A hybrid multi-criteria decision-making framework for prioritising clinical determinants of medical futility in the intensive care unit

Determinations of medical futility in the intensive care unit (ICU) are high-stakes, ethically charged, and clinically variable. Contemporary literature increasingly frames such situations as potentially inappropriate treatment (PIT), particularly when physiologically feasible interventions are judged unlikely to provide meaningful benefit. This study aimed to construct a transparent decision-analytical framework for identifying the clinical factors ICU physicians prioritise when assessing futility and PIT, using a hybrid multi-criteria decision-making (MCDM) approach. A decision-analytical modelling study was conducted in Türkiye with a national expert panel of 20 ICU physicians with 1–18 years of post-speciality experience. A two-round Delphi pre-process with five intensivists, using an a priori 80% consensus threshold, identified 15 clinical factors and 8 evaluation criteria. Criterion weights were derived using Shannon Entropy, and six MCDM techniques were applied in parallel. Bootstrap resampling, ±20% sensitivity analysis, and Spearman rank correlation were used to assess stability and agreement. Terminal malignancy (F4) and irreversible neurological injury (F5) ranked first and second across all six methods (mean ranks 1.000 and 2.000, σ = 0.00, 100% bootstrap stability). Inter-method agreement was very strong (Spearman ρ = 0.964–1.000, p < 0.001), while inter-group agreement across experience strata was moderate to strong ( ρ = 0.643–0.807, p < 0.01). Documented treatment limitation preference (F15) ranked last across every method and experience group. Early-career clinicians ranked clinical frailty (F2) first, whereas advanced clinicians elevated refractory shock (F6) to fourth. Advanced age (F1) showed the widest experience-related variation, ranking 5th, 14th, and 11th across early-career, mid-career, and advanced groups, respectively. The Top-5 factors remained stable under ± 20% weight perturbation. In this expert-panel MCDM framework, ICU physicians appeared to prioritise biological non-recoverability when assessing futility and PIT, while treating treatment limitation preference as a separate ethical domain. The framework captures clinicians’ cognitive prioritisation rather than outcome-validated prognostic accuracy and should be read as a decision-support scaffold, not a bedside prediction tool. Prospective outcome validation and recalibration across different legal and cultural settings are required before clinical implementation.