Mid-Regional Proadrenomedullin and Mid-Regional Proatrial Natriuretic Peptide Clearance Predicts Poor Outcomes Better Than Single Baseline Measurements in Critically Ill Patients With Pneumonia: A Retrospective Cohort Study

Abstract

Background We assessed the ability of baseline and serial measurements of mid-regional proadrenomedullin (MR-proADM) and mid-regional proatrial natriuretic peptide (MR-proANP) to predict 28-day mortality in critically ill patients with pneumonia compared with Acute Physiological and Chronic Health Evaluation IV (APACHE IV) model and Sequential Organ Failure Assessment (SOFA) score. Methodology Biomarkers were collected for the first five days in this retrospective observational cohort study. Biomarker clearance (as a percentage) was presented as biomarker decline in five days. We investigated the relationship between biomarkers and mortality in a multivariable Cox regression model. APACHE IV and SOFA were calculated after 24 hours from intensive care unit admission. Results In 153 critically ill patients with pneumonia, 28-day mortality was 26.8%. Values of baseline MR-proADM, MR-proANP, and APACHE IV were significantly higher in 28-day nonsurvivors, but not significantly different for SOFA score. Baseline MR-proADM and MR-proANP, APACHE IV, and SOFA had a low area under the curve in receiver operating characteristics (ROC) curves. No optimal cut-off points could be calculated. Biomarkers and severity scores were divided into tertiles. The highest tertiles baseline MR-proADM and MR-proANP were not significant predictors for 28-day mortality in a multivariable model with age and APACHE IV. SOFA was not a significant predictor in univariable analysis. Clearances of MR-proADM and MR-proANP were significantly higher in 28-day survivors. MR-proADM and MR-proANP clearances had similar low accuracy to identify nonsurvivors in ROC curves and were divided into tertiles. Low clearances of MR-proADM and MR-proANP (first tertiles) were significant predictors for 28-day mortality (hazard ratio [HR]: 2.38; 95% confidence interval [CI]: 1.21-4.70; p = 0.013 and HR: 2.27; 95% CI: 1.16-4.46; p = 0.017) in a model with age and APACHE IV. Conclusions MR-proADM and MR-proANP clearance performed better in predicting 28-day mortality in a model with age and APACHE IV compared with single baseline measurements in a mixed population of critically ill with pneumonia.