A novel joint optimization of bed scheduling and patient re-allocation for hospital multiple departments based on homogeneous Markov chain

The scheduling of hospital beds is hindered by issues such as suboptimal utilization of clinical spaces and prolonged patient waiting times, resulting in an inequitable allocation of medical resources. To improve the quality of patient care, the adoption of efficient resource allocation and management strategies is imperative. Thus, in response to the current issues of uneven resource allocation among different departments and the inability of patients to receive timely treatment, this paper proposes a hospital multi-department bed scheduling and patient reallocation model based on homogeneous Markov chains. Firstly, the total capacity of the ward is considered to classify patients’ hospitalization statuses. Patients who are admitted for the first time are designated as Level 1, while those who are either transferred due to failure of initial admission or reassigned after admission are categorized as Level 2. A homogeneous Markov chain is employed to model the patient transfer process between different departments, with a dynamic factor introduced to simulate the deterioration of patients’ conditions during the waiting period. And, a heuristic optimization model is designed to provide an objective evaluation of the patient transfer model. Again, for the salp swarm algorithm (SSA) algorithm, a chaotic strategy is proposed to enhance the population’s average distribution, a dynamic search strategy is introduced to address the issue of the algorithm getting trapped in local optima, and a dynamic learning strategy is implemented to improve the influence of elite individuals on unknown leaders, thereby accelerating the algorithm’s convergence rate. The modified salp swarm algorithm is employed to solve the model, and tests are conducted to evaluate the performance of the improved algorithm. Finally, through numerical examples, the practicality of the proposed model and the superiority of the algorithm’s performance are demonstrated. The research results show that the model increases the success rate of first-time patient bed allocations by 122 %, confirming the effectiveness and applicability of the patient transfer process model. Moreover, a moderate increase in bed resources can reduce the hospital’s patient reallocation volume by 402 %.