Priority setting in critical care: a multicriteria approach to ranking access to Intensive Care Unit beds

There is a situation in public health services where there is a high demand for services but not enough capacity to meet that demand. Within the critical care landscape in Brazil, there exists a significant disparity: of the total of 45,848 beds in intensive care units, only 49% are accessible to the Unified Health System (Sistema Único de Saúde - SUS), while the remaining 51% are exclusively allocated to a mere 23% of the population. Therefore, multicriteria decision analysis has the potential to be utilized as a tool in the decision-making process since it has been stated as a tool that promotes transparency and brings rationality to the decision-making process. The objective of this study was to create a multi-criteria model that can assist in the allocation of beds in Adult Intensive Care Units in the state of Rio Grande do Norte, Brazil.

A twelve-step framework was used to carry out the decision-aiding process. The PROMETHEE I and II methods were employed to build the multicriteria model. Three stages were followed: problem structuring, preference modeling and aggregation, and finalization. The decision problem was approached from the perspective of the P.γ ranking problem in which six clinical vignettes based on actual patients were ranked in order of preference, defined by the decision-maker's preference system.

Two pre-orders were recommended: a partial one, in which four pairs of actions were incomparable with each other; and a complete one, in which relationships between each pair of alternatives were described. The sensitivity analysis revealed considerable robustness of the model because of the range of the stability interval.

The partial ranking proposed by PROMETHEE I showed that Patient 6 outranks all the patients within the set of alternatives. When the complete ranking was generated, aided by PROMETHEE II, Patient 6 remained as the alternative that outranks all the others. Therefore, considering the resource-constrained scenario, Patient 6 was recommended to be admitted. The proposed model formalized the decision-making process of prioritizing access to Adult Intensive Care Unit beds, ensuring transparency and rationality by structuring all relevant factors.