Population age structure dependency of the excess mortality P-score

Area of Science:

• Demography
• Epidemiology
• Public Health

Background:

• The excess mortality P-score measures pandemic burden by comparing observed to expected deaths.
• Its age dependency is crucial for accurate cross-country comparisons, particularly during the COVID-19 pandemic.

Purpose of the Study:

• To formally and empirically assess the population structure bias of the P-score.
• To evaluate the P-score's suitability for comparing excess mortality across European countries during the COVID-19 pandemic.

Main Methods:

• Calculated P-scores for European countries (2021-2023) using UN World Population Prospects and HMD data.
• Estimated expected deaths via a Lee-Carter forecast model.
• Decomposed P-score differences using Kitagawa-type analysis and compared rank correlations between age-standardized and classical P-scores.

Main Results:

• The P-score is an average of age-specific excess deaths weighted by expected deaths.
• Differences in population structure played a marginal role in European P-score comparisons.
• Excess mortality effects were the dominant factor, with similar country rankings for age-standardized and classical P-scores.

Conclusions:

• The P-score is suitable for European excess mortality comparisons as it primarily reflects mortality differences, not structural biases.
• Findings should not be extrapolated to global comparisons due to potential variations in death distributions.
• Age-standardization can be employed to mitigate P-score biases in specific comparison scenarios.