In this prospective, noninterventional study, fresh samples from asymptomatic or symptomatic patients with suspected SARS-CoV-2 infection were collected in a POC setting across two sites (Mayo Clinic, Jacksonville, Florida, USA, and Charité University Hospital, Berlin, Germany). All SARS-CoV-2 tests were ordered by an HCP aligned with real-world standard-of-care (SOC) procedures. Suspicion of SARS-CoV-2 infection was defined as an individual with signs and/or symptoms consistent with respiratory infection (e.g., fever, cough, shortness of breath, or myalgia) or recent known exposure (i.e., close contact with an individual with confirmed COVID-19 or who had respiratory infection symptoms). Frozen, archived samples from individuals with suspected infection who had previously tested positive for SARS-CoV-2 were also used in the analysis.

The study was conducted in compliance with the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) and Good Clinical Practice Guidelines, and the Helsinki Declaration of 1964 and its later amendments. Institutional review board approval was obtained for each participating study site before study start (Institutional Ethical Review Board of the Charité-Universitätsmedizin Berlin [reference number: EA2/119/23] and Mayo Clinic Florida Institutional Review Board (reference number: 23-000144). All participants aged 18 years or older provided written informed consent; all participants aged less than 18 years had written informed consent provided on their behalf by a legal guardian.

During the study, the two participating sites used either the POC SARS-CoV-2 or 68/8800 tests as their standard method for detecting SARS-CoV-2. Clinical samples were prospectively collected at each site using their respective SOC procedures. Sample testing with POC SARS-CoV-2 was performed in accordance with the manufacturer’s assay instructions for use. All system operators were provided with the relevant user materials prior to the study and were blinded to all results from other test systems.

Eligible participants provided fresh nasopharyngeal samples. These were collected using each site’s established SOC procedures for upper respiratory swab sampling. Flocked swab samples were eluted in BD/Copan Universal Transport Medium (UTM), Remel Viral Transport Medium (VTM), or a similar 3 mL viral transport medium compatible with molecular testing for SARS-CoV-2. Testing of fresh, non-archived samples (3 mL sample; 200 µL for testing) was conducted on the Cobas Liat System as soon as possible, and up to a maximum of 4 h, after sample collection. Subsequent 68/8800 testing was conducted at a central laboratory within 72 h of collection (stored at 2–8 °C).

The study protocol allowed for the use of archived specimens if prospectively collected samples were insufficient to reach the required number of positive samples. De-identified, frozen specimens were eligible for inclusion if they had previously tested positive (using any method) for SARS-CoV-2 and were collected in BD/Copan UTM, Remel VTM, or a similar viral transport media compatible with SARS-CoV-2 molecular testing (stored at −70 °C or colder). A full list of specimen exclusion criteria is available in the Supplementary Material (Table S1). Thawed samples were first retested for SARS-CoV-2 positivity if the previous test method for establishing SARS-CoV-2 positivity was not 68/8800. Following the 68/8800 test to establish SARS-CoV-2 positivity, POC SARS-CoV-2 was conducted within 4 h for thawed samples stored at room temperature or within 4–24 h for thawed samples stored at 2–8 °C.

For both fresh and frozen specimens, repeat tests with either assay were conducted if an invalid result was generated in the first test; however, a second invalid result led to exclusion from the analysis.

POC SARS-CoV-2 has been authorized for use under the Food and Drug Administration (FDA) Emergency Use Authorization (EUA), US 510(k) clearance and Conformité Européenne in vitro diagnostic (CE-IVD) in laboratory and POC settings for in vitro detection of SARS-CoV-2 in people with or without symptoms of COVID-19 [17, 18]. In the US, it is also authorized for use at POC in settings operating under a Clinical Laboratory Improvement Amendments (CLIA) certificate of waiver, compliance, or accreditation [17]. Under FDA EUA and CE-IVD, the 68/8800 laboratory test is approved for in vitro detection of SARS-CoV-2 in individuals suspected of COVID-19 and those without symptoms [19,20,21].

The real-world POC setting of this study gave rise to site-specific variation in some procedures not otherwise specified in the protocol; these were broadly related to sample handling and laboratory testing procedures that were instead conducted per site SOC. At the German site, all samples following POC SARS-CoV-2 testing were diluted for 68/8800 testing. Prior to dilution, a 600 µL aliquot was prepared and frozen (− 20 °C). Diluted samples consisted of ~ 2.2 mL of the remaining original sample and 4.3 mL of added Cobas PCR Media. These diluted samples (stored at 2–8 °C) were retested using POC SARS-CoV-2 before 68/8800 laboratory testing was completed. This was not the case at the US site where source (non-diluted) samples were used for both POC SARS-CoV-2 and 68/8800 testing as per manufacturer’s assay instructions for use.

At the US site, fresh specimens were initially processed using the Cobas SARS-CoV-2 & Influenza A/B test on the Cobas Liat System (Roche Diagnostics, USA); different than that stipulated by the study protocol. The same specimens were also processed on the Cobas 6800 System, which enabled the qualitative detection of SARS-CoV-2. Positive SARS-CoV-2 samples using this assay were archived according to SOC procedures. Following identification of the assay error, all positive samples were rerun following a single freeze–thaw cycle with POC SARS-CoV-2 as described above for frozen specimens (conducted at a different central laboratory site: Mayo Clinic, Rochester, MN, USA).

The diagnostic performance of POC SARS-CoV-2 was evaluated relative to 68/8800. Subjects/specimens with both valid POC SARS-CoV-2 test results and valid 68/8800 test results were considered evaluable. Pairwise agreement between the two assays was evaluated by estimating the positive percent agreement (PPA), negative percent agreement (NPA), and overall percent agreement (OPA) with two-sided 95% confidence intervals (CI). PPA and NPA were calculated as the percentage of specimens with positive or negative results, respectively, on 68/8800 that also had positive or negative results on POC SARS-CoV-2. Similarly, OPA was calculated as the percentage of total specimens with the same SARS-CoV-2 outcome in both assays. The distribution of cycle threshold (Ct) values for POC SAR-CoV-2 and 68/8800 were recorded for the exploratory analysis of discordant samples.

In terms of PPA, the acceptance criteria for POC SARS-CoV-2 were set at PPA ≥ 95% with a lower-bound 95% CI ≥ 88.5%. For NPA, the acceptance criteria were NPA ≥ 95% with a lower-bound 95% CI ≥ 90.0%. To achieve the defined PPA lower 95% CI boundary, an estimated total sample size of approximately 100 positive samples, out of a maximum of 1000 prospectively collected samples, was required for the analysis. A minimum of 60 positive samples (including fresh and frozen) was required for the analysis; there was no requirement for the minimum number of negative samples.

For the German site, diagnostic performance was also explored for the retested, diluted POC SARS-CoV-2 samples relative to 68/8800. In addition, pairwise agreement between the diluted POC SARS-CoV-2 samples and the corresponding source (non-diluted) samples was explored; statistical significance at p < 0.05 was determined using a McNemar test.

Demographic characteristics, self-reported symptoms and vital signs were collected for each patient following enrolment or taken from electronic medical records; these data were summarized using descriptive statistics. Clinical raw study data were captured in Microsoft Excel, then transferred and analyzed using SAS/STAT® software.