Improving Survival Prognostication in Patients With Metastatic Cancer Through Clinical Judgment

Area of Science:

• Oncology
• Clinical Prognostics
• Cancer Survival Prediction

Background:

• The NEAT model is a validated tool for estimating survival in metastatic cancer patients.
• Clinical judgment is crucial but its predictive value relative to prognostic models is not well-quantified.
• This study aimed to compare clinician predictions with the NEAT model's accuracy.

Purpose of the Study:

• To evaluate if experienced clinicians (nurses and physicians) can predict patient survival more accurately than the NEAT prognostic model.
• To quantify the added value of clinical judgment in survival prediction for metastatic cancer patients.
• To assess the impact of survival predictions on treatment decisions, specifically radiation dose intensity.

Main Methods:

• Prospective cohort study of 73 metastatic cancer patients referred to radiation oncology.
• Nurses and physicians prospectively predicted if patients would survive longer or shorter than NEAT model estimates.
• McNemar's chi-squared test was used to compare prediction accuracy.

Main Results:

• Clinician predictions (nurses and physicians) showed similar accuracy (around 61%).
• Prediction accuracy was not significantly different between confident and less confident judgments.
• Patients receiving higher radiation dose intensity (EQD2≥40) had significantly longer median survival (17 months vs. 2 months).

Conclusions:

• Experienced clinicians provide insights that modestly enhance the accuracy of validated prognostic models for advanced cancer survival.
• Clinical judgment, while valuable, does not entirely replace the need for validated prognostic tools.
• Survival predictions influenced treatment decisions, leading to improved outcomes in some cases.