Abstract / Summary
Abstract Purpose: The vascular norepinephrine response index [VNERi = diastolic arterial pressure / (norepinephrine dose × heart rate)] is a bedside index computed from already-monitored signals. It has been studied only at baseline. We described its 48-hour course and tested whether it adds prognostic information beyond dynamic clinical measures. Methods: We studied a single-center prospective cohort of 94 adults with septic shock started on first-line norepinephrine. VNERi was computed at hours T0–T48. Course parameters (mean level, minimum, slope) were compared for 28-day mortality against the change in Sequential Organ Failure Assessment (SOFA) score (ΔSOFA) and lactate clearance, using the area under the receiver operating characteristic curve (AUC) with the DeLong test, nested logistic regression, and time-dependent Cox regression adjusted for time-varying SOFA(t) and lactate(t). Phenotypes came from a latent class mixed model. Results: Mortality was 32% (30/94). Baseline VNERi did not separate outcomes (AUC 0.59; most had a preserved response). The course parameters did (mean-level AUC 0.73) but did not outperform ΔSOFA (AUC 0.84, window-matched DeLong p = 0.02). Added to a dynamic clinical model, VNERi gave a small gain (integrated discrimination improvement 0.05–0.06, likelihood-ratio p ≈ 0.02). In time-dependent Cox, a higher VNERi(t) was linked to survival (hazard ratio 0.48) but weakened to non-significance after adjusting for SOFA(t) (0.68, p = 0.08). Three exploratory course phenotypes separated survival (log-rank p = 0.002). Conclusion: Serial bedside VNERi monitoring is feasible and carries a partly independent signal that adds a small amount to dynamic clinical severity without outperforming it. These findings are hypothesis-generating and need validation in a hyporesponsive population.