Capillary and venous bilirubin values. Are they really different?

We measured total serum bilirubin values in paired capillary and venous samples from 79 untreated jaundiced newborn infants (group 1) and in 29 infants who were receiving phototherapy (group 2). While bilirubin values from the two sites correlated significantly for both groups, capillary samples underestimated venous bilirubin values when the latter exceeded 170 mumol/L (10 mg/dL) (mean and 95% confidence limits: group 1, -15.1 mumol/L [-0.9 mg/dL] and -24.7 to -5.5 mumol/L [-1 to -0.3 mg/dL]; group 2, -10.3 mumol/L [-0.6 mg/dL] and -17.1 to -3.4 mumol/L [-1 to -0.2 mg/dL]). Furthermore, capillary samples underestimated venous bilirubin levels by more than 17 mumol/L (1 mg/dL) in eight of 16 group 1 patients and five of 18 group 2 patients when venous bilirubin values exceeded 170 mumol/L (10 mg/dL). Lower capillary values at higher bilirubin levels might be due to the influence of environmental light. As clinical treatment decisions may be made on the basis of differences in serum bilirubin level of about 17 mumol/L (1 mg/dL) and as capillary samples may underestimate venous bilirubin levels by a similar amount, it may be prudent to measure venous rather than capillary bilirubin levels when the total serum bilirubin level exceeds 170 mumol/L (10 mg/dL).     

Use of buffy-coat smears in the diagnosis of septicaemia in children

Gram-stained buffy-coat smears from venous and capillary blood samples in 105 children suspected of septicaemia were examined for the presence of bacteria and the results compared with blood culture isolates. Gram-positive and Gram-negative bacteria were identified in 18 venous (44%) and 19 capillary (46%) buffy-coat preparations in 41 instances where bacterial organisms were isolated from the blood cultures. It is concluded that the examination of buffy-coat smears for bacteria in children suspected of septicaemia is a useful adjunct to blood cultures and, in areas where no facilities exist for culture of blood, may be a simple and rapid method of establishing the diagnosis of bacteraemia in suspected patients.     

Monitoring long-term therapy with theophylline preparations in children with asthma

In 78 children (4 to 17 years of age) with moderate or severe asthma who were additionally treated with sustained-release theophylline preparations, different ways of drug monitoring were examined. Analysis of plasma and saliva theophylline was performed by means of high performance liquid chromatography. Saliva theophylline turned out to permit a reliable prediction of plasma theophylline, if an individual regression is calculated for each patient, basing on 3 simultaneously performed measurements of theophylline levels in saliva and plasma within the therapeutic range of 8 to 20 mg/l. In 25 patients theophylline levels were determined in venous and capillary blood. There was an excellent agreement (r = 0.97). Thus, a convenient monitoring of theophylline treatment in children is possible.     

Red blood cell distribution width is not a reliable biomarker for low iron stores in children with cystic fibrosis

Low iron stores in children, absolute iron deficiency (AID), can lead to impaired neurodevelopment and requires iron therapy. In the presence of infection/inflammation, like in cystic fibrosis (CF), serum ferritin (SF) is not a reliable biomarker for AID. Red blood cell distribution width (RDW) is a promising alternative reported not to be influenced by infection in healthy children. Currently, there are no data on the diagnostic capacity of RDW to detect AID in pediatric CF patients. This was a prospective observational study that investigated iron status biomarkers in 53 Dutch pediatric CF patients. AID was defined using World Health Organization criteria for SF in stable patients (no recent pulmonary exacerbation) and C-reactive protein (CRP) ≤10 mg/l. Patients with AID had higher RDW levels than patients without AID (p = 0.019). An RDW ≥13.2% showed the following test statistics: sensitivity 100%; specificity 39.4%; positive predictive value 20%; and negative predictive value 100%. Furthermore, we found a correlation between RDW and CRP in the total group that originated from the stable patients (r = 0.308; p = 0.042). In conclusion, the diagnostic capacity of RDW for detecting AID in pediatric CF patients seems limited because RDW levels might also be influenced by chronic infection/inflammation in these patients.     

C reactive protein in the evaluation of febrile illness.

We studied prospectively 154 febrile children to determine the diagnostic value of the quantitative serum C reactive protein concentrations (CRP). Children with acute otitis media, acute tonsillitis, or treated with antibiotics during the two previous weeks and infants less than 2 months of age were excluded. Ninety seven children were from private paediatric practice and 57 were patients who had been admitted to hospital. The comparison group consisted of 75 children with confirmed bacterial infections whose CRP values were recorded retrospectively. In the study group 35 (23%) children had a confirmed viral infection, 92 (59%) had a probable viral infection as judged from the clinical picture and outcome of the illness, and 27 (18%) had a bacterial or probable bacterial infection. When the duration of the disease was more than 12 hours and the CRP value less than 20 mg/l, all children had a confirmed or probable viral infection. Nine children (one from the study group and eight from the comparison group) were found to have a septic infection and a CRP value of 20 mg/l or less. In all these cases, however, the duration of the symptoms was less than 12 hours. In addition CRP less than or equal to 20 mg/l was found in five (14%) children with urinary tract infection in the comparison group. CRP values of 20-40 mg/l were recorded in children with both viral and bacterial infections. A CRP value greater than or equal to 40 mg/l detected 79% of bacterial infections with 90% specificity. Our data show that determination of serum CRP concentrations is a valuable tool in evaluating children who have been ill for more than 12 hours.     

Capillary (heelstick) versus venous blood sampling for the determination of glucose concentration in the neonate.

Various sites may be used to obtain blood (plasma) for the determination of the glucose concentration in the neonate. Because multiple sites may be sampled in the same neonate, it is important to determine the variability in blood glucose concentration which may result from such sampling. Since pain and mechanical forces may be different because of the method used to obtain the capillary (heelstick) blood compared to the venous specimen, the two sites were sampled, and the blood glucose concentration was determined simultaneously in 25 asymptomatic well neonates whose mean birth weight was 2,562 +/- 152 g and whose gestational age was 35.5 +/- 1.5 weeks. There was a significant (p less than 0.0001), but relatively weak correlation (r2 = 0.64) between capillary (heelstick) blood and venous blood relative to blood glucose concentration. When the capillary (heelstick)-venous glucose concentration difference was compared to the mean of the capillary (heelstick) and venous glucose concentrations, a difference of +/- 0.5 mM (9 mg/dl) was noted in 3 of 25 neonates. Appropriately obtained capillary (heelstick) blood samples provide measurement of blood glucose concentration which are variable compared to venous samples, but which are probably not significant physiologically.     

Serum C-reactive protein in the differential diagnosis of childhood meningitis

BACKGROUND: Although determination of serum C-reactive protein (CRP) is considered one of the most useful tests for differentiating between bacterial and aseptic meningitis, its diagnostic accuracy in comparison with other laboratory parameters is yet to be further evaluated. METHODS: A total of 192 pediatric cases, aged between 2 months and 14 years, comprising patients with bacterial meningitis (n = 66) and aseptic meningitis (n = 126), were retrospectively analyzed on the basis of data from the initial examination. The area under the best fit binormal curve of the receiver operating characteristics (Az) for CRP was determined and compared with that for several other analytic parameters, including white blood cell count and erythrocyte sedimentation rate of peripheral blood, standard cerebrospinal fluid analysis variables and the combination test (probability of acute bacterial meningitis (pABM)) derived from Hoen's model. RESULTS: Compared with each of the other variables, the Az for serum CRP (0.97 +/- 0.02) was found to be significantly greater (P < 0.01) for all except pABM (0.99 +/- 0.01; P > 0.05). False-negative cases among the CRP test results were found to have been examined too early. CONCLUSIONS: The diagnostic accuracy of a single CRP determination was found to be equivalent to that of the most effective combination test. Patients with meningitis in whom serum CRP values are determined at least 12 h after the onset of fever and are < 2 mg/dL are far less likely to have bacterial meningitis.     

C-reactive protein is useful in distinguishing Gram stain-negative bacterial meningitis from viral meningitis in children.

OBJECTIVE: To clarify to what extent Gram stain-negative bacterial meningitis can be distinguished from viral meningitis by assessment of cerebrospinal fluid (CSF) and blood indices and serum C-reactive protein (CRP) in children over 3 months of age. DESIGN: Common CSF indices, blood leukocyte counts, and serum CRP values were compared between patients with bacterial meningitis who had a positive CSF bacterial culture but a negative Gram stain and patients with viral meningitis. POPULATION: Three hundred twenty-five consecutive patients with CSF culture-proven bacterial meningitis, for whom Gram stain was negative in 55 cases, and 182 children with proven or presumed viral meningitis. RESULTS: Significant differences between patients with bacterial and viral meningitis were found in all indices with large overlap in all except serum CRP. In patients with bacterial meningitis, the mean CSF glucose concentration, protein concentration, leukocyte count, blood leukocyte count, and serum CRP were 2.9 mmol/L (52 mg/dL), 1.88 g/L, 4540 x 10(6)/L, 18.0 x 10(9)/L, and 115 mg/L; and in those with viral meningitis, mean values were 3.3 mmol/L (59 mg/dL), 0.52 g/L, 240 x 10(6)/L, 10.6 x 10(9)/L, and <20 mg/L, respectively. Of the tests investigated in this study, only serum CRP was capable of distinguishing Gram stain-negative bacterial meningitis from viral meningitis on admission with high sensitivity (96%), high specificity (93%), and high negative predictive value (99%). CONCLUSION: Exclusion of bacterial meningitis with only the conventional tests is difficult. Combined with careful physical examination and CSF analyses, serum CRP measurement affords substantial aid.     

Serum C-reactive protein concentrations in Malaysian children with enteric fever

To investigate the role of serum C-reactive protein (CRP) in the diagnosis of typhoid fever, we studied 227 febrile Malaysian children hospitalized during a 12-month period. The children were: culture-positive for Salmonella typhi (Group 1; n = 108); culture-negative but with typical clinical features of typhoid fever (Group 2; n = 60); or had non-typhoidal illness (Group 3; n = 59). Group 1 children had the highest serum CRP concentrations (geometric mean [SD range]; 43 [12-150] mg/l vs. 26 [8-85] mg/l in Group 2 and 21 [4-110] mg/l in Group 3; p < 0.001). In regression analysis, age, patient group and fever duration were independently associated with serum CRP (p < 0.05) but gender was not. In Group 1 patients, there was a significant positive association between serum CRP and Widal O and H agglutinin titres. In receiver-operator characteristic (ROC) analysis of serum CRP for Groups 1 and 2 combined, compared with Group 3, the area under the curve (AUC) was 0.65. These data show that the serum CRP is highest in culture-positive children with enteric fever and reflects the immune response to the infection in this group. Nevertheless, serum CRP had relatively low sensitivity and specificity for confirmed or clinically diagnosed typhoid fever (68 and 58 per cent, respectively at 'cut-off' concentration 30.0 mg/l), and an AUC value only moderately above that associated with no predictive power (0.5). Although of limited use as a primary diagnostic test, a raised serum CRP may still have a place as one of a range of features that facilitate assessment of a febrile child in a typhoid-endemic area.     

Acute phase proteins as markers of systemic illness in acute diarrhoea

Fifty-seven Tanzanian children, 6-25 months, hospitalized with acute diarrhoea were grouped according to whether there was clinical evidence of systemic infection (SI) (n = 35) or not (n = 22). Serum acute phase proteins were measured in samples taken within 48 h of admission. Means for C-reactive protein (CRP) and serum amyloid A (SAA) were significantly higher in children with SI compared to those without (geometric means (95% CI); CRP, mg/l: 22.1 (13.6-35.5) vs. 7.4 (4.4-12.4); SAA, mg/l: 12.2 (6.8-22.1) vs. 4.9 (2.5-9.7)). Levels of alpha1-acid glycoprotein were similar in both groups (1.16 g/l (0.95-1.43) vs. 1.04 (0.83-1.29), respectively). CRP > or =30 mg/l had a positive predictive value of 95%, and specificity of 96% for correctly identifying SI, but a low sensitivity (51%) and negative predictive value (55%). Clinical outcome of diarrhoea was worse in children with SI: more needed intravenous fluids (23% vs. 5%), the duration of diarrhoea was longer (59.4 vs. 34.2 h) and mortality was higher (6% vs. 0%). APPs were not found to be useful markers of systemic illness in acute diarrhoea in this population.     

C-reactive protein (CRP) in early diagnosis of neonatal septicemia.

The usefulness of CRP in early detection of neonatal septicemia/meningitis and urinary tract infection was studied in a neonatal unit using a semiquantitative latex-agglutination as a rapid screening method, and electroimmuno assay as reference method for CRP determination. In 94% of non-infected infants CRP was less than or equal to 15 mg/l and 82% had CRP less than 10 mg/l up to 3 days of age. After 3 days of age 96% had CRP less than 10 mg/l. The initial CRP level was increased in 16 out of 18 patients (89%) with bacterial septicemia. Low CRP was seen in one patient with total agranulocytosis and septicemia from Streptococcus type B and in one patient with Staphylococcus albus sepsis. A rise in CRP was also seen in very pre-term infants with septicemia. Increased initial CRP was uncommon in neonatal urinary tract infection (2 of 9), but a rise was seen in 3 additional patients. A comparison between CRP, total neutrophil blood cell count and band neutrophil count as diagnostic parameters was in favour of CRP at this early stage of infection. CRP is of definite value as an aid in early diagnosis of neonatal septicemia and bacterial meningitis.     

Evaluation of the capillary microhematocrit as a screening test for anemia in pediatric office practice

The capillary microhematocrit test is widely used to screen pediatric patients for anemia. Recently, it has been suggested that this method produces spuriously elevated values compared with venous hematocrits measured by a Coulter electronic counter and might consequently fail to detect children who are truly anemic. To address this issue we studied 66 white children 9 months to 14 years of age whose capillary hematocrits were either below, equal to, or one or two points above the lower limit of normal for age. Venous specimens were obtained simultaneously with the capillary sample; hemoglobin, hematocrit, and mean corpuscular volume results were obtained using a Coulter electronic counter. Using published standards of venous hemoglobin, we determined the sensitivity, specificity, and predictive values of the capillary microhematocrit in this population of patients with low or borderline values. Twenty of the 66 patients had venous hemoglobin values less than the lower limit of normal. The sensitivity of the microhematocrit was 90.0%; the specificity was 43.5%. The predictive values for a normal (negative) hematocrit was 90.1%; the predictive value for a low (positive) hematocrit was 40.9%. We conclude that the microhematocrit method using capillary blood will miss very few patients with significantly low venous hemoglobin values and is thus an acceptable screening test for anemia. Because it does not require expensive equipment or special skill to obtain the specimen or perform the test, it is ideal for physicians' offices or nonhospital-based clinics.     

Comparison of EML 105 and Advantage analysers measuring capillary versus venous whole blood glucose in neonates

AIM: Near-patient blood glucose monitoring is an essential component of neonatal intensive care but the analysers currently used are unreliable and inaccurate. The aim of this study was to compare a new glucose electrode-based analyser (EML 105) and a non-wipe reflectance photometry method (Advantage) as opposed to a recognized laboratory reference method (Hexokinase). We also investigated the effect of sample route and haematocrit on the accuracy of the glucose readings obtained by each method of analysis. METHODS: Whole blood glucose concentrations ranging from 0 to 3.5 mmol/l were carefully prepared in a laboratory setting and blood samples from each respective solution were then measured by EML 105 and Advantage analysers. The results obtained were then compared with the corresponding plasma glucose reading obtained by the Hexokinase method, using linear regression analysis. An in vivo study was subsequently performed on 103 neonates, over a 1-y period, using capillary and venous whole blood samples. Whole blood glucose concentration was estimated from each sample using both analysers and compared with the corresponding plasma glucose concentration estimated by the Hexokinase method. Venous blood was centrifuged and haematocrit was estimated using standardized curves. The effect of haematocrit on the agreement between whole blood and plasma glucose was investigated, estimating the degree of correlation on a scatterplot of the results and linear regression analysis. RESULTS: Both the EML 105 and Hexokinase methods were highly accurate, in vitro, with small proportional biases of 2% and 5%, respectively. However, in vivo, both study analysers overestimated neonatal plasma glucose, ranging from at best 0.45 mmol/l (EML 105 venous) to 0.69 mmol/l (EML capillary). There was no significant difference in the agreement of capillary (GD = 0.12, 95% CI, [-0.32,0.08], p = 0.2) or venous samples (GD = 0.05, 95% CI. [0.09, 0.19], p = 0.49) with plasma glucose when analysed by either study method (GD = glucose difference between study analyser and reference method) However, the venous samples analysed by EML 105 estimated plasma glucose significantly better than capillary samples using the same method of analysis (GD = 0.24, 95% CI. [0.09,0.38], p < 0.01). The relationship between haematocrit and the resultant glucose differences was non-linear with correlation coefficients of r = -0.057 (EML 105 capillary), r = 0.145 (EML 105 venous), r = -0.127 (Advantage capillary) and r = -0.275 (Advantage venous). There was no significant difference in the effect of haematocrit on the performance of EML 105 versus Advantage, regardless of the sample route. CONCLUSION: Both EML 105 and Advantage overestimated plasma glucose, with no significant difference in the performance of either analyser, regardless of the route of analysis. Agreement with plasma glucose was better for venous samples but this was only statistically significant when EML 105 capillary and venous results were compared. Haematocrit is not a significant confounding factor towards the performance of either EML 105 or Advantage in neonates, regardless of the route of sampling. The margin of overestimation of blood glucose prohibits the recommendation of both EML 105 and Advantage for routine neonatal glucose screening. The consequences include failure accurately to diagnose hypoglycaemia and delays in the instigation of therapeutic measures, both of which may potentially result in an adverse, long-term, neurodevelopmental outcome.     

C-reactive protein and bacterial infection in preterm infants

Serum C-reactive protein (CRP) concentration was measured by a new solid phase ligand-binding radiometric monoclonal antibody immunoassay in a prospective study of 193 consecutively born preterm infants. In 104 with no clinical or laboratory evidence of infection the median CRP in cord serum was 0.125 mg/l (range 0.011–6.0 mg/l), at 24 h it was 1 mg/l (0.016–7.0) and at 48 h 2 mg/l (0.400–8.0). The present highly sensitive assay has enabled these normal ranges to be defined for the first time, at levels below the threshold of non-labelled immunoassays and of all commercially available CRP assays. The values in cord serum were significantly lower than in normal healthy adults (median 0.8 mg/l, range 0.07–29 mg/l,n=468) [20]. Arterial catheterisation and endotracheal intubation, in the absence of infection, did not appear to elevate CRP, nor did cerebral germinal layer or intraventricular haemorrhage. Among nine infants with confirmed septicaemia eight had a serum CRP level raised at least once during the first 48 h and serum CRP in the other one increased 250-fold in 24 h before treatment was started. Using this assay, serum CRP is a useful and rapidly available adjunct to clinical assessment in diagnosis and exclusion of bacterial infection in the early neonatal period, has encouraged us to withhold or discontinue antibiotics and also has a role in monitoring response to treatment.     

C-reactive protein (CRP) as tumor marker in pediatric and adolescent patients with Hodgkin disease

BACKGROUND: C-reactive protein (CRP) is an acute phase protein produced in the liver. An elevated CRP is a nonspecific marker of inflammation. Additionally, it also appears to be a prognostic marker in several malignancies. Elevated CRP levels in adult patients with Hodgkin disease (HD) were reported previously. However, levels of CRP have not been evaluated in pediatric and adolescent HD patients. PROCEDURE: We analyzed CRP serum levels in 95 consecutive pediatric and adolescent patients with Hodgkin disease. CRP levels were correlated with stage, absence or presence of B symptoms, and prognosis. RESULTS: At the time of diagnosis increased serum CRP levels were found in 64 % (61/95) of the patients with a median of 21 mg/L (range: <5-211). Serum C-reactive protein levels correlated with stage and were higher in patients with B symptoms. Higher CRP levels were associated with an increased risk of relapse. CONCLUSION: In addition to soluble interleukin 2 receptor (sIL-2R) levels, CRP holds promise as a diagnostic and prognostic index and follow-up monitor in pediatric and adolescent patients with Hodgkin disease, and merits further investigation.     

C-REACTIVE PROTEIN (CRP) IN EARLY DIAGNOSIS OF NEONATAL SEPTICEMIA

Abstract. The usefulness of CRP in early detection of neonatal septicemia/meningitis and urinary tract infection was studied in a neonatal unit using a semiquantitative latex-agglutination as a rapid screening method, and electroimmuno assay as reference method for CRP determination. In 94% of non-infected infants CRP was 15 mg/l and 82 % had CRP <10 mg/l up to 3 days of age. After 3 days of age 96% had CRP < 10 mg/l. The initial CRP level was increased in 16 out of 18 patients (89%) with bacterial septicemia. Low CRP was seen in one patient with total agranulocytosis and septicemia from Streptococcus type B and in one patient with Staphylococcus albus sepsis. A rise in CRP was also seen in very pre-term infants with septicemia. Increased initial CRP was uncommon in neonatal urinary tract infection (2 of 9), but a rise was seen in 3 additional patients. A comparison between CRP, total neutrophil blood cell count and band neutrophil count as diagnostic parameters was in favour of CRP at this early stage of infection. CRP is of definite value as an aid in early diagnosis of neonatal septicemia and bacterial meningitis.     

Determination of gentamycin levels in capillary blood

Gentamycin serum levels were determined in capillary and venous blood samples of 37 children of different ages using the Immuno-Assay-Merckotest - EMIT. 97 specimens were compared with each other. Gentamycin concentrations in capillary blood were insignificantly lower than those determined in venous blood samples. There was no correlation between the hematocrit and the peak concentration of Gentamycin.     

Serum C-Reactive Protein in the Early Diagnosis of Neonatal Septicemia and Bacterial Meningitis

The usefulness of serum C-reactive protein (CRP) in the early detection of neonatal infection was studied using a special laser nephelometric apparatus (CRP-1), by which CRP concentrations could be quickly determined in the nursery, with only a small amount of serum (20 μL). Initial serum CRP concentrations of samples obtained from 90 infants suspected to have sepsis and/or meningitis were evaluated. Of the 90 infants, 25 showed culture-proven septicemia and/or bacterial miningitis, while 18 were considered to be infectious based on clinical signs and positive sepsis work-up even though cultures were negative. 47 infants had negative cultures and sepsis work-up and showed a favorable clinical courses. Statistical analysis for the evaluation of serum CRP at the level of one mg/dL was performed. False negative CRP was demonstrated in seven of 25 infants with culture-proven sepsis and/or meningitis (28%) and in 4 of 18 infants with other infections (22%). On the other hand, seven of 47 (15%) non-infected infants showed false positive results. The specificity and sensitivity of serum CRP determination were 85% and 74%, respectively, for all patients, and 85% and 72%, respectively, for patients with sepsis and/or meningitis. The sensitivity varied with the pathogens. We conclude that, while the initial CRP values alone are unsatisfactory for deciding the need for antibiotic therapy, CRP is useful in the early detection of neonatal infections, and its measurement by this new equipment should available in the nursery.     

Comparisons between serum concentrations of thyroxine and thyroxine-binding proteins in samples simultaneously obtained from capillary, peripheral vein, central vein and aorta in newborn infants.

A total number of 40 newborn infants with various maturity were studied: 13 babies without perinatal events, 19 infants recovered from transient diseases, 6 infants with idiopathic respiratory distress syndrome and 2 infants with asphyxia indicating artificial ventilation. Comparisons were performed between serum concentrations of thyroxine (T4), thyroxine-binding globulin (TBG), prealbumin (TBPA) and albumin (Alb) in capillary versus peripheral vein, aorta versus central vein and, finally, in peripheral versus central veins. In healthy infants serum T4 concentrations in capillary blood and peripheral vein did not differ significantly. Although serum concentrations of thyroid hormone-binding proteins tended to be increased in aortic compared to central venous specimens no statistically significant differences appeared. In infants in good clinical conditions serum T4, TBG, TBPA, and Alb levels were 6--8% higher in peripheral than in central veins, possibly primarily due to a hemo-concentrating effect of venous stasis. Therefore, in evaluation of the thyroid variables in newborn infants the technique of blood sampling must be considered. In most infants with idiopathic respiratory distress syndrome and in one asphyxiated baby a remarkable tendency to a low serum TBG and T4 concentration in peripheral compared to central vein samples were observed.     

The role of procalcitonin as a predictor of nosocomial sepsis in preterm infants

AIM: To assess the role of procalcitonin in detecting nosocomial sepsis in preterm infants, after the onset of clinical symptoms. SUBJECTS: 100 preterm infants, 24-36 wk of gestation, were followed from the age of 3 d until discharge. Procalcitonin and C-reactive protein (CRP) levels were measured within 3 d of sepsis workup events. RESULTS: 141 blood samples were drawn from 36 infants during 85 episodes of sepsis workup performed between 4 and 66 d of life. Of these episodes, 51 (60%) were not a result of documented sepsis and thereby served as the negative comparison group. Median procalcitonin levels were higher in the septic group compared with the non-septic group at the time of the sepsis workup (2.7 vs 0.5 ng/ml, p=0.003), at 1-24 h after the sepsis workup (4.6 vs 0.6 ng/ml, p=0.003), and at 25-48 h (6.9 vs 2.0 ng/ml, p=0.016). Using high cutoff levels, both procalcitonin (2.3 ng/ml) and CRP (30 mg/l) had high specificity and positive predictive value (97%, 91% and 96%, 87%, respectively) but low sensitivity (48% and 41%, respectively) to detect sepsis. Areas under the ROC curve for procalcitonin and CRP were 0.74 and 0.73, respectively. CONCLUSION: Procalcitonin >2.3 ng/ml or CRP >30 mg/l indicates a high likelihood for neonatal sepsis, and antibiotic therapy should be continued even in the presence of sterile cultures.