Procalcitonin levels to predict bacterial infection in Surgical Intensive Care Unit patients
| Autor: | Cian Huei Shih, Hsiao Hui Yang, Lee Ying Soo, Jhen Da Yang, Yen Ta Huang, Guan Jin Ho, Jin You Jhan, Shin-Jie Lin, Yi Tsen Lin |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
medicine.medical_specialty
lcsh:Surgery Procalcitonin hospital surgery departments 03 medical and health sciences 0302 clinical medicine Bandemia Bacterial infections Interquartile range White blood cell medicine 030212 general & internal medicine 030219 obstetrics & reproductive medicine Leukopenia Receiver operating characteristic business.industry Incidence (epidemiology) lcsh:RD1-811 medicine.disease Surgery critical care medicine.anatomical_structure Bacteremia medicine.symptom business procalcitonin |
| Zdroj: | Formosan Journal of Surgery, Vol 50, Iss 4, Pp 135-141 (2017) |
| ISSN: | 1682-606X |
| DOI: | 10.4103/fjs.fjs_54_17 |
| Popis: | Background: Infection-induced inflammatory response might be aggravated by surgery insults. The clinical presentation of Surgical Intensive Care Unit (SICU) patients might be different from medical critically ill patients. Purpose: To evaluate the diagnostic and prognostic values of procalcitonin (PCT) to predict bacterial infection in SICU patients. Methods: We retrospectively analyzed the 2-year (2013 and 2014) records of 342 adult SICU cases with suspected bacterial infection in SICU of Hualien Tzu Chi Hospital. The past histories, the first infection-related parameters when SICU admission, culture results, infection-related laboratory examinations, and outcomes were collected. Results: Median of PCT level in patients with negative and any positive culture was 0.84 (interquartile range [IQR] 0.18–6.21) and 2.27 (IQR 0.54–9.93) ng/ml, respectively. Infection from blood, urine, and skin/soft tissue elicited significantly higher PCT levels. PCT in receiver operating characteristic (ROC) curve demonstrated the most accurate to predict bacterial infection (area under the ROC curve [AUC]: 0.61; 95% confidence interval [CI]: 0.54–0.63) and bacteremia (AUC: 0.73; 95% CI: 0.66–0.80) compared to white blood cell count, ratio of neutrophils, and neutrophil-to-lymphocyte count ratio (NLCR). Significantly higher PCT levels (4.12 ng/ml, 1.12–19.99; median, IQR) were observed in mortality cases. Higher PCT levels were significantly accompanied with higher NLCR, as well as higher incidence of leukopenia and bandemia. Using Kaplan–Meier analysis, significantly higher intrahospital mortality was observed in cases with above the cutoff PCT levels of 0.5 and 2 ng/ml cases, respectively. Conclusion: PCT is a relatively more useful tool to predict bacterial and particularly bloodstream infection compared to other infection-related parameters in routinely clinical practice. Initial PCT levels may be a prognostic factor of SICU patients with bacterial infection. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|