Nasal or throat sampling is adequate for the detection of the human respiratory syncytial virus in children with acute respiratory infections

Abstract Human respiratory syncytial virus (HRSV) is one of the most important causes of acute respiratory infections (ARI) in young children. HRSV diagnosis is based on the detection of the virus in respiratory specimens. Nasopharyngeal swabbing is considered the preferred method of sampling, although there is limited evidence of the superiority of nasopharyngeal swabs (NPS) over the less invasive nasal (NS) and throat (TS) swabs for virus detection by real‐time reverse transcription quantitative polymerase chain reaction (RT‐qPCR). In the current study, we compared the three swabbing methods for the detection of HRSV by RT‐qPCR in children hospitalized with ARI at Mahosot Hospital, Vientiane, Laos. In 2014, NS, NPS, and TS were collected from 288 children. All three samples were tested for HRSV by RT‐qPCR; 141 patients were found positive for at least one sample. Almost perfect agreements (κ > 0.8) between the swabs, compared two by two, were observed. Detection rates for the three swabs (between 93% and 95%) were not significantly different, regardless of the clinical presentation. Our findings suggest that the uncomfortable and technically more demanding NPS method is not mandatory for HRSV detection by RT‐qPCR.


| INTRODUCTION
Human respiratory syncytial virus (HRSV) is a common respiratory pathogen in children under the age of 5 years. In 2015, there were estimated to be 33.1 million new episodes of HRSV-associated acute lower respiratory infections worldwide, of which 3.2 million were hospitalized and 59 600 patients died. 1 HRSV diagnosis is based on the detection of the virus in respiratory specimens using cell culture, immunofluorescence, immunoenzymatic, or molecular assays. During the past decade, polymerase chain reaction (PCR), a fast and accurate detection tool, has been widely used in the diagnosis and is often chosen over conventional methods for the detection of respiratory pathogens. 2 Collection of nasopharyngeal swabs (NPS) is considered the preferred sampling method for the detection of respiratory viruses, 3 although it requires experienced staff and can be uncomfortable, especially for young children. There is limited evidence of the superiority of NPS over the less invasive nasal (NS) and throat (TS) swabs for virus detection by real-time PCR, with only a few studies evaluating HRSV detection in children [4][5][6][7] and two including NS. 6,7 We are not aware of studies that have compared all three sampling methods.
In 2014, we conducted a study on children (<5-year-old) hospitalized at Mahosot Hospital, Vientiane, Laos, with an acute respiratory infection (ARI). 8 Three different samples (NS, TS, and NPS) were collected from a large proportion of these patients. Since HRSV was one of the most common pathogens detected, this gave us the opportunity to compare the performance of these three sampling techniques for the detection of HRSV by real-time reverse transcription quantitative polymerase chain reaction (RT-qPCR).

| Specimen collection
From December 2013 to December 2014, 383 children younger than 5 years of age, with a clinical presentation of ARI were enrolled, as previously described. 8 At inclusion, samples were collected at the same time in the following sequence: TS, NS, then NPS. They were available for 288 (75.2%) patients who were included in this study. NS and TS were placed separately in 1 mL viral transport medium (Sigma Virocult [MWE]), Corsham, England vials. NPS was placed in 1 mL of skim-milk tryptone glucose glycerol medium (STGG), to allow subsequent bacterial and viral investigations from the same sample. 9 Virocult vials and STGG were transported to the laboratory within 2 hours in a cool box. Swabs were squeezed, and the media were aliquoted and stored at −80°C before performing the laboratory assays.

| Testing for HRSV
Nucleic acids were extracted from 100 µL of each swab medium using the Cador Pathogen 96 QIAcube HT kit (Qiagen, Hilden, Germany) following the manufacturer's instructions, with an elution volume of 90 µL. RT-qPCR for HRSVA/B detection was performed using specific primers and probes as described by Bonroy et al 10 Testing was performed following the manufacturer's instructions, using the Express One-Step Superscript qRT-PCR Universal Kit   κ Coefficient measures the agreement of the HRSV RT-qPCR results between the swabs compared two by two.
c Det rate = Detection rate of each swab for the detection of HRSV by RT-qPCR calculated over the number of patients positive in any of the three swabs.
characteristics of the patients are presented in supplemental data (  (Figure 1 and Tables 1 and S1).
Detection rates of the three sampling techniques were analyzed according to demographic and clinical patient characteristics along with their 95% CI ( Figure 1 and Table S1). No significant difference was observed between the three different swabs for any of the characteristics analyzed, even after stratification by age, gender, and clinical presentation (including with or without coryza). However, only seven HRSV positive patients presented with no coryza.
HRSV viral load was on average significantly lower in TS than in NS and in NPS (P < .001, t test). When patients were sorted by increasing TS viral load, we observed that the viral load in TS was lower than in NS and NPS for most patients, 90% and 91%, respectively ( Figure 2).

| DISCUSSION
Although it is often considered a preferred method for the detection of respiratory pathogens, our study showed that NPS was not significantly better than NS or TS for the detection of HRSV by RT-qPCR in Lao children, with almost perfect agreements (κ > 0.80) between the swabs observed. The detection rates for the three swabs (between 93% and 95%) were not significantly different. In addition, the HRSV viral load detected in NS was not significantly different to that in NPS (P > .05, t test), but significantly higher than that in TS (P < .001, t test).
Our findings are in accordance with previous publications. We also investigated whether the choice of the sampling method should be based on particular patient characteristics, such as young age, specific respiratory symptoms, or signs of severity. For this, the detection rates of the three swabs were calculated and compared within different groups of patients sharing the same characteristics.
No significant difference was observed between the three different F I G U R E 1 Detection rate of the three swabs tested for the detection of HRSV by RT-qPCR according to patient characteristics. Only the characteristics which were observed in more than 30 HRSV-positive patients are displayed. # Detection rate of each swab for the detection of HRSV by RT-qPCR calculated over the total number of positive patients (positive in at least one of the three swabs tested). •"PCV13 received" if they had received at least two doses of vaccine for children less than 1-year-old or at least one dose of vaccine for children between 1-to 2-year old. *wet season: from May to October. ■Low birth weight: defined by the World Health Organisation (WHO) as weight at birth less than 2500 g. ❖Fever: defined as body temperature 38°C or higher per axilla. ➤HRSVpositive patients = positive for HRSV by RT-qPCR for at least one of the three swabs tested. ♣Pneumonia and severe pneumonia were defined according to WHO criteria: children who presented with cough or difficulty breathing and had fast breathing (aged 2-11 months: ≥50 breaths/minute, aged 1-4 years: ≥40 breaths/min) or chest indrawing, were classified as having pneumonia; children who presented with cough or difficulty breathing and had at least one of the following criteria were classified as severe pneumonia: oxygen saturation 90% or lesser , while breathing room air, or central cyanosis; severe respiratory distress; signs of pneumonia with a general danger sign (inability to breastfeed or drink, lethargy or reduced level of consciousness, convulsions, vomiting). Children younger than 2-month old who presented with cough or difficulty breathing and fast breathing (≥60 breaths/min) were classified as severe pneumonia. HRSV, human respiratory syncytial virus; RT-qPCR, reverse transcription quantitative polymerase chain reaction swabs for any of the patient groups tested. However, most of the