We report here the results of our clinical and biological, multicenter cohort, performed in primary care, in France, with 36 general practitioners, who included 685 of all ages, presenting with symptoms of ARI over two epidemical seasons. We found correlations between clinical diagnosis and respiratory viruses: influenza was associated with the diagnosis of flu-like syndrome (p < 0.001), HRV with rhinitis (p < 0.05), and RSV with bronchiolitis (p < 0.001) and bronchitis (p < 0.05). Respiratory distress was associated with RSV (p = 0.002), and a cough at the inclusion was significantly not associated with the influenza virus (p = 0.009). Antibiotic prescriptions were not associated with any specific virus. By day 7, persistent cough was significantly associated with active and passive smoking (respectively p = 0.01 and p < 0.001), influenza and RSV-positive samples (p < 0.05) and an age of less than 2 years (p < 0.01).

The main characteristic of our cohort, from primary care, is the important proportion of children (about 30%). The profile of our cohort is so unique, firstly because of the age of patients (11% were between 6 and 17 years old), and the majority were adults between 18 and 64 years old (57%): this is a real-life study. Most studies are performed at hospital, on patients at the extreme ages of life, or with comorbidities, and patients between 6 and 65 years old are rarely hospitalised15. Unfortunately, our study had to be interrupted early due to the containment of the SARS-CoV2 pandemic, which is why we did not include the entire epidemic season in the second season.

The viruses responsible for ARI belong to some fifteen major species with a preferential tropism for the respiratory tract. However, most of data is hospital-based, due to the high cost of virological diagnosis. Results from primary care are rare, so the knowledge of these viruses is limited in ambulatory medicine. Around 60% of samples performed in children presenting to their GP with symptoms of ARI between 2011 and 2013 in England were positive for at least one virus12. The type of sampling used in primary care also raises concerns about reproducibility, feasibility, and ease of use. Oropharyngeal sampling is simple and easily accessible, particularly in children, and is little feared by primary care practitioners13. Nasopharyngeal swabbing, not known before the SARS-CoV2 pandemic, is now easily performed by primary care practitioners.

We noticed a high level of antibiotic prescription, specifically in young children aged from 2 to 5 years old. Surprisingly, these age groups are characterized by highest level of positive viral tests. This highlights the real value of implementing qualitative virus detection tests in primary care, to guide GPs in their prescriptions. This could be POCT (point of care tests): these tests are designed to be performed by non-professional staff (patients themselves) or paramedical staff (nurses). The value of microbiological POCT, particularly in terms of antibiotic savings, is currently being evaluated16. A review published in 2022 estimated the interest of the POCT for the C-reactive protein in primary care in patients consulting for acute respiratory infections17: the POCT could decrease significantly the prescription of antibiotics, but an algorithm seems necessary. Combined rapid diagnosis tests exists, like influenza/RSV/SARS-CoV2 rapid tests18. Compared with the results obtained with specific RT-PCR, the positive and negative concordance percentages of these rapid tests, carried out in the GP’s cab, were high19.

Some diagnoses established by the GPs were associated with respiratory viruses. Influenza viruses were significantly associated with flu-syndrome, RSV was significantly associated with bronchitis and bronchiolitis, and HRV was associated with rhinitis. So, correlations between clinical and virological diagnoses exist. These findings suggest that clinical observations alone may not reliably predict the presence of a bacterial infection that would warrant antibiotic treatment. Further research is needed to identify clinical predictors of bacterial infection in patients with ARI. This information could be promoted to the GPs, to limit antibiotics prescriptions, particularly in children20. Use of antibiotics in children do modify their respiratory microbiota, with an impact on their respiratory development21. Interestingly, RSV is responsible for other diagnosis than bronchiolitis, and is not responsible for all cases of bronchiolitis. New therapies are now available to prevent the infections due to RSV: some are designed for the infants22,23, and other designed for pregnant women24 or the elderly25. The burden of RSV in the elderly has been assessed, and active prevention is necessary26. We should keep in mind that RSV can induce respiratory distress in all ages, and its prevention should be a priority for the public health policies.