Clinical Sampling Challenges in Diagnosing Severe Respiratory Viral Infections

Background: Respiratory viruses are a primary cause of pneumonia in both children and adults, particularly in cases of severe acute respiratory infection (SARI) requiring hospitalization. Although molecular diagnostic technologies have clearly enhanced detection capacity, the selection process of the sampling site substantially influences diagnostic accuracy. Research question: How do diagnostic yield and concordance for respiratory viruses differ between paired upper and lower respiratory tract samples in patients with SARI? Study design and methods: A multicenter, retrospective study was conducted across 9 Italian hospitals. The analysis is based on molecular diagnostic data obtained from paired lower respiratory tract (LRT) and upper respiratory tract (URT) samples collected between 2014 and 2022 and analyzed using multiplex polymerase chain reaction (PCR) assays for respiratory virus detection. Viral detection rates, concordance, and discordance were compared across sample types and viruses, as were quantification cycle (Cq) values obtained by real-time PCR assay (as a proxy for viral load). Results: Of the 346 paired samples, 67.9% were concordant and 32.1% were discordant. A virus was detected in the LRT but not in the URT samples in 21.7% of cases. The most prevalent viruses detected were influenza viruses type A (IAV), rhinovirus (RV)/enterovirus (EV), and respiratory syncytial virus (RSV). Median Cq values were significantly lower in LRT samples (24.5 vs 27.5 in URT; P < .001), indicating higher viral loads. In 67.3% of concordant sample pairs, viral loads were higher in the LRT sample, with more than a third exceeding a 10-fold (1 Log10) difference. Interpretation: Sampling only the URT may miss significant viral infections in patients with SARI. Sampling the LRT improves diagnostic sensitivity, particularly in later stages of infection, and should be considered in diagnostic protocols to enhance clinical decision-making.