Comparison of Predictive Powers for Mortality between Systemic Vascular Resistance Index and Serum Lactate in Children with Persistent Catecholamine-Resistant Shock

Persistent catecholamine-resistant shock in children causes severe morbidity and mortality. We aimed to analyze the association between hemodynamics and serum lactate at different time points and 28-day mortality in children with persistent catecholamine-resistant shock. Methods. Twenty-six children with persistent catecholamine-resistant shock were enrolled, and their hemodynamics were monitored using the pulse index continuous cardiac output. Serial cardiac index (CI), systemic vascular resistant index (SVRI), and vasoactive-inotropic score (VIS) were analyzed for the first 24 hours. Associations between hemodynamics, serum lactate, and 28-day mortality were analyzed. Results. The 28-day mortality rate was 53.8%. SVRI and VIS were independent predictors of 28-day mortality. The mortality group had lower serial SVRI and higher VIS than the survival group (p < 0.05). Serial SVRI had the highest area under the receiver operating characteristic curve (AUC) for predicting mortality during the first 24 hours of persistent catecholamine-resistant shock. Three important hemodynamic parameters, CI, SVRI and perfusion pressure (MAP-CVP), were significantly correlated with lactate, of which SVRI had the best correlation (r = −0.711, p < 0.001). According to the AUC, SVRI was a more powerful predictor of mortality than lactate in persistent catecholamine-resistant shock. After 24 hours of treatment for persistent catecholamine-resistant shock, an SVRI lower than 1284 dyn·s·cm(−5)·m(2) was associated with 28-day mortality. Conclusions. SVRI was an early factor associated with mortality in the pediatric patients with persistent catecholamine-resistant shock potentially and had the good correlation with serum lactate. An SVRI more than 1284 dyn·s·cm(−5)·m(2) during the first 24 hours of persistent catecholamine-resistant shock was associated with favorable outcomes. The result should be used with caution due to the small sample size.