Coincidence analysis: a new method for causal inference in implementation science

Area of Science:

• Implementation Science
• Public Health Research
• Mathematical Modeling

Background:

• Multifaceted interventions involve complex, interrelated elements.
• Real-world implementation research requires methods to understand causal inference.
• Coincidence Analysis (CNA) is a novel mathematical method for analyzing causal pathways.

Purpose of the Study:

• To apply Coincidence Analysis (CNA) to human papillomavirus (HPV) vaccination data.
• To compare CNA findings with traditional regression analysis.
• To identify necessary and sufficient conditions for vaccination uptake.

Main Methods:

• Applied CNA to county-level data on HPV vaccination campaigns and uptake in Sweden (2012-2014).
• Compared results from CNA with previously published regression findings.
• Utilized a cross-case method for causal inference.

Main Results:

• CNA identified multiple causal paths to high vaccination uptake, unlike regression analysis.
• High HPV vaccination rates were achieved by offering vaccines in all schools.
• An alternative path to high uptake involved offering vaccines in some schools combined with media coverage.

Conclusions:

• CNA provides a novel comparative approach for implementation researchers.
• The method aids in understanding how implementation conditions interact to influence outcomes.
• CNA can uncover empirical findings missed by other methods.