Developing Personas Based on the Identified Characteristics and Needs of ICU Patients: A Mixed-Methods Study

The high demand for critical care, limited medical resources and patient variability among Intensive Care Units (ICUs) suggest that standardized resource allocation may not fully address individual ICU patient needs. Patient persona, a representative profile synthesized from individuals with similar characteristics and needs, is gaining traction in the healthcare field but remains underexplored in ICUs.

To develop patient personas of ICU patients to support stratified resource management.

We employed an explanatory sequential mixed-methods design. Quantitatively, we analysed 48 709 ICU patients from the MIMIC-IV database using K-prototypes clustering to identify subgroups. Qualitatively, semi-structured interviews were conducted with 15 ICU nurses from three ICUs in China between May and July 2025 and analysed using conventional content analysis. Quantitative and qualitative findings were integrated to comprehensively characterize ICU patients.

Three distinct personas were identified. (1) Critical & Complex Support: elderly, highest disease severity, most life-support interventions, complete loss of disease management capacity, psychosocial disruption, longest hospitalization, highest mortality and need comprehensive support. (2) Severe & Intensive Support: significantly older, Medicare coverage, high disease severity and comorbidity burden, many life-support interventions, partial disease management capacity, dysfunctional psychosocial adaptation, short hospital stays then discharged to nursing facilities and need intensive support. (3) Sub-intensive & Basic Support: middle-aged, considerable disease severity but few life-support interventions, collaborate with healthcare providers yet show psychosocial maladaptation, shortest hospital stays, discharged home and need self-management support.

The three patient personas developed in this study may provide a reference for delivering precise care and optimizing ICU resource allocation. Further research is needed to evaluate the generalizability of these personas across different clinical settings.

Based on these personas, ICU managers can dynamically allocate resources (beds, staff, equipment) by subgroup characteristics and needs, while ICU nurses can prioritize and individualize bedside care.