Şenel İ. K.
Archives of health science and research (Online), cilt.12, sa.1, ss.1-8, 2025 (Scopus, TRDizin)
Özet
Objective: To examine spatial clustering patterns of physician distribution across Turkish provinces from 2002 to 2023 using spatial autocorrelation analysis and identify geographic disparities requiring policy intervention. Methods: Cross-sectional spatial analysis with longitudinal components across 81 Turkish provinces. Global Moran’s I and Local Indicators of Spatial Association (LISA) statistics quantified spatial autocorrelation using k-nearest neighbors spatial weights (k = 5). Z-score standardization enabled temporal comparisons. Monte Carlo permutation testing (999 iterations) assessed statistical significance at α = 0.05.Results: Significant positive spatial autocorrelation exists in physician distribution (Moran’s I = 0.235, P = .003). The LISA analysis identified 4 cluster types: Low-Low clusters (29 provinces, 35.8%), High-High clusters (25 provinces, 30.9%), Low-High outliers (15 provinces, 18.5%), and High-Low outliers (12 provinces, 14.8%). The gap between highest and lowest performing provinces spanned 5.38 standard deviations. This value refers to the difference between the average z-scores of the top- and bottom-ranked provinces across the study period, whereas the 7.02 SD figure represents the difference between annual extremes across the entire period. Conclusion: Turkish physician distribution exhibits significant spatial clustering with persistent Low-Low clustering in southeastern provinces. Spatial autocorrelation methods effectively identify priority areas for targeted health workforce interventions. The 10 provinces with significant Low-Low clustering require immediate policy attention to address systematic regional disadvantages in physician access.
Cite this article as: Şenel İK. Annual choropleth mapping and spatial autocorrelation of physicians in Türkiye: A 22-year analysis of provincial physician distribution patterns (2002-2023). Arch Health Sci Res. 2025, 13, 0170, doi: 10.5152/ArcHealthSciRes.2025.25170.