A systematic review and meta-analysis published in the Journal of Clinical Microbiology showed that nasopharyngeal (NP) swabs are the superior method for diagnosing SARS-CoV-2.
Study authors searched available literature in publication databases through October 1, 2020 for studies of alternative nucleic acid amplification testing (NAAT) for SARS-CoV-2. Their search, which turned up 1253 unique citations, included analyses of NAAT performed on saliva (n=25), nasal (n=11), oropharyngeal (n=6), or nasal/oropharyngeal swabs (n=4).
The studies of saliva comprised 4528 paired saliva and NP specimens (except 2 studies that used nasal/oropharyngeal specimens as the comparator). Positive SARS-CoV-2 saliva samples were detected among 88% of tests (95% CI, 81-93), which was similar to the results obtained from paired NP samples (94%; 95% CI, 90-98). However, the dual positivity rate indicated relatively poor test agreement (79%; 95% CI, 71-86) and significant heterogeneity (I2, 88.6%).
Stratified by saliva collection methods, studies of cough or deep throat specimens yielded higher positivity (94%; 95% CI, 87-99) than studies that did not specify collection method (86%; 95% CI, 78-92). Studies in which saliva was not preprocessed showed substantially lower positivity rates (60%; 95% CI, 49-70) than those for which the nucleic acid extraction step was performed (89%; 95% CI, 83%-94%).
The studies where oral swabs were used had a positivity rate of 84% (95% CI, 57-100), which was similar to studies in which paired NP swabs were used (88%; 95% CI, 73-98) but they showed a low dual positivity rate (68%; 95% CI, 36-93).
The researchers compared nasal midturbinate or anterior nares swabs with NP swabs in 11 studies. Midturbinate or anterior nares swabs had a lower positivity rate (82%; 95% CI, 73-90) than the NP swabs (98%; 95% CI, 96-100), had lower dual swab agreement (79%; 95% CI, 69-88), and high heterogeneity (I2, 87%). Stratified by type, anterior nares outperformed (90%; 95% CI, 84-94) nasal midturbinate (84%; 95% CI, 65-97) samples.
These nasal swab studies performed more poorly among patients with a low viral load (<1000 copies/mL; 61%; 95% CI, 40-79) than among patients with a high viral load (85%; 95% CI, 82-91).
Nasal swab tests administered by healthcare workers returned fewer positive results (68%; 95% CI, 47-86) than self-administered tests (93%; 95% CI, 85-98).
These data should be interpreted with caution because study populations and laboratory procedures varied greatly between studies. There was also little consensus on sampling methods or laboratory procedures, which should be streamlined and standardized to ensure optimal viral detection.
The study authors concluded that their review of head-to-head comparisons of SARS-CoV-2 diagnostic sampling indicated NP swab NAAT remained the most effective method and that testing for COVID-19 in saliva may be the next best alternative.
Lee RA, Herigon JC, Benedetti A, Pollock NR, Denkinger CM. Performance of saliva, oropharyngeal swabs, and nasal swabs for SARS-CoV-2 molecular detection: a systematic review and meta-analysis. J Clin Microbiol. Published online January 27, 2021. doi:10.1128/JCM.02881-20