Additive and multiplicative models for relative survival rates

Area of Science:

• Biostatistics
• Epidemiology
• Survival Analysis

Background:

• Relative survival rate (RSR) estimates disease impact on mortality using a reference population.
• RSR is vital when cause of death is uncertain.
• Disease effects on mortality hazards can be additive or multiplicative.

Purpose of the Study:

• To examine both additive and multiplicative models for relative survival.
• To develop statistical methods for estimating disease effects on mortality.
• To explore disease effect constancy over follow-up periods.

Main Methods:

• Comparison of additive and multiplicative hazard models for RSR.
• Development of maximum likelihood and moment-based estimators.
• Analysis of disease effects assumed constant or piecewise constant over time.
• Exploration of relationships between moment-based and logrank scores.

Main Results:

• Additive models are presented as biologically more plausible than multiplicative ones.
• Methods for estimating disease effects under constant and piecewise constant assumptions are detailed.
• Moment-based statistics are shown to be related to logrank scores.

Conclusions:

• The study provides statistical tools for analyzing relative survival data.
• It offers a framework for understanding disease-specific mortality effects.
• The findings contribute to more accurate survival analysis in epidemiology.