key: cord-0770558-rxxucjew
authors: Stokes, W.; Berenger, B.; Portnoy, D.; Scott, B.; Singh, T.; Venner, A.; Turnbull, L.; Pabbaraju, K.; Shokoples, S.; Wong, A. A.-T.; Gill, K.; Guttridge, T.; Proctor, D.; Hu, J.; Tipples, G.
title: Real-World Clinical Performance of the Abbott Panbio with Nasopharyngeal, Throat and Saliva Swabs Among Symptomatic Individuals with COVID-19
date: 2021-01-04
journal: nan
DOI: 10.1101/2021.01.02.21249138
sha: fc62bc44cc3b565d039df68522a2c15d7673b41b
doc_id: 770558
cord_uid: rxxucjew

BACKGROUND Point of Care Testing (POCT) SARS-CoV-2 antigen tests, such as the Abbott Panbio, have great potential to help combat the COVID-19 pandemic. The Panbio is United States Food and Drug Administration (FDA) approved for the detection of SARS-CoV-2 in symptomatic individuals within the first 7 days of COVID-19 symptom onset(s). METHODS Symptomatic adults recently diagnosed with COVID-19 in the community were recruited into the study. Paired nasopharyngeal (NP), throat, and saliva swabs were collected, with one paired swab tested immediately with the Panbio, and the other transported in universal transport media and tested using reverse-transcriptase polymerase chain reaction (RT-PCR). Positive percent agreement (PPA) was calculated. Subsequently, individuals within 7 days of symptom onset who presented to community assessment centres for SARS-CoV-2 testing had Panbio testing completed and paired with RT-PCR results from parallel NP or throat swabs. RESULTS 145 individuals were included in the study. Collection of throat and saliva was stopped early due to poor performance (throat PPA 57.7%, n=61, and saliva PPA 2.6%, n=41). NP swab PPA was 87.7% [n=145, 95% confidence interval 81.0% - 92.7%]. There were 1,641 symptomatic individuals tested by Panbio in community assessment centres, with 268/1641 (16.3%) positive for SARS-CoV-2. There were 37 false negatives, corresponding to a PPA of 86.2% [81.5% - 90.1%]. CONCLUSIONS The Panbio test reliably detects most cases of SARS-CoV-2 from adults in the POCT community setting presenting within 7 days of symptom onset using nasopharyngeal swabs. Throat and saliva swabs are not reliable specimens for the Panbio.

The Panbio (Abbott, Ill, United States) is approved by the United States Food and Drug 73 Administration (FDA) for the detection of SARS-CoV-2 antigen in individuals who are within 74 the first 7 days of symptom(s) onset. The immunochromatographic assay detects SARS-CoV-2 75 nucleocapsid protein and is indicated only for nasopharyngeal (NP) swabs. The test should be 76 conducted either immediately after collection, or up to 2 hours if the NP swab is placed in the 77 Panbio extraction tube filled with extraction buffer at room temperature. 1 78 79 Based on a study by Abbott that was conducted on 585 NP specimens collected from individuals 80 exposed to SARS-CoV-2 or having COVID-19 symptoms within 7 days, sensitivity of the 81

Panbio was found to be 91.4% and specificity 99.8% when compared to a reverse transcriptase 82 real-time polymerase chain reaction (RT-PCR) reference method. Sensitivity increased to over 83 94% when samples with cycle threshold (Ct) >33 were excluded. 1 There is currently a paucity of 84 data available from external third parties on the Panbio's performance. The few published studies 85 have demonstrated Panbio sensitivity ranges from 72.6% -86.5% among symptomatic 86 individuals or exposed asymptomatic individuals. 2-4 87 88 We sought to assess the positive percent agreement (PPA) of the Panbio by comparing its 89 performance to RT-PCR testing among individuals in the community using two separate 90 evaluations. The first by testing individuals with recently confirmed COVID-19 while adhering 91 as closely as possible to manufacturer recommendations (testing of symptomatic individuals 92 within 7 days of symptom(s) onset). The second setting was a prospective evaluation using the 93 Panbio to diagnose COVID-19 in SARS-CoV-2 assessment/screening centres. We also tested the 94 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted January 4, 2021. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Cobas SARS-CoV-2 test, as per the manufacturer, a positive result was defined as 2/2 targets 140 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted January 4, 2021. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted January 4, 2021. One hundred and sixty-three individuals were recruited for the first clinical evaluation. Eighteen 184 individuals were excluded: Three were asymptomatic at the time of COVID-19 diagnosis and at 185 time of study recruitment, nine were symptomatic at the time of COVID-19 diagnosis but 186 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted January 4, 2021. ; https://doi.org/10.1101/2021.01.02.21249138 doi: medRxiv preprint asymptomatic at time of study recruitment, four had Panbio results that were not recorded, one 187 had the Panbio reported as negative before 15 minutes, and another was unable to be processed 188 by RT-PCR. Individual characteristics of the remaining 145 individuals is provided in Table 1 . were 70.0% of throat swabs tested on the Panbio that were collected before the reference throat 202 swab. In addition, 14.6% of swabs used for saliva testing on the Panbio were collected before 203 both throat swabs, and the rest of the saliva swabs were tested after the two throat swabs. The 204 PPA of throat and saliva swabs (95% confidence intervals (CI)) was 57.7% (95% CI 43.2% -205 71.3%) and 2.6% (95% CI 0.06% -13.5%), respectively (see supplementary material). 206

Of the 145 that underwent a NP swab, 121 were positive on both the Panbio and RT-PCR (Table  208 2). The PPA of the Panbio compared to RT-PCR was 87.7% (95% CI 81.0% -92.7%) ( Table 3) . 209 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. were negative on the Panbio and RT-PCR. All 11 retrospective samples containing other 232 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. used. All studies examined paired NP swabs tested with the Panbio and a PCR-based platform, 256 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted January 4, 2021. ; https://doi.org/10.1101/2021.01.02.21249138 doi: medRxiv preprint with two of the three studies using the Allplex (Seegene, South Korea). 3,4 One study 2 used the 257 VitaPCR SARS-CoV-2 (Credo diagnostics, Singapore) that has limited data available on its 258 performance, which may potentially explain why 7 samples were Panbio positive but VitaPCR 259 SARS-CoV-2 negative. 9 Gremmels et al. performed testing on the Panbio up to 2 hours from 260 collection, which may account for the lower sensitivity detected (72.6%). 3 Of the studies that 261 examined positivity rate based on symptom duration, one study found no difference in positivity 262 rate with duration of symptom onset 3 , whereas another found higher sensitivity in individuals 263 with symptom onset < 7 days (sensitivity 86.5%) compared to individuals with symptom onset ≥ 264 7 days (sensitivity 53.8%). 4 We found no difference in Panbio sensitivity among individuals with 265 symptoms > 7 days, but this was limited to a very small pool of samples (N=15). All studies, 266 including ours, found decreases in Panbio sensitivity among SARS-CoV-2 samples with higher 267

Ct values, with sensitivity dropping when Ct values are approximately > 26. 268

Our study contributes to the literature on the Panbio's performance by using alternative 270 collection methods, such as throat and saliva swabs. Unfortunately, these specimens were proven 271 to be inferior to NP swabs and should not be used on the Panbio. Further studies are required to 272 determine if an alternative way to test saliva on the Panbio could prove effective (e.g. direct 273

inoculation of saliva onto the Panbio test cartridge or saliva collected in a media). We did not 274 evaluate nasal swabs on the Panbio because the results could have been altered by collecting NP 275 swabs also. However, previous work done by our laboratory has shown nasal swabs to be 276 inferior to throat swabs for the detection of SARS-CoV-2, so it would be surprising if nasal 277 swabs proved to be as effective a specimen as NP swabs for testing on the Panbio. 6 278 279 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted January 4, 2021. ; https://doi.org/10.1101/2021.01.02.21249138 doi: medRxiv preprint

Our study was predominately restricted to individuals within the community who had symptoms 280 ≤ 7 days. As such, our study was unable to provide any conclusions about the Panbio 281 performance among individuals admitted to hospital, in congregate living facilities, who are 282 asymptomatic, and individuals with symptoms > 7 days. Our study did not evaluate the negative 283 percent agreement of the Panbio when applied in real-world settings. The strengths of our study 284 include the large number of COVID-19 positive individuals recruited. In addition, we included 285 prospective data taken from real-world settings (symptomatic individuals presenting to COVID 286 assessment centres) and found similar results, which further reinforces our study findings. We settings. Given the speed, low-complexity and acceptable performance, the Panbio test is suitable 300 for use in the POCT setting, especially when rapid identification of positive patients is critical. 301

As such, they will play an impactful role in combating the COVID-19 pandemic by improving 302 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 314  315  316  317  318  319  320  321  322  323  324  325  326  327  328  329  330  331  332  333  334  335  336  337 All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted January 4, 2021. ; https://doi.org/10.1101/2021.01.02.21249138 doi: medRxiv preprint All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted January 4, 2021. ; https://doi.org/10.1101/2021.01.02.21249138 doi: medRxiv preprint

COVID-19 Ag Rapid Test Device