Table 1.

Characteristics of different sampling methods for diagnosis of COVID-19.

Specimen	Advantages	Disadvantages	Sensitivity a	Specificity	Ref.
Nasopharyngeal swabs (NPS)	Gold standard	Supervised sample collection, requires specialized medical personnel with PPE, expensive, reflex sneezing/coughing, high risk of viral transmission, patient discomfort	98% (CI: 89–100%)	98.1% (CI: 96.5–99.0%)	[15,35,36]
Oropharyngeal swabs (OPS)	High sensitivity if performed along with NPS	Supervised sample collection, requires specialized medical personnel with PPE, expensive, highest rate of aerosol transmission, more likely to have nausea and vomit, reflex sneezing/coughing, patient discomfort	21.1% (CI: 10.5–31.6%)	97.6% (CI: 93.9–99.5%)	[15,35,37,38]
Nasal swabs	Less invasive, less expensive, self-collection, no patient discomfort	Less accurate	87.1% (CI: 79.57–93.55%)	100% (CI: 69.2–100%)	[18,39,40,41]
Saliva	Self-collection, easy to obtain, cheap, non-invasive, low rate of aerosol transmission, cost-effective, does not require healthcare workers or PPE, no patient discomfort	Relatively less sensitive than NPS	91% (CI: 80–99%)	97.6% (CI: 95.5–98.9%)	[19,20,36,42,43]
Sputum	Less invasive than NPS, painless	Not all patients can provide it	97.2% (CI: 90.3–99.7%)	90.0% (CI: 73.5–97.9%)	[26,37,41,44]

^a: sensitivity and specificity values are for real-time reverse transcriptase–polymerase chain reaction (RT-PCR) tests, and they are highly variable. Many factors can affect these values, such as time of sampling or clinical characteristics of patients (symptomatic or asymptomatic). CI: 95% Confidence Interval, PPE: Personal Protective Equipment.