We study a home health care (HHC) problem that is characterized by prioritized patients and uncertain demands. In practice, HHC supply chain networks often struggle to meet high demand due to a shortage of service vehicles. Additionally, disruptions caused by natural calamities and pandemics (e.g., COVID-19) further compound these challenges, necessitating the consideration of real-life characteristics such as patient priorities, infrastructure locations, and transportation of medical supplies with uncertain demands. To formulate the problem, we propose a multi-depot and multi-period chance-constrained optimization model with precedence constraints, assuming that the demand quantities for medical supplies are random variables. Since patients’ medical conditions vary in severity, the priority of each patient is translated into a time-dependent potential healthcare cost that changes dynamically over the planning horizon. The solution to the proposed model determines the optimal locations for the base Mobile Health Facilities (MHFs) and the fleet size of HHC vehicles, and generates scheduling and routing plans to visit patients within specified time windows. We propose a unique three-phase solution approach, integrated with stochastic simulation, to address the problem. We then assess the robustness of the proposed model based on a realistic case of HHC service provision in Hong Kong and explore the optimal values for two model parameters, namely the Vehicle Threshold Index and the MHF Threshold Index. The performance evaluation tests show that the proposed solution method is efficient and effective for solving real-world problems.