Multimorbidity analysis with low condition counts: a robust Bayesian approach for small but important subgroups.

Area of Science:

• Gerontology
• Biostatistics
• Epidemiology

Background:

• Examining long-term condition associations is vital for multimorbidity intervention but challenging with sparse data.
• The oldest old population presents unique challenges due to limited available data on comorbidities.

Purpose of the Study:

• To develop and apply a Bayesian inference framework robust to sparse data for quantifying morbidity associations.
• To compare the proposed Associations Beyond Chance (ABC) measure with standard Relative Risk (RR) in analyzing comorbidity networks.

Main Methods:

• Retrospective cross-sectional study of 12,009 primary care patients aged 90+ in Scotland (March 2007).
• Analysis of 40 long-term conditions, stratified by sex, comparing RR and the novel ABC measure.
• Construction of association networks to explore condition interactions and differences between RR and ABC estimations.

Main Results:

• The Bayesian framework demonstrated appropriate caution with sparse data, especially for uncommon conditions.
• This cautious approach impacted aggregated multimorbidity measures and network representations, including sex-specific differences.
• Differences in association analysis were observed when using RR versus ABC estimations.

Conclusions:

• Incorporating uncertainty is crucial in multimorbidity research to prevent misleading findings in small subgroups.
• The proposed Bayesian framework enhances the reliability of association estimations and research into disease mechanisms and multimorbidity.