HAprog: A New Prognostic Application to Assist Oncologists in Routine Care

Area of Science:

• Oncology
• Palliative Care
• Medical Informatics

Background:

• Palliative care (PC) is increasingly initiated earlier in disease progression.
• The Barretos Prognostic Nomogram (BPN) aids survival prognostication in advanced cancer but requires manual calculations and lab tests.
• Existing tools limit routine clinical application due to complexity and data requirements.

Purpose of the Study:

• To refine the Barretos Prognostic Nomogram (BPN).
• To develop a user-friendly smartphone application (App) for prognostic calculations.
• To enhance survival prognostication for advanced cancer patients in clinical settings.

Main Methods:

• Reanalysis of two advanced cancer patient cohorts (n=215 development, n=276 validation).
• Revision of metastasis variables and replacement of Karnofsky Performance Status (KPS) with Eastern Cooperative Oncology Group Performance Status (ECOG-PS).
• Multivariable regression analyses (Cox, Log-logistic) and backward variable elimination for model selection; calibration and discrimination assessed.

Main Results:

• Two models were developed: a 6-parameter 'full version' and a 5-parameter 'clinical version' excluding lab variables.
• Both models demonstrated good calibration and discrimination (c-indices 0.778 and 0.739).
• The HAprog App provides rapid, offline prognostic calculations for clinicians, usable in under a minute.

Conclusions:

• The HAprog App integrates refined prognostic models with strong calibration and discrimination.
• These tools offer practical benefits for oncologists managing advanced cancer outpatients.
• The App facilitates informed decision-making in clinical practice.