk could be the low prediction rate, which would indicate a low number of pCR cases (10-15-20% in usual cases), which would mean patients would be referred for surgery instead of more frequent follow-up. |
| Out-of-scope use cases |
Automated clinical decision making |
| Description |
Model Card implementation in CEDAR forms |
| Model Input |
Generic Primary Tumor TNM Finding |
| Model Input |
Generic Regional Lymph Nodes TNM Finding |
| Model Input |
Tumor Length |
| Created on |
2024-08-19T06:11:28-07:00 |
| License |
Creative Commons |
| Primary intended use |
Prediction of pathologic complete response before start of treatment |
| Created by |
Johan van Soest |
| Ethical considerations - Human life |
Model is intended to support clinicians in deciding whether adjuvant treatment (after radiation therapy) is necessary. If a pathologic complete response is predicted, clinicians might opt for watch-and-wait (more follow-up CT/MRI scans) instead of surgery. In case of pCR it could save the patient from potential unnecessary surgery |
| Outcome |
Pathologic Complete Response |
| Reference |
https://doi.org/10.1016/j.radonc.2010.12.002 |
| Primary intended users |
Clinicians |
| Title |
Prediction of pathologic complete response for rectal cancer patients |
| Created by |
https://metadatacenter.org/users/d005592d-1c56-420c-9520-52f43c300f34 |
| Modified by |
https://metadatacenter.org/users/d005592d-1c56-420c-9520-52f43c300f34 |
| Foundational model or algorithm used |
Logistic Regression |
| Last updated on |
2024-08-19T13:20:39-07:00 |
| Ethical considerations - Mitigations |
If pCR is predicted, the risk of false-positives (prediction of pCR, but tumor tissue still available) can be mitigated by watch-and-wait approach with more frequent medical imaging. |
| Ethical considerations - Data |
Data used during model development was derived from clinical practice (patients). All data was retrospectively analyzed, and approved by the clinical trial for which patients did provide informed consent. |
| http://purl.obolibrary.org/obo/IAO_0000116 |
Based on Stiphout et al. (2011) - clinical variables only |
| https://schema.metadatacenter.org/properties/48fa46ee-5e7b-48af-bbb6-9ea0cd2da3b8 |
genid-7a2ed455eca74060a2cb07e8a3779fad6309164-b0 |