| BackgroundSepsis is among the most common causes of death in hospitalized patients. Conceptually, sepsis is defined as a systemic inflammatory host response to infection and characterized by alterations in physiologic parameters such as temperature, heart rate, respiratory rate, etc. Early diagnosis of sepsis and prompt appropriate treatment lead to reduce morbidity and mortality. Every hour of delay in antibiotic administration demonstrated an increase in mortality of 7.6% in septic shock. Although blood culture (BC) is considered the standard of choice for bacteremia detection, there is no "gold standard" for the diagnosis of sepsis. Due to the time required for the implementation and interpretation of the results of blood and other bacterial cultures, delays between blood sampling and information obtained and returned to the clinicians represent an important disadvantage, with no currently available alternative. The Sepsis Occurrence in Acutely Ill Patients study revealed that approximately 40% of the sepsis patients remain culture negative. Moreover, low specificity of BC due to contamination is a problem in diagnosis of sepsis.Procalcitonin (PCT), the precursor of the hormone calcitonin, is produced by C-cells of the thyroid gland or neuroendocrine cells in the lung or intestine. It is produced in the parenchymal cells in response to microbial toxins or inflammatory mediators such as interleukin (IL)-1? and tumor necrosis factor (TNF)-a. Because up-regulation of PCT is attenuated by interferon (INF)-y, a cytokine released in response to viral infections, PCT is more specific for bacterial infections and may help to distinguish bacterial infections from viral illnesses.Numerous studies on PCT have been carried out, although few with large sample size. To deal with the complexity of sepsis, an understanding of PCT in heterogenous clnical condition is required. Thus this study was designed as a cross-sectional study with large sample size and heterogenous clinical condition.Intention1. Analyze PCT levels in patients with positive BC. Analyze their optimal cut-off value. Analyze PCT levels in patients with blood infection with Acinetobacter baumannii, Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, Enterobacter cloacae, Staphylococcus aureus, Enterococcus faecium, Enterococcus faecalis respectively.2. Analyze PCT levels in patients with positive body fluid culture. Verify whether they are statistically different with negative culture. Analyze their optimal cut-off value. Analyze PCT levels in patients with positive hydrothorax culture, positive ascites culture,, positive bile culture and positive cerebrospinal fluid culture respectively.3. Analyze PCT levels in patients with sepsis and severe sepsis. Analyze their optimal cut-off values. Verify whether PCT level was associated with the severity of sepsis.Method1. SubjectMedical records of patients hospitalized in Zhujiang Hospital (Southern Medical University, Guangzhou, China) from January 1st 2012 to June 12th 2014, from whom had PCT assay in plasma and BC or body fluid culture obtained, were retrospectively observed. Patients aged 10 to 79 were enrolled. Patient records/information was anonymized and de-identified prior to analysis. Patients were screened for if they had PCT assayed within the surrounding two days when BC (blood culture) or body fluid culture were obtained. The results were based on the first PCT sample taken within an individual hospitalization. Organisms of the BC results as coagulase-negative staphylococci, aerobic and anaerobic diphtheroids, Micrococcus sp, Bacillus sp were considered as "contaminants".2952 cases (from 2538 patients) were available in this study.2. Method2.1 Data RecordingThe records of clinical data included age, sex, PCT level and results of BC or body fluid culture. As for the cases with positive BC or body fluid culture results, further investigation on their temperature, plasma WBC (white blood cell) count and clinical diagnoses including records of sepsis, septic shock, MODS (multiple organ dysfunction syndrome) and death was carried on.2.2 Data CategorizationAll data were classified into three groups, "positive BC" group, "negative all culture" group, and "positive body fluid culture" group (negative BC). The latter group was further classified into 4 groups, "positive hydrothorax culture" group, "positive ascites culture" group, "positive bile culture" group and "positive cerebrospinal fluid culture" group. "Positive BC" group was further classified according to different bacteria. Signs of systemic inflammatory response syndrome (SIRS), such as body temperature, plasma WBC and etc. were notified in cases of a positive BC or body fluid culture. Referring to 2012 surviving sepsis campaign guidelines, sepsis is defined as the presence (probable or documented) of infection together with systemic manifestations of infection and severe sepsis is defined as sepsis plus sepsis-induced organ dysfunction or tissue hypoperfusion. In this study, cases with both bacterial infection and SIRS were categorized into "sepsis" group. Cases in "sepsis" group with MODS or septic shock were categorized into "severe sepsis" group.2.3 Technical MethodsPCT was assayed by VIDAS(?) B.R.A.H.M.S PCT assay (bioMerieux, Marcy L'Etoile, France). The principle of the assay combines a one-step immunoassay sandwich method with a final fluorescent detection (ELFA). Routinely for the blood cultures, two pairs of aerobic and anaerobic bottles were obtained and incubated for at least five days with a maximum of seven days or until positive. Blood or body fluid samples were cultured with the automatic BacT/Alert 3D system. Pathogenic bacteria were identified by the VITEK ? system.2.4 Statistical MethodData was analyzed using SPSS 20.0 statistics software. PCT values were nonparametric in our study, thus they were evaluated by median with interquartile range, and applied Mann-Whitney-Wilcoxon test on two independent samples, and Kruskal-Wallis H test on three or more independent samples for categorical analysis. A p-value of less than 0.05 was considered statistically significant. ROC curve (receiver operating characteristic curve) displayed sensitivity versus 1-specificity, and was applied to identify the optimal PCT cut-off value. AUC (area under the curves) with higher values indicates increased discriminatory ability. PLR (positive likelihood ratio) and NLR (negative likelihood ratio) were examined. For the reason that PCT measurement range of VIDAS B.R.A.H.M.S is 0.05-200ng/mL, PCT values lower than 0.05ng/mL were calculated as 0.05ng/mL, values higher than 200ng/mL were calculated as 200ng/mL.Result1. Descriptive DataA total of 2538 patients including 1592 males and 946 females were enrolled. 346 patients had more than one PCT tests. Median age of the study population was 51 (interquartile range was 28) years old.2. PCT Levels in Patients with Positive BC440 cases were in "positive BC" group and 2389 cases in "negative all culture" group. Median PCT values of them respectively were 4.53ng/mL and 0.49ng/mL (p<0.001) (Table 1, Figure 1). AUC of PCT for distinguishing positive BC was 0.713, with optimal cut-off value 1.46ng/mL (70.0% sensitivity,64.5% specificity,1.972 PLR,0.465 NLR).8 kinds of bacteremia were specifically analyzed including Acinetobacter baumannii, Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, Enterobacter cloacae, Staphylococcus aureus, Enterococcus faecium, Enterococcus faecalis (Table 2). PCT values in patients infected with Enterobacter cloacae were generally the highest (10.17ng/mL), with Klebsiella pneumoniae were second highest (8.96ng/mL). These 8 kinds of bacteremia were categorized into two groups, "gram-negative BC group" with 301 cases and "gram-positive BC group" with 86 cases. Median PCT value of them respectively were 6.99ng/mL and 2.96ng/mL (p<0.001). Median PCT value of 55 patients with fungemia was 3.43ng/mL. However, the possibility of the fungemia cases infected with bacteria on the same time cannot be excluded.3. PCT Levels in Patients with Positive Body Fluid CultureA total of 123 cases with positive body fluid cultures were classified into 4 groups:15 cases in "positive hydrothorax culture" group,56 cases in "positive ascites culture" group,14 cases in "positive bile culture" group,38 cases in "positive cerebrospinal fluid culture" group (Table 1, Figure 1). AUC of PCT for distinguishing positive body fluid culture is 0.661, with optimal cut-off value 1.32ng/mL (63.4% sensitivity,63.3% specificity,1.728 PLR,0.578 NLR). PCT values were especially high in "positive ascites culture" group (median PCT 8.32ng/mL) and "positive bile culture" group (median PCT 5.98ng/mL), but low in "positive hydrothorax culture" group (median PCT 1.39ng/mL) and lowest in "positive cerebrospinal fluid culture" group (median PCT 0.46ng/mL). Both "positive ascites culture" group and "positive bile culture" group did not differ from that of "positive BC culture" group (p>0.05). Both "positive hydrothorax culture" group and "positive cerebrospinal fluid culture" group did not differ from that of "negative all culture" group (p>0.05).4. PCT Levels in Patients with Sepsis or Severe Sepsis357 cases were classified into "sepsis" group,150 of them were classified into "severe sepsis" group. Median PCT values of them respectively were 5.63 ng/mL and 11.06 ng/mL (Table 1). AUC of PCT to predict sepsis was 0.731, with optimal cut-off value 1.07ng/mL (78.4% sensitivity,59.2% specificity,1.922 PLR,0.365 NLR). AUC of PCT to predict severe sepsis was 0.790, with optimal cut-off value 1.12ng/mL (91.3% sensitivity,57.7% specificity,2.158 PLR,0.151 NLR) (Figure 2). No statistical difference of PCT values between "positive BC" group and "sepsis" group was seen (p>0.05). PCT values in "severe sepsis" group were significantly higher than "positive BC" group and "sepsis" group (both p<0.05).22 cases were recorded dead according to clinical diagnosis. The median PCT value was 9.50ng/mL (Table 1). AUC of PCT to predict death was 0.583, with optimal cut-off value 7.64ng/mL (59.1% sensitivity,62.1% specificity, PLR 1.559, NLR 0.659).Conclusion1. PCT is able to indicate positive BC.1.46ng/mL was suggested as cut-off to predict positive BC (70.0% sensitivity,64.5% specificity). A case with a very high PCT value suggests more likely of gram-negative bacteria blood infection.2. Different PCT levels could be related to different infection sites. Despite the possibility of blood infection, high PCT levels are more likely to indicate abdominal infection or biliary infection, while low PCT levels are more likely to indicate thoracic infection or CNS (central nervous system) infection.3. PCT assay is an effective tool to help to diagnose sepsis. The optimal cut-off value for sepsis in our study was 1.07ng/mL (78.4% sensitivity,59.2% specificity). Different PCT levels could be related to different severity of sepsis. |
PDF Full Text Request