Conditional Survival: A Useful Concept to Provide Information on How Prognosis Evolves over Time

Area of Science:

• Biostatistics
• Clinical Epidemiology
• Survival Analysis

Background:

• Conditional survival (CS) provides dynamic predictions of patient prognosis.
• It is relevant in both absolute (clinical) and relative (public health) forms.
• CS incorporates disease progression and biomarker data up to a specific time point.

Purpose of the Study:

• To define and illustrate the application of absolute conditional survival.
• To highlight CS as a dynamic prognostic tool for chronic diseases.
• To demonstrate CS estimation using clinical cohort data.

Main Methods:

• Utilizing conditional Kaplan-Meier estimates within defined patient strata (e.g., age, stage).
• Employing conditional regression models (e.g., Cox models) for time-to-event data.
• Estimating CS as a function of prediction time 's' parametrically and nonparametrically.

Main Results:

• Absolute CS was estimated in a large multiple myeloma patient cohort.
• The study illustrates CS as a function of prediction time 's'.
• Methodological considerations for CS investigation include long-term follow-up and accounting for migration and treatment changes.

Conclusions:

• Conditional survival offers crucial, evolving prognostic information for patients with chronic diseases.
• It serves as a foundation for identifying factors influencing long-term survival.
• Accurate CS estimation requires robust long-term patient follow-up and consideration of dynamic clinical factors.