A FRAMEWORK TO SUPPORT SPECIALISTS IN THE PROBLEM OF HOSPITAL EXAMINATION SCALING

Abstract

Efficient management of patient flow in hospital settings, especially in the scheduling of diagnostic tests, represents a critical challenge given the high demand and limited resources. The Hospital Test Scheduling Problem consists of determining the processing sequence of tests across multiple stations, considering clinical priorities and arrival times. The objective of this work is to identify an alternative capable of scheduling hospital tests using heuristic algorithms. The methodology included a systematic literature review for theoretical grounding, followed by the implementation of three scheduling algorithms: Arrival Order Priority (AOP), based on arrival order; Minimum Processing Time Priority (MPTP), based on the shortest processing time; and Roulette Selection Strategy (RSS), of a probabilistic nature. To analyze the generated solutions, a framework capable of dynamically reconstructing the processing flow from the produced data was developed. The experiments were conducted in 150 distinct scenarios, varying the number of tests and processing stations. The results indicated that AOP presented the best average performance in terms of total processing time in all scenarios analyzed, while RSS obtained the longest average times, although it expanded the exploration of the solution space. It is concluded that the scheduling policy directly impacts operational performance and that the proposed framework contributes to the analysis, validation, and integration of decision support strategies in hospital environments.