Beyond Spatial Proximity: Optimizing Nursing Home Bed Allocation with a Proposed ‘4A’ Model—Insights from Guangzhou, China

Background/objectives: As Guangzhou’s population ages rapidly, the supply-and-demand gap in elderly care facilities has become severe. We propose a ‘4A’ model that draws on accessibility, affordability, acceptability, and availability from the classic ‘5A’ access framework to optimize nursing home bed allocation from existing facilities. Methods: We applied the ‘4A’ model to 2630 communities/villages and 243 nursing homes in Guangzhou, using linear programming to maximize a matching score under capacity, demand, and occupancy constraints. Inequality was assessed using the Gini coefficient, Theil index, Moran’s I, and hot spot analysis. Four policy scenarios were simulated under resource-scarce and resource-abundant conditions. Results: A total of 753 (28.6%) communities/villages had no access to nursing homes. The unmet demand rates followed a ‘dual-center’ clustering pattern, and hot spots were not only in the urban core, where ‘scale disadvantage’ and ‘matching disadvantage’ coexisted, but also in outer suburban districts to form a ‘supply vacuum.’ Inequity was moderate, and it originated from between-district disparities. Compared with a ‘distance-only variant’ model, our ‘4A’ model allocated 3813 more beds. Policy simulations showed that minimum service (α=0.1) eliminated all unserved communities/villages; targeted bed expansion increased satisfaction rates to 74.8% and reduced the Gini coefficient to 0.280; and subsidies and quality upgrades became effective only when bed supply was abundant. The four simulations were phased into a three-phase policy roadmap. Conclusions: The ‘4A’ model transforms the qualitative ‘5A’ framework into a computable allocation matrix, offering actionable recommendations for equitable access to elderly care resources.