Marked Elevation in Serum Procalcitonin Levels Do Not Correlate With Severity of Disease or Mortality in Hospitalized Patients: A Retrospective Study.

Background: Bacteremia and sepsis are significant contributors to the morbidity, mortality, and economic burden of health care systems worldwide. Procalcitonin has been identified as a potentially useful marker of disease and severity in sepsis. However, the assumption that greater procalcitonin levels correlate with greater burden of disease has not been adequately studied.

Methods: A retrospective chart review of adult patients admitted to an urban teaching hospital with suspected sepsis was undertaken to test the association of elevated procalcitonin (>30 ng/mL) with other markers of sepsis (lactic acid, white blood cell count, percent bands), severity of disease (Sequential Organ Failure Assessment [SOFA] and Acute Physiology and Chronic Health Evaluation-II [APACHE II] scores), and mortality.

Results: In total, 168 patients were identified over 18 months (42% ward, 11% Stepdown, 44% medical intensive care unit [MICU], 2% surgical intensive care unit (STICU), 1% gynecology [GYN]). The Spearman correlation analysis showed that serum procalcitonin level did not correlate with SOFA ( = .238) or APACHE II ( = .918) scores on admission, and did not correlate with survival (Kruskal-Wallis test,  = .937). However, higher serum procalcitonin levels were associated with patients who had positive blood cultures (Kruskal-Wallis test,  = .0016 for Gram-positive and  = .0007 for Gram-negative bacteria). Lactic acid levels on admission strongly correlated with SOFA APACHE II (the Spearman correlation,  < .0001 for both) and mortality ( = .0001 for both).

Conclusions: Higher serum procalcitonin levels above 30 ng/mL failed to correlate with indicators of sepsis, severity of disease (SOFA and APACHE II scores), and mortality but were associated with positive blood cultures. Lactic acid levels did show correlation to both severity of disease and mortality. Serum procalcitonin levels >30 ng/mL do not appear to correlate with the severity of disease in a sample of patients with markedly elevated initial procalcitonin levels.