Letters to the Editor
To the Editor: In their recent article in Shock, Day et al. (1) assess the value of the addition of a set of inflammatory and endothelial biomarkers to clinical data when predicting the infectious etiologies of abnormal vital signs in Emergency Department (ED) patients. The authors found that the combination of interleukin-6 and soluble E-selectin, the initial temperature, and a history of fever improved identifying infection. We agree with their conclusion that in future studies clinical data should be integrated into prediction models of infection. Recently, we reviewed the literature on biomarkers in sepsis and wondered whether adding the relatively new biomarkers for sepsis and multiorgan failure (MOF), tissue inhibitor of metalloproteinase-2, angiopoietin-2 (Ang-2), neutrophil gelatinase-associated lipocalin (NGAL), kidney injury molecule-1 (KIM-1), and insulin-like growth factor-binding protein-7 (IGFBP-7) to routine laboratory determinations could help to identify severity of infection, complications, need for ICU admission, and outcome. In patients admitted to our university medical center-based ED with a clinical suspection of infection, venous blood samples for biomarkers were collected at admittance to the ED, in combination with blood for routine laboratory analysis. A total of 94 patients were included: 40 patients with sepsis, 51 with severe sepsis, and 3 patients with septic shock.
Lactate and blood urea nitrogen (BUN) were the routine markers that distinguished best between sepsis and severe sepsis (P < 0.001 and P = 0.002): lactate accounted for a large (r = −0.51) effect size and BUN for a medium (r = −0.32) effect size. Creatinine was a predictor for MOF. C-reactive protein was a predictor for both respiratory failure and admittance to the ICU. The biomarkers KIM-1 and IGFBP-7 significantly distinguished between sepsis and severe sepsis (P = 0.001 and P = 0.015): KIM-1 showed a medium (r = −0.35) effect size and IGFBP-7 a small-to-medium (r = −0.25) effect size. IGFBP-7 was the only biomarker that predicted for the primary endpoint, 6-month mortality. Plasma NGAL was the best biomarker to predict MOF. The only biomarker to predict acute kidney injury was IGFBP-7. Ang-2 was both a predictor for respiratory failure and admittance to the ICU.
The review article by Pierrakos and Vincent (2) lists 178 biomarkers, proposed for (diagnosis of) sepsis: almost none of them showed a sensitivity and specificity over 90%.
In conclusion, combinations of biomarkers might contribute and—in the future—may be part of (machine learning) models helping us in taking the right patient treatment and placement decisions in patients with infections in the ED. It is paramount that these biomarkers will be available fast during patients’ stay in the ED. Until that time clinical judgment and vital signs are most important in diagnosing (severity of) infection (3). In the last 10 years, the early recognition of the septic patient, reevaluating, and adjusting treatment in a timely manner using treatment bundles has improved outcome substantially (4); now it is time for fine-tuning our diagnostic modalities during acute ED admission of these large group of patients.
1. Day DE, Oedorf K, Kogan S, Novack V, Sanchez LD, Wolfe RE, Shapiro NI, Henning DJ. The utility of inflammatory and endothelial markers to identify infection in emergency department patients. Shock
2015; 44 3:215–220.
2. Pierrakos C, Vincent JL. Sepsis biomarkers: a review. Crit Care
2010; 14 1:R15.
3. Van der Vegt AE, Holman M, ter Maaten JC. The value of the clinical impression in recognising and treating sepsis patients in the emergency department. Eur J Emerg Med
2012; 19 6:373–378.
4. Ligtenberg JJM, ter Maaten JC, Zijlstra JG. Back to basics in sepsis treatment: critically ill patients need intensive care. Crit Care