Table 5.
Summary of Findings Table of Laboratory-Based RT-PCR, Rapid RT-PCR, Rapid Isothermal NAAT, Rapid Antigen Tests, and Antibody Tests
| Test | Laboratory-based RT-PCRa | Rapid RT-PCRa | Rapid Isothermal NAATa | Rapid Antigen Testsb | IgM Antibodiesc,d | IgG Antibodiesc,d |
|---|---|---|---|---|---|---|
| Assuming 1% prevalence: Effect per 1000 patients tested | ||||||
| Sensitivity | 0.99 (0.96–0.99) | 0.98 (0.95–1.00) | 0.81 (0.75–0.86) | 0.52 (0.42–0.62) | 0.33 (0.25–0.41)c 0.73 (0.66–0.78)d |
0.23 (0.16–0.32)c 0.68 (0.62–0.73)d |
| No. of studies (no. of patients) |
6 studies (376 patients) | 4 studies (230 patients) | 4 studies (288 patients) | 5 studies (271 patients) | 12 studies (919 specimens)c 16 studies (2309 specimens)d |
13 studies (1343 specimens)c 16 studies (2708 specimens)d |
| True positives (patients with SARS-CoV2) |
10 (10 to 10) | 10 (10 to 10) | 8 (8 to 8) | 5 (4 to 6) | 3 (3 to 4)c 7 (7 to 8)d |
2 (2 to 3)c 7 (6 to 7)d |
| False negatives (patients incorrectly classified as not having SARS-CoV2) |
0 (0 to 0) | 0 (0 to 0) | 2 (1 to 2) | 5 (4 to 6) | 7 (6 to 7)c 3 (2 to 3)d |
8 (7 to 8)c 3 (3 to 4)d |
| Test accuracy Certainty of evidence |
⊕⊕⊕◯e | ⊕⊕⊕◯e | ⊕⊕⊕◯e | ⊕⊕◯◯f,g | ⊕◯◯◯e,g,h | ⊕◯◯◯e,g,h |
| Specificity | 0.98 (0.94–0.99) | 0.97 (0.89–0.99) | 0.99 (0.96–1.00) | 1.00 (0.99–1.00) | 0.98 (0.97–0.99) | 0.99 (0.99–0.99) |
| No. of studies (no. of patients) |
6 studies (296 patients) | 4 studies (164 patients) | 4 studies (209 patients) | 5 studies (6675 patients) | 21 studies (7165 specimens) | 25 studies (11,887 specimens) |
| True negatives (patients without SARS-CoV2) |
970 (931 to 980) | 960 (881 to 980) | 980 (950 to 990) | 990 (980 to 990) | 970 (960 to 980) | 980 (980 to 980) |
| False positives (patients incorrectly classified as having SARS-CoV2) |
20 (10 to 59) | 30 (10 to 109) | 10 (0 to 40) | 0 (0 to 10) | 20 (10 to 30) | 10 (10 to 10) |
| Test accuracy Certainty of evidence | ⊕⊕⊕◯e | ⊕⊕⊕◯e | ⊕⊕⊕◯e | ⊕⊕⊕◯f | ⊕◯◯◯i,j,k | ⊕◯◯◯i,j,k |
| Considerations | Most patients were symptomatic | Most patients were symptomatic | Most patients were symptomatic | Most patients were asymptomatic; suboptimal reference standard | Case-control studies; suboptimal reference standard | Case-control studies; suboptimal reference standard |
NOTE. These data do not represent comparative differences between tests.
⊕⊕⊕⊕, high certainty; ⊕⊕⊕◯, moderate certainty; ⊕⊕◯◯, low certainty; ◯◯◯◯, very low certainty.
Compared with a composite reference of multiple laboratory-based RT-PCR tests in symptomatic individuals.
Compared with rapid or laboratory-based RT-PCR reference standard in asymptomatic adults.
Compared with rapid or laboratory-based PCR reference in wk 1 after symptom onset.
Compared with rapid or laboratory-based PCR reference in wk 2 after symptom onset.
Rated down for serious indirectness, as the studies included mainly symptomatic individuals.
Rated down for serious risk of bias as the reference was single RT-PCR tests (rapid or laboratory-based).
Rated down for observed serious unexplained inconsistency with considerably variable sensitivity.
Rated down for very serious risk of bias as most of the studies had case-control design, reported results per specimens rather than individual patients, and the reference was single RT-PCR tests (rapid or laboratory-based).
Rated down for very serious risk of bias as most of the studies had case-control design and reported results per specimens rather than individual patients.
Rated down for observed serious unexplained inconsistency with considerably variable specificity.
Rated down for serious indirectness as the many of the studies included stored specimens from time periods before the COVID-19 pandemic.