Estimation of age-standardized net survival, even when age-specific data are sparse

Area of Science:

• Epidemiology
• Biostatistics
• Cancer Research

Background:

• Age-standardization is crucial for international cancer survival comparisons.
• Standard methods struggle with data sparsity, leading to missing estimates.

Purpose of the Study:

• To demonstrate a pre-weighting approach as a viable alternative for external age-standardization in population-based cancer data.
• To show the pre-weighting approach performs well in sparse data scenarios.
• To couple pre-weighting with the Pohar Perme estimator for cohort and period analyses.

Main Methods:

• Utilized a pre-weighting approach for external age-standardization.
• Integrated the Pohar Perme estimator for both cohort and period analyses.
• Employed SEER public use data, including sparse scenarios (e.g., Connecticut by race).

Main Results:

• The pre-weighting approach yields comparable estimates to traditional methods with sufficient data.
• It successfully produces estimates throughout follow-up in sparse data situations where traditional methods fail.

Conclusions:

• Recommends adopting the Pohar Perme estimator with pre-weighting for international and national cancer survival studies.
• This approach overcomes non-estimation issues in sparse data.
• Facilitates more consistent and reliable survival estimate comparisons.