← All tools

ADAPT-EM — Adjuvant Therapy Decision Support for Endometrial Carcinoma

Molecular-anchored net-benefit modeling · shared decision-making with patient preferences, trial evidence (PubMed-linked) & a plain-language patient summary

Research decision-support prototype — not a validated clinical device. Built on published summary-level trial data + literature-anchored priors. Supports, not replaces, multidisciplinary tumor-board judgment. No patient data are stored.

Tumor & pathology

Patient preferences & goals

Set what matters most to this patient — each choice is plain-language, and the recommendation updates to match their priorities.
1 · Overall goal
Sets the overall balance between fighting the cancer and limiting side effects.
2 · How much does the patient want to AVOID each side effect?
5 · Deal-breakers (rule an option out)

Ranked strategies by net clinical benefit

Strategy & supporting evidence5-yr RFS5-yr OSΔRFSΔOSMeets goalGr≥3Net QALY
How to read this
Recommendation is the best evidence-based option after applying your preferences. Immunotherapy/PARP in early stage and other emerging options are tagged Investigational; deal-breakers mark options Excluded. "Meets goal" flags whether an option clears your minimum-benefit threshold. Net benefit = quality-adjusted life-years vs observation after subtracting preference-weighted toxicity. Trial names link to PubMed.
Strategies: observation; vaginal brachytherapy; pelvic EBRT; VBT + chemotherapy (GOG-249); chemoradiation (PORTEC-3); chemotherapy (GOG-122); chemo + immunotherapy (RUBY, NRG-GY018); chemo + immunotherapy + PARP (DUO-E). Engine: molecular-anchored baseline risk × subgroup-conditional trial HRs × preference-weighted toxicity, integrated by a 3-state competing-outcomes QALY model; identical to the reference Python implementation (max difference <10−¹⁵). Not for clinical use without external validation. Created by Adrian Kohut, MD. © 2026 Adrian Kohut, MD.