VHA Study Found That Technique More Sensitive
U.S. Army Pfc. Alandes T. Rowsey, a combat medic specialist assigned to the 223rd Engineer Battalion of the Mississippi National Guard administers a COVID-19 test at the Mississippi State Veterans Home in Kosciusko, MS, in May 2020. Army National Guard photo by Spc. Jovi Prevot
WASHINGTON — Identifying SARS-CoV-2 infections quickly and accurately is critical to managing the COVID-19 pandemic and reducing community spread of the virus. Nasopharyngeal swabs are considered the “gold standard” for diagnosing SARS-CoV-2 infections, but collecting those samples, which involve guiding a long swab through the nostrils to the nasopharynx, is often impractical, given the large-scale testing required by the pandemic.
A new meta-analysis1 compares the sensitivity of nasopharyngeal swabs and anterior or mid-turbinate nasal swabs, which only penetrate about an inch into the nostrils, for initial diagnosis of SARS-CoV-2.
“Early in the pandemic, there was a shortage of nasopharyngeal swabs, so people looked for an alternative,” explained Timothy O’Leary, MD, PhD, chief research and development officer of the VHA Office of Research and Development, and corresponding author of the study, which was published in PLOS One.1
In July 2020, a letter to the editor published in the New England Journal of Medicine suggested that nasal swabs could be self-collected. “This meant that healthcare workers could remain distanced from the patients providing samples,” O’Leary said. “Though they would still wear personal protective equipment, they didn’t need to change masks, gowns, gloves, etc., for every sample.” Nasal swabs also cause less discomfort for patients than nasopharyngeal swabs, which go through the nose all the way to the back of the throat.2
While some studies have demonstrated that the accuracy of anterior nares swabs mirrors that of nasopharyngeal swabs, other studies have suggested that nasal swabs are significantly inferior at detecting SARS-CoV-2, particularly later in the course of the disease.
Relative Sensitivity
The meta-analysis, which reviewed 11 studies published between Jan. 1, 2020, and March 20, 2021, set out to establish the relative sensitivity of the two swabbing methods for detecting SARS-CoV-2 using nucleic acid amplification RT-PCR tests for initial diagnosis. The researchers focused on initial diagnosis to align with the way nucleic acid testing is currently used for identifying SARS-CoV-2 infections, rather than for monitoring COVID-19 treatment outcomes or determining whether to release patients from isolation or quarantine.
The data examined in the study supports the higher accuracy of nasopharyngeal swabs for this purpose. In every study and cohort included in the analysis, anterior nasal swabs were inferior to nasopharyngeal swabs. Estimates of nasopharyngeal swab sensitivity ranged from 79% to 100%. Only one study estimated the sensitivity of nasopharyngeal swabs at less than 90%. Sensitivity estimates for anterior nares samples ranged from 42% to 94%, with most studies showing sensitivity between 83% and 89%. However, there was no evidence that a particular type of swab construction or transportation device produced superior results.
“This is reassuring for practical reasons, since not all supplies may be available at any given time,” says Yaolin Zhou, M.D., co-author of the study and associate professor in the Department of Pathology at East Carolina University’s Brody School of Medicine in Greenville, NC.
O’Leary pointed out that the results are consistent with what experts know about the SARS-CoV-2 virus. “As COVID-19 infections progress, the peak virus concentration seems to move around the respiratory tract,” he told U.S. Medicine. “The nasopharyngeal swab will pick up virus from all of the upper respiratory tract as it goes through—nasopharynx, mid-turbinate, and anterior nares.”
Samples that contain a higher viral load also require fewer amplification cycles to deliver a positive test result. “In real-time reverse transcription polymerase chain reaction (RT-PCR) tests, as the DNA copies are made it gets detected by a PCR instrument,” according to Zhou. “If you start out with more virus, then it is more likely that, as it is amplified, it will exceed a threshold that will be called ‘positive’ by the instrument earlier in the process of amplification.”
Still, the study authors noted that nasal swabs might be reasonable alternatives, given the challenges posed by obtaining nasopharyngeal samples.
“As molecular pathologists, we were confident that nasopharyngeal swabs are better than nasal swabs. But we have to operate in the reality of practical limitations,” Zhou added.
According to O’Leary, when deciding to use a nasopharyngeal swab rather than a nasal swab, providers should ask: Will it make a difference to the patient if I use a less sensitive test right now and it comes out negative?
“If the answer is no, then one should use a nasal (anterior nares) swab,” he said. “If you are considering the use of a drug for which a positive test result is necessary, and if you can’t afford a delay if you have to test again, then it is probably worth getting the somewhat more sensitive nasopharyngeal swab.”
- Zhou Y, O’Leary TJ. Relative sensitivity of anterior nares and nasopharyngeal swabs for initial detection of SARS-CoV-2 in ambulatory patients: Rapid review and meta-analysis. PLoS One. 2021 Jul 20;16(7):e0254559. doi: 10.1371/journal.pone.0254559. PMID: 34283845; PMCID: PMC8291630.
- Tu YP, Jennings R, Hart B, Cangelosi GA, et. al. Swabs Collected by Patients or Health Care Workers for SARS-CoV-2 Testing. N Engl J Med. 2020 Jul 30;383(5):494-496. doi: 10.1056/NEJMc2016321. Epub 2020 Jun 3. PMID: 32492294; PMCID: PMC7289274.
Table 1
| Study | Study Site | Patient Characteristics | Risk of Spectrum Bias | Total Number of Subjects | Percent positive | ANS Sensitivity | NPS Sensitivity | MTS Sensitivity | Self or health care worker | Platform (LOD) |
|---|---|---|---|---|---|---|---|---|---|---|
| Federman [20] | Hospital and drive-through testing site | Ambulatory and hospitalized. Symptomatic or high exposure risk | High | 81 | 25 | 19/20 | 18/20 | Healthcare workers | Simplexa 1 COVID-19 Direct Kit 6000 NDU/mL) | |
| 95% (73–100%) | 90% (67–98%) | |||||||||
| Berenger [15] | Participant’s homes | Previously tested positive | High | 36 | 81 | 24/29 | 27/29 | Healthcare workers | LDP | |
| 83% (63–93%) | 93% (76–99%) | |||||||||
| Péré [14] | Hospital | Patients suspected for COVID19 | Unclear | 44 | 84 | 33/37 | 37/37 | Healthcare workers | Allplex 2019-nCoV assay | |
| 89% (74–96%) | 100% (88–100%) | |||||||||
| Tu [2] | Five ambulatory clinics | Ambulatory subjects with upper respiratory symptoms | Low | 498 | 10 | 48/51 | 51/52 | 50/52 | Self | Samples were sent to a reference laboratory for RT-PCR testing; testing specifics were not described |
| 94% (83–98%) | 98% (88–100%) | 96% (86–99%) | ||||||||
| Callahan [3] | Drive-through and walk-up testing site | Ambulatory patients with suspected COVID19 or previous positive | High | 308 | 32 | 47/98 | 92/98 | Healthcare workers | The Abbott Real-Time | |
| SARS-CoV-2(2700 NDU/mL) | ||||||||||
| 48% (38–58%) | 94% (87–97%) | |||||||||
| Griesemer [9]—Albany | Medical center testing tent | Ambulatory symptomatic and asymptomatic | Low | 236 | 5.2 | 5/12 | 12/12 | Healthcare workers | LDP | |
| 42% (16–71%) | 100% (70–100%) | |||||||||
| Griesemer [9]–New Rochelle | Drive-through testing site | Ambulatory symptomatic and close contacts | Low | 227 | 41 | 81/93 | 91/93 | Healthcare workers | LDP | |
| 87% (78–92%) | 98% (92–100%) | |||||||||
| Kojima [16] | Participant’s homes | Previously tested ambulatory subjects | High | 45 | 64 | 23/29 | 23/27* | Healthcare workers | Modified CDC (unmodified CDC is 18000 NDU/mL) | |
| 79% (60–91%) | 85% (65–95%) | |||||||||
| McCulloch [13] | Emergency department and drive-through testing center | Ambulatory symptomatic patients, including 27 known-positives | High | 185 | 21 | 35/38 | 31/38 | Self | LDP | |
| 92% (78–98%) | 82% (65–92%) | |||||||||
| Hanson [10] | Drive-through testing center | Symptomatic patients | Low | 354 | 24 | 70/81 | 80/81 | Self | Hologic Aptima SARS-CoV-2 | |
| TMA test (600 NDU/mL) | ||||||||||
| 86% (77–93%) | 99% | |||||||||
| (92–100%) | ||||||||||
| Pinninti [22] | Hospital | Symptomatic and asymptomatic hospitalized patients | High | 40 | 85 | 34/34 | 29/34 | Healthcare workers | Modified CDC (200 copies/mL) | |
| 100% (87–100%) | 85% (68–94%) | |||||||||
| Liu [21] | Hospital | Confirmed or highly suspected SARS-CoV-2 infection | High | 48 | 54 | 23/26 | 26/26 | Healthcare workers | Sansure Biotechnology (200 copies/mL) | |
| 88% (69–97%) | 100% (84–100%) |
*Two MT specimens were deemed as having insufficient material for analysis. Had these specimens been excluded from the computations for NPS, both would have demonstrated the same performance level (23/27).
** Only data from the AM specimen collection is included in this table.
