RESUMO
BACKGROUND: Midregional proadrenomedullin (MR-proADM) is a prognostic biomarker in patients with community-acquired pneumonia (CAP) and sepsis. In this paper, we examined the ability of MR-proADM to predict organ damage and long-term mortality in sepsis patients, compared to that of procalcitonin, C-reactive protein and lactate. METHODS: This was a prospective observational cohort, enrolling severe sepsis or septic shock patients admitted to internal service department. The association between biomarkers and 90-day mortality was assessed by Cox regression analysis and Kaplan-Meier curves. The accuracy of biomarkers for mortality was determined by area under the receiver operating characteristic curve (AUROC) analysis. RESULTS: A total of 148 patients with severe sepsis, according to the criteria of the campaign to survive sepsis, were enrolled. Eighty-five (57.4%) had sepsis according to the new criteria of Sepsis-3. MR-proADM showed the best AUROC to predict sepsis as defined by the Sepsis-3 criteria (AUROC of 0.771, 95% CI 0.692-0.850, p <0.001) and was the only marker independently associated with Sepsis-3 criteria (OR = 4.78, 95% CI 2.25-10.14; p < 0.001) in multivariate analysis. MR-proADM was the biomarker with the best AUROC to predict mortality in 90 days (AUROC of 0.731, CI 95% 0.612-0.850, p <0.001) and was the only marker that kept its independence [hazard ratio (HR) of 1.4, 95% CI 1.2-1.64, p <0.001] in multivariate analysis. The cut-off point of MR-proADM of 1.8 nmol/L (HR of 4.65, 95% CI 6.79-10.1, p < 0.001) was the one that had greater discriminative capacity to predict 90 days mortality. All patients with MR-proADM concentrations ≤0.60 nmol/L survived up to 90 days. In patients with SOFA ≤ 6, the addition of MR-proADM to SOFA score increased the ability of SOFA to identify non-survivors, AUROC of 0.65 (CI 95% 0.537-0.764) and AUROC of 0.700 (CI 95% 0.594-0.800), respectively (p < 0.05 for both). CONCLUSIONS: MR-proADM is a good biomarker in the early identification of high risk septic patients and may contribute to improve the predictive capacity of SOFA scale, especially when scores are low.