From stakeholder mapping to statistical modeling: an illustrative demonstration of end-to-end Net-Map methodology for health governance analysis.

backgroundCoordinating primary cancer prevention in Europe requires alignment across the EU, national, and local actors. Yet governance relations are challenging to observe. We used a participatory network approach to make relationships visible and test which ties are associated with funding. This study serves as a methodological demonstration to illustrate the complete Net-Map research cycle from data collection to statistical analysis.

methodsThree Net-Map sessions (January-February 2024) with a national NGO expert panel elicited stakeholders and ties for authority, influence, and funding, followed by validation. Canvases were used to record the information, and they were digitized afterwards into edge and attribute files. We described network structure and estimated four nested exponential random graph models (ERGMs) for the directed funding network, progressively adding governance level (EU, national, local), formal authority, and informal influence as predictors of tie formation. Model diagnostics included goodness-of-fit assessment, structural extensions controlling for degree heterogeneity and clustering, and robustness checks across network subsets.

resultsThe final network comprises 128 organizations and 836 funding ties (density 5.14%). Funding is predominantly cross-level (71.4%). EU actors show the strongest outward activity (EU-national density 17.8%; EU-local 16.1%), while national-local is lower (6.2%). In ERGMs, influence is the strongest correlate of funding (OR 168.54, p < 0.001), authority is positive but smaller (OR 5.76, p < 0.001), and same-level funding is less likely (OR 0.35, p < 0.001). Model comparisons favored the full specification. Goodness-of-fit diagnostics were adequate for the modeled dyadic statistics. Supplementary structural extensions controlling for degree heterogeneity and clustering attenuate the influence OR to 41.49, while the authority OR remains stable (5.76–6.46), confirming that both effects are robust to model specification.

conclusionsThis illustrative analysis claims that participatory Net-Map workshops, combined with formal network analysis are helpful in diagnosing coordination patterns in decentralized health systems. The findings suggest that authority and influence serve complementary functions in governance, with implications that warrant further investigation with broader panels and longitudinal designs. The results primarily serve as an example of methodological application rather than providing generalizable insights about Romanian cancer prevention governance.