Capture-recapture using multiple data sources: estimating the prevalence of diabetes

Area of Science:

• Epidemiology
• Biostatistics

Background:

• Accurate estimation of diagnosed diabetes prevalence is crucial for public health.
• Traditional methods may underestimate disease burden due to data fragmentation.

Purpose of the Study:

• To assess the feasibility of using multiple data sources combined with capture-recapture techniques.
• To estimate the prevalence of diagnosed diabetes in a specific population.

Main Methods:

• Employed a model-averaging procedure with adjusted Akaike's Information Criterion (QAICc).
• Integrated capture-recapture estimates from log-linear models analyzing four distinct patient lists.
• Applied the methodology to patient data from Otago, New Zealand.

Main Results:

• Analyzed data from 5,716 individuals across four diabetes patient lists.
• The model-averaged prevalence estimate for diagnosed diabetes was 3.70% (95% CI: 3.36-4.04%).
• Prevalence in adults aged 15+ was estimated at 4.45% (95% CI: 4.03-4.86%).

Conclusions:

• Capture-recapture methods with model averaging provide a cost-effective and efficient tool for prevalence estimation.
• This technique allows for monitoring the diabetes epidemic and informing resource allocation.
• The findings align with national survey results, validating the approach.