Validation of the prognostic model "oncologiq palliative" for head and neck cancer patients.

Abstract

Purpose: Patients diagnosed with incurable head and neck squamous cell carcinoma (HNSCC) have a poor prognosis, with a median survival of approximately five months. Physicians often struggle to predict prognosis accurately and tend to overestimate survival. Timely sharing of validated accurate individual information on life expectancy could aid in facilitating better patient counseling. However, this knowledge is lacking. Therefore, the aim of this study is to conduct temporal and external validation of the prognostic model OncologIQ Palliative to assess its accuracy and generalizability.

Methods: The validation procedure involved temporal assessment in a retrospective cohort of 355 palliative HNSCC patients from Erasmus MC (2017-2020), followed by external validation in a retrospective cohort of 44 patients from Leiden University Medical Center (2019-2021).The discriminative ability of OncologIQ Palliative was assessed using the C-index and calibration was evaluated through graphical assessment, intercept, and slope.

Results: The temporal cohort had a median follow-up of 115 days, and the external cohort 143 days. The model showed moderate discriminative ability in temporal validation (C-index 0.66) and better discrimination in external validation (C-index 0.71). Reasonable agreement was observed between predicted and observed 6-month and 1-year survival rates, with some deviations from the perfect calibration line.

Conclusion: The validation methods conducted in this study underscored the reliability of OncologIQ Palliative. They showed adequate calibration and discrimination in both validation procedures, thereby facilitating the provision of more accurate prognostic counselling for head and neck cancer patients in the palliative phase.