Procalcitonin in children with suspected novel influenza A (H1N1) infection.

Rapid diagnosis of children with flu-like symptoms is of great clinical importance. It has been shown that the discrimination of influenza from other viral and bacterial infections in children during an influenza epidemic, either based on clinical findings or on laboratory parameters, is unreliable.2, 3, 4

In a number of pediatric studies, circulating levels of procalcitonin (PCT) were raised in bacterial infection and only mildly raised in the absence of bacteria.5, 6, 7 We hypothesized that PCT might be a useful tool for the discrimination of children with bacterial infection from children with viral infection during the Novel Influenza A (H1N1) pandemic.

From October until December 2009, 30 consecutive children with suspected Novel Influenza A (H1N1) infection were included at the freely accessible (i.e., no referral needed) influenza outpatient clinic. Standard blood tests at presentation included C-reactive protein (CRP), differentiated white blood count and PCT. A pharyngeal swab for polymerase chain reaction (PCR) for Influenza A and B, rhinovirus, adenovirus, para-influenza virus 1–4, enterovirus, human coronavirus, human metapneumovirus and respiratory syncytial virus, as well as for Legionella spp., M. pneumoniae and C. pneumoniae, were standardly performed, using primers as advised by the World Health Organization. Blood and sputum cultures, lumbar punctures and radiological examinations were routinely ordered at the discretion of the treating physician.

Values are presented as numbers with percentages or as medians with interquartile ranges (IQR). Correlations between CRP, leukocytes and PCT were assessed using Spearman’s correlation test. Student’s t-test was used for inter-group comparison. P < 0.05 was considered significant.

30 Consecutive patients were included (mean age 24 months). 25 Patients (83.3%) were diagnosed with a confirmed viral illness. Three patients were diagnosed with a – clinically suspected – viral upper airway infection without positive PCR, one patient was diagnosed with a M. Pneumoniae infection and one patient was diagnosed with a non-confirmed bacterial pneumonia.

In the subgroup with confirmed viral infections, nine cases of novel influenza A (H1N1)(30%), eleven cases of rhinovirus (36.7%), five cases of respiratory syncytial virus (16.7%), four cases of adenovirus (13.3%) and one case of human metapneumovirus were diagnosed (Table 1 ).

Table 1

Detected viruses by means of PCR in children with suspected Novel Influenza A (H1N1) during the H1N1 pandemic. Total exceeds 100% due to double infections.

Virus	n
Influenza A (H1N1)	9
Rhinovirus	11
Adenovirus	4
Respiratory syncytial virus	5
Human metapneumovirus	1
Suspected viral infection	3

Median PCT values in the subgroup with confirmed viral infections were 0.153 ng/mL with IQR 0.094–0.261 ng/mL, median CRP values were 19.0 mg/L with IQR 5.6–28.3 mg/L and median leukocyte count was 13.1 giga/L with IQR 9.5–16.0 giga/L. The one case with non-confirmed bacterial pneumonia showed a marked acute phase response (PCT 32.17 ng/mL, CRP 444.1 mg/L). The negative predictive value (NPV) of PCT for bacterial infection was 88%, at a cut-off <0.71 ng/mL – the earlier described lower cut-off point in healthy neonates – whereas the NPV of CRP was 28% at a cut-off <10 mg/L and 78% at a cut-off <30 mg/L (Fig. 1 ).

Figure 1

PCT levels (ng/mL), CRP levels (mg/mL) and leukocyte counts (giga/L) in patients with a confirmed viral infection at presentation.

A significant correlation between PCT and CRP (ρ 0.531, P = 0.003), between CRP and leukocyte count (ρ 0.445, P = 0.016), but not between PCT and leukocyte count (ρ 0.041, P = 0.83) could be observed.

Our study was aimed to assess the value of PCT in children with suspected Novel Influenza A (H1N1) infection during the pandemic. Based on the results of this small cohort study, we can conclude that the acute phase response in children with Novel Influenza A (H1N1) infection and other viral infections is mild, with low levels of CRP, PCT and leukocyte count. For CRP and leukocyte count, this is in conjuncture with earlier studies in children and adults.2, 4, 8, 9 We show that values of PCT are generally below <0.71 ng/mL – the earlier described lower cut-off point in healthy neonates – in children with viral infection of any cause.

Levels of CRP were considerably low in the subgroup with viral infections and showed a significant correlation with PCT levels. However, seven patients had CRP values >30 mg/L, whereas the generally accepted lower cut-off point is <10 mg/L.

PCT measurement in children with suspected Novel Influenza A (H1N1) infection seems to be of added value. Low circulating levels of PCT appear to correlate with the presence of viral infection and/or the absence of bacterial infection. Although circulating levels of CRP show a correlation with PCT and are generally low in patients with viral infection, a considerable amount of patients had moderately elevated levels of CRP. The same holds true for leukocyte counts.

This study has some limitations. Apart from the small cohort size, a major drawback is the fact that we almost uniformly included patients with viral infections. To be able to analyze the diagnostic properties of PCT, CRP and leukocyte counts in differentiating viral and bacterial infections, a group of children with bacterial infection from the same cohort is needed. However, many earlier studies have shown that values of PCT during bacterial infection in children can be elevated.5, 6, 7

In children with flu-like symptoms, a rapid and adequate diagnosis is of great importance. Reliable biomarkers that are readily available may enhance and quicken the diagnostic process. Based on our findings, PCT may be of added value when the treating physician is uncertain whether to withhold antibiotics.

Disclosures

The authors do not have any conflict of interest to declare. No financial funding for this research was received.