Saliva is more sensitive for SARS-CoV-2 detection in COVID-19 patients than nasopharyngeal swabs
Authors: Wyllie et al.
Journal/ Pre-Print: Medrxiv
1. Increased viral load detected in patient’s saliva samples when compared with nasopharyngeal samples of the same patients
2. Self-collected saliva samples have comparable SARS-CoV-2 detection sensitivity to nasopharyngeal swabs collected by healthcare workers from mild and subclinical cases
3. There is less SARS-CoV-2 temporal variability when testing saliva from inpatients
This paper investigates the use of saliva swabbing as an alternative, safer and easier method of sampling patients for the presence of COVID-19. Using a hospitalised cohort of 44 people they show that levels of SARS-CoV-2 were generally higher in saliva than in matched nasopharyngeal swabs, and that in repeat sample collections detection was less variable in saliva. In a validation cohort of 98 asymptomatic healthcare workers detected very low levels of SARS-CoV-2 in saliva of three people, suggesting that saliva swabbing could be an appropriate, sensitive and safer method of screening asymptomatic or pre-symptomatic SARS-CoV-2 infections.
Impact for SARS-CoV-2/COVID19 research efforts
· Develop diagnostic tools for SARS-CoV-2/COVID19:
Study shows saliva sampling is at least as effective as current nasopharyngeal swabbing for picking up SARS-CoV-2 infection, but is easier to perform, and safer for healthcare professionals sampling patients as it is less likely to induce coughing or sneezing that could spread virus.
· Clinical Cohort study (e.g. drug trials)
Strengths and limitations of the paper
Novelty: Previous reports have suggested saliva may be a suitable alternative sample for testing, but this study has investigated in more patients and over time, showing at least comparable results to nasopharyngeal swabs, less variation and possibly higher sensitivity.
Standing in the field: This supports previous preliminary studies of saliva as a reliable tool in SARS-CoV-2 detection and in other respiratory infections.
Appropriate statistics: Statistics are clearly explained where used.
Viral model used: Original SARS-CoV-2 strain
Translatability: Very translatable as could be a much easier sampling method: Reduced exposure to health workers, reduced discomfort to the patients and reduced need for consumables that are affected by global shortages (PPE, swabs etc.)
Main limitations: Main focus is on hospitalised patients, but the use might be wider if it could reliably detect asymptomatic or pre-symptomatic individuals who would generally be expected to have a lower viral load. More data are required for this. Although the number of patients studied is reasonable, the number of matched samples between saliva and nasopharyngeal swabs is much lower, and again more data between matched samples would be useful.