Accuracy of procalcitonin in detecting severe bacterial infections among critically ill children

Kantimas Sitthikool; Chalermthai Aksilp

doi:10.4103/prcm.prcm_11_19

ORIGINAL ARTICLE

Year : 2020 | Volume : 4 | Issue : 1 | Page : 13-17

Accuracy of procalcitonin in detecting severe bacterial infections among critically ill children

Kantimas Sitthikool, Chalermthai Aksilp
Department of Medical Services, Division of Pulmonology and Critical Care, Queen Sirikit National Institute of Child Health, College of Medicine, Rangsit University, Bangkok, Thailand

Date of Submission	26-Aug-2019
Date of Decision	21-Jan-2020
Date of Acceptance	08-Jun-2020
Date of Web Publication	08-Dec-2020

Correspondence Address:
Kantimas Sitthikool
Department of Medical Services, Division of Pulmonology and Critical Care, Queen Sirikit National Institute of Child Health, College of Medicine, Rangsit University, 420/8 Rajavithi Road, Rajatevee, Bangkok 10400
Thailand

Source of Support: None, Conflict of Interest: None

Check

DOI: 10.4103/prcm.prcm_11_19

Abstract

Objectives: The aims of this study are to determine the accuracy of serum procalcitonin (PCT) in the early detection of severe bacterial infections among critically ill children and to establish the correlation between PCT changes and clinical outcomes. Design: This was a prospective, observational study at Queen Sirikit National Institute of Child Health, Bangkok, Thailand, between March 2014 and December 2014. Materials and Methods: Children aged between 1 month and 15 years with acute severe life-threatening conditions were included. Microbiologic specimens were sent for multiplex polymerase chain reaction and bacterial culture on day 1 of hospitalization. PCT was obtained on days 1, 2, 3, and 5. Measurement and Main Results: A total of 61 patients with a mean age of 21.2 months were enrolled. Microbiologic specimens were sent for multiplex polymerase chain reaction and bacterial culture on day 1 of hospitalization. PCT was obtained on days 1, 2, 3, and 5. The medians of PCT levels on days 1 and 2 from the bacterial infections group were significantly higher than those from the viral infections group and the mixed infections group. The sensitivity, specificity, and area under the PCT curve (cutoff value ≥1.1 ng/ml) employed to predict bacterial infections were 67.7%, 73.7%, and 0.72, respectively. The percentage changes of PCT levels on days 2–5 correlated with those of pediatric logistic organ dysfunction (PELOD) scores on days 1–5 but did not correlate significantly with the lengths of PICU stay. Conclusions: PCT is a moderately accurate option for the early detection of bacterial infections among children with acute severe life-threatening conditions since there is a correlation between the percentage changes of PCT levels and PELOD scores but no significant correlation between the percentage changes of PCT levels and the length of PICU stay.W

Keywords: Acute severe life-threatening conditions, critically ill children, procalcitonin, procalcitonin, severe bacterial infections

How to cite this article:
Sitthikool K, Aksilp C. Accuracy of procalcitonin in detecting severe bacterial infections among critically ill children. Pediatr Respirol Crit Care Med 2020;4:13-7

How to cite this URL:
Sitthikool K, Aksilp C. Accuracy of procalcitonin in detecting severe bacterial infections among critically ill children. Pediatr Respirol Crit Care Med [serial online] 2020 [cited 2023 May 29];4:13-7. Available from: https://www.prccm.org/text.asp?2020/4/1/13/302703

Introduction

Severe bacterial infections and sepsis are common causes of morbidity and mortality in children. These conditions are usually accompanied by clinical and laboratory signs, including fever or subtemperature, abnormal heart and respiratory rates, and leukocytosis. However, these manifestations may be noninfectious and traumatic in origin and are neither specific nor sensitive to bacterial infections or sepsis.^[1-4]

Traditional diagnosis includes assessment of clinical symptoms and culturing techniques. However, although a positive culture from blood, bronchoalveolar lavage, and cerebrospinal fluid generally leads to a reliable result, certain limitations, such as false negatives and timing delay in diagnosis, may arise.^[5] To deal with such issues, biomarkers have been used to diagnose and monitor the evolution of infection processes.

Serum procalcitonin (PCT) is a 114–116-amino acid and a propeptide of calcitonin produced in the C-cell of the thyroid gland. This biomarker of bacterial infections correlates with the extent and severity of microbial invasion in several types of infection. Thus, PCT levels are very low (<0.1 ng/ml) in healthy humans.^[6],[7] In contrast, bacterial infections induce systemic inflammation reactions, elevating plasma PCT within 2–6 h.^[8],[9]

The objective of this study is to determine the accuracy of PCT in the early detection of severe bacterial infections among critically ill children. In addition, it seeks to establish the correlation between PCT changes and clinical outcomes.

Materials and Methods

This research was a prospective, observational study conducted on patients with acute severe life-threatening conditions admitted in the Queen Sirikit National Institute of Child Health, Bangkok, from March 2014 to December 2014. The patients included in the study met the following criteria: (1) children aged 1 month to 15 years and (2) a diagnosis of an acute severe life-threatening condition, namely acute lower respiratory infection (ALRI) with acute respiratory failure, severe sepsis, septic shock, myocarditis, and severe encephalitis. In contrast, the patients excluded from the research were those with an underlying disease, such as congenital heart disease, chronic lung disease, and immunodeficiency. The study was approved by the institutional review board.

Procalcitonin measurement

The measurement of the patients’ PCT concentrations was conducted in the laboratory using BRAHMS PCT immunofluorescent assays. PCT levels were initially measured on day 1 of admission, before being assessed daily on days 2, 3, and 5. The research personnel performing the measurement were blinded to the clinical information.

Pathogen testing

On day 1 of hospitalization, microbiological specimens were obtained from the following sources: (1) serum, (2) nasal or throat swab, (3) tracheal suction (in the event of intubated patients), and (4) cerebrospinal fluid (in the event of suspected central nervous system infection), for the detection of bacteria, viruses, and fungi using multiplex polymerase chain reaction (PCR). In addition, a bacterial culture could be dispatched for the purpose of routine clinical care at the discretion of the attending clinicians.

Definitions and outcome measures of different pathogen groups

The patients were classified into one of the following pathogen groups based on their test results: (1) bacterial infections, referring to the detection of bacteria from any sterile sources; (2) viral infections, referring to the detection of viruses without any co-detection of bacteria; (3) mixed infections, referring to the detection of both bacterial and viral infections or (4) negative, referring to the detection of no pathogens.

Statistical analysis

Statistical analysis was performed using SPSS 16.0 (SPSS Inc., Chicago, IL, USA) to compare the demographic characteristics and laboratory measurements of the subjects across four groups. For continuous variables, a Chi-square was employed. For two-group comparisons (nonnormal distribution), the Mann–Whitney test and the Kruskal–Wallis test were utilized. Group differences associated with a P ≤ 0.05 were considered statistically significant. The test accuracy of PCT in identifying bacterial and mixed infection groups was reviewed using a receiver operating characteristic (ROC) curve analysis. The sensitivity, specificity, PPV, and NPV were calculated to determine the PCT cutoff value. The Pearson’s correlation analysis was performed to determine the correlation of PCT changes and clinical outcomes.

Results

A total of 61 patients with a severe life-threatening condition were recruited to participate in the research. The median age of the patients was 1.8 years (0.92–4.41 years), and 58.9% were male. The mean lengths of hospital stay and PICU stay were 13.57 and 4.16 days, respectively. The overall mortality rate was 6.5%. The most common diagnosis was ALRI with respiratory failure (67.3%), followed by pneumonia with sepsis (21.3%), as shown in [Table 1]. The majority of the patients (31.2%) were classified into the viral infection group, as shown in [Table 2]. The positive results for viral infection group were detected from nasal/throat swab, tracheal suction, or serum PCR; meanwhile, bacterial infections were detected by positive results in any sterile sources. No patients had positive nasal/throat swabs for bacteria. The baseline characteristics of the four groups of patients were not statistically different, as shown in [Table 3].

Table 1: Subjects' diagnoses (n=61)

Click here to view

Table 2: Pathogen groups of the subjects

Click here to view

Table 3: Baseline characteristics of the patients by pathogen group

Click here to view

The median PCT levels on day 1 were 9.16 (1.60–100.97), 0.51 (0.15–1.18), and 1.12 (0.25–12.2) ng/ml for the bacterial infection, viral infection, and mixed infection groups, respectively. In comparison, the figures on day 2 were 18.12 (2.20–56.48) ng/ml, 0.50 (0.24–6.26) ng/ml, and 2.08 (0.67–13.61) ng/ml for the bacterial infections, viral infections, and mixed infections groups, respectively. Further analysis showed that the PCT levels associated with the bacterial infection group were significantly higher than those associated with the viral infection group and the mixed infection group on both days 1 and 2. The findings are presented in [Table 4].

Table 4: Procalcitonin levels from microbiological results by pathogen group

Click here to view

The cutoff value of the PCT levels on both days 1 and 2 was 1.1 ng/ml, with sensitivity, specificity, PPV, NPV, and AUC figures on day 1 standing at 67.7%, 73.7%, 72.6%, 68.8%, and 0.72, respectively, and on day 2 equaling 74.2%, 68.4%, 70.0%, 71.9%, and 0.74, respectively, as shown in [Figure 1]. In addition, the percentage changes of PCT levels on days 2–5 correlated with those of PELOD scores on days 1–5, as shown in [Figure 2].

Figure 1: Receiver operating characteristic curves of procalcitonin days 1 and 2 for the bacterial infections group and the mixed infection group combined.

Click here to view

Figure 2: Correlation between the percentage changes of procalcitonin levels and PELOD scores.

Click here to view

Discussion

This prospective, observational research aims to assess the accuracy of PCT in detecting the early stages of severe bacterial infections among critically ill children and to determine the correlation between PCT changes and clinical outcomes. All patients in this study presented with life-threatening condition. At that time, we did not know the cause of diseases. Furthermore, clinical manifestations cannot distinguish between bacterial and viral infection. Therefore, we obtained serum PCT in every patient who presented with severe life-threatening condition. It was found that the accuracy of PCT was moderate with the cutoff value of ≥1.1 ng/ml and sensitivity, specificity, and AUC figures equaling 67.7%, 73.7%, and 0.72, respectively. Subgroup analysis was analyzed in patients aged ≤2 years, and we found that PCT is not more useful in this population. Such findings resonate with those discovered in previous studies. A case in point is Moulin et al.’s work, in which PCT >1 ng/ml could differentiate between viral and bacterial community-acquired pneumonia (AUC = 0.93).^[10] In addition, PCT ≥2.5 ng/ml could be employed to predict systemic inflammatory response syndrome (SIRS) with moderate accuracy (sensitivity = 68%, specificity = 74%, AUC = 0.71).^[11] In another study, higher PCT levels were found to be higher among patients in the bacterial infection group (cutoff ≥0.5 ng/ml, sensitivity = 67.8%, specificity = 80.4%, AUC = 0.764).^[12] Although a sizeable number of critical ill patients were included in this research, most received antibiotics before the PCT collection process, leading to reductions in PCT levels, and hence lower PCT accuracy and sensitivity. These findings suggest that cutoff, sensitivity, and specificity values are likely to vary across studies depending on the population, the timing of PCT measurement, and the severity scores associated with patients. In this study, we also found low PCT levels in mixed infection group due to the fact that PCT is released in response to bacterial infections via a direct stimulation of cytokines, such as interleukin (IL)-1b, tumor necrosis factor-alpha, and IL-6. On the other hand, interferon-g, a cytokine released in response to viral infections, blocks the upregulation of PCT, resulting in low PCT in viral infection. Therefore, PCT levels in mixed infection group were not as high as bacterial infection group. However, it can be seen that PCT levels associated with the mixed infection group were higher than those associated with the viral infection group on both days 1 and 2.

As for the correlation between PCT levels and PELOD scores, in this study, the percentage changes of PCT levels on days 2–5 were found to correlate with the percentage changes of PELOD scores on days 1–5 but insignificantly with the lengths of PICU stay. Similar results were reported by Qi, who discovered a poor prognosis in pediatric sepsis patients with high PCT levels and mild pediatric clinical illness scores.^[13] Based on these findings, it can thus be presumed that the incorporation of the percentage changes of both PCT levels and PELOD scores is vital for enhancing the prediction of an adverse outcome.

Owing to its satisfactory accuracy, PCT has been widely used as an indicator of bacterial infections. In the present research, the PCT levels caused by bacterial infections were significantly higher than those resulting from viral and mixed infections. In addition, PCT was also proven to increase markedly among pediatric patients with sepsis and bacterial sepsis patients.^[14-17] Further, increased PCT with a positive pneumococcal urinary antigen was found to correlate with pneumococcal community-acquired pneumonia.^[18] In addition, meta-analysis findings demonstrate that a PCT assay is a potentially effective test for distinguishing between bacterial and viral meningitis in pediatric patients.^[19] Finally, PCT is regarded as an accurate biomarker for telling bacterial and viral pneumonia apart.^[20]

Despite the foregoing discussion, it is worth mentioning several limitations to the present study. First, a subgroup analysis could not be conducted because of the small subgroup size. Second, the lower respiratory tract specimens were collected through tracheal suction, contributing to low sensitivity and specificity in pathogen detection as well as possible upper respiratory tract contamination. Third, a urine culture was not collected from all the patients, potentially undermining the diagnosis of urinary tract infection. Fourth, multiplex PCR is not pertinent to all respiratory pathogens, such as Chlamydia trachomatis and Streptococcal gr A. Finally, it was not possible to determine the correlation between PCT levels and mortality rates in this research since the latter were insignificant.

Conclusions

PCT levels associated with the bacterial infection group were significantly higher than those associated with the viral infection and the mixed infection group. PCT is a moderately accurate option for the early detection of bacterial infections among children with acute severe life-threatening conditions since there is a correlation between the percentage changes of PCT levels and PELOD scores.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

References

1.	Steinberg W, Tenner S. Acute pancreatitis. N Engl J Med 1994;330:1198-210.
2.	Moore FA, Sauaia A, Moore EE, Haenel JB, Burch JM, Lezotte DC. Postinjury multiple organ failure: A bimodal phenomenon. J Trauma 1996;40:501-10.
3.	Saffle JR, Sullivan JJ, Tuohig GM, Larson CM. Multiple organ failure in patients with thermal injury. Crit Care Med 1993;21:1673-83.
4.	Reinhart K, Karzai W, Meisner M. Procalcitonin as a marker of the systemic inflammatory response to infection. Intensive Care Med 2000;26:1193-200.
5.	Carrigan SD, Scott G, Tabrizian M. Toward resolving the challenges of sepsis diagnosis. Clin Chem 2004;50:1301-14.
6.	Snider RH Jr., Nylen ES, Becker KL. Procalcitonin and its component peptides in systemic inflammation: Immunochemical characterization. J Investig Med 1997;45:552-60.
7.	Jacobs JW, Lund PK, Potts JT Jr., Bell NH, Habener JF. Procalcitonin is a glycoprotein. J Biol Chem 1981;256:2803-7.
8.	Dandona P, Nix D, Wilson MF, Aljada A, Love J, Assicot M, et al. Procalcitonin increase after endotoxin infection in normal subjects.J Clin Endocrinol Metab1994;79:1605-8.
9.	Brunkhorst FM, Heinz U, Forycki ZF. Kinetics of procalcitonin in iatrogenic sepsis. Intensive Care Med 1998;24:888-9.
10.	Moulin F, Raymond J, Lorrot M, Marc E, Coste J, Iniguez JL, et al. Procalcitonin in children admitted to hospital with community acquired pneumonia. Arch Dis Child 2001;84:332-6.
11.	Simon L, Saint-Louis P, Amre DK, Lacroix J, Gauvin F. Procalcitonin and C-reactive protein as markers of bacterial infection in critically ill children at onset of systemic inflammatory response syndrome. Pediatr Crit Care Med 2008;9:407-13.
12.	Oshita H, Sakurai J, Kamitsuna M. Semi-quantitative procalcitonin test for the diagnosis of bacterial infection: Clinical use and experience in Japan. J Microbiol Immunol Infect 2010;43:222-7.
13.	Qi YZ. Prognostic values of serum procalcitonin level and pediatric critical illness score in children with sepsis. Zhongguo Dang Dai Er Ke Za Zhi 2014;16:190-3.
14.	Enguix A, Rey C, Concha A, Medina A, Coto D, Diéguez MA. Comparison of procalcitonin with C-reactive protein and serum amyloid for the early diagnosis of bacterial sepsis in critically ill neonates and children. Intensive Care Med 2001;27:211-5.
15.	Alizadeh N, Memar MY, Moaddab SR, Kafil HS. Aptamer-assisted novel technologies for detecting bacterial pathogens. Biomed Pharmacother 2017;93:737-45.
16.	Lopez AF, CubellsCL, GarcíaJJ, Pou JF. Procalcitonin in pediatric emergency departments for the early diagnosis of invasive bacterial infections in febrile infants: results of a multicenter study and utility of a rapid qualitative test for this marker. Pediatr Infect Dis J 2003;22:895-903.
17.	Memar MY, Varshochi M, Shokouhi B, Asgharzadeh M, Kafil HS. Procalcitonin: The marker of pediatric bacterial infection. Biomed Pharmacother 2017;96:936-43.
18.	Galetto-Lacour A, Alcoba G, Posfay-Barbe KM, Cevey-Macherel M, Gehri M, Ochs MM, et al. Elevated inflammatory markers combined with positive pneumococcal urinary antigen are a good predictor of pneumococcal community-acquired pneumonia in children. Pediatr Infect Dis J 2013;32:1175-9.
19.	Henry BM, Roy J, Ramakrishnan PK, Vikse J, Tomaszewski KA, Walocha JA. Procalcitonin as a serum biomarker for differentiation of bacterial meningitis from viral meningitis in children: Evidence from a meta-analysis. Clin Pediatr (Phila) 2016;55:749-64.
20.	Zhu F, Wei H, Li W. Significance of serum procalcitonin levels in differential diagnosis of pediatric pneumonia. Cell Biochem Biophys 2015;73:619-22.