SafeSwab, a sample collection and dispensing device for near-patient testing (preprint)

Background The COVID-19 pandemic has accelerated the pace of innovation around virtual care visits and testing technology. Here we present the SafeSwab (Safe Health Systems, Los Angeles, CA), an integrated, universal sample collection and dispensing device that is designed to minimize user error and enable rapid testing in a point of care or self-testing format. Methods The SafeSwab was used with the Safe Health Systems HealthCheck digital health application to enable self-testing by patients using lateral flow tests for SARS-CoV-2 antigen or for antibodies against SARS-CoV-2. Results Patients (n=74) using the SafeSwab produced a valid rapid test result in 96% of attempts, and 96% of patients felt confident that they had collected a good sample. The Safe HealthCheck app has an integrated image analysis algorithm, AutoAdapt LFA, that interprets a picture of a rapid test result, and the algorithm interpreted the result correctly 100% of the time. Conclusion The SafeSwab was found to be versatile and easy to use for both self-collected nasal sampling as well as fingerstick blood sampling. The use of Safe Health Systems HealthCheck app allows an integrated solution for patient instruction and test interpretation

Adaptable Automated Interpretation of Rapid Diagnostic Tests Using Few-Shot Learning (preprint)

Point-of-care lateral-flow assays (LFAs) are becomingly increasingly prevalent for diagnosing individual patient disease status and surveying population disease prevalence in a timely, scalable, and cost-effective manner, but a central challenge is to assure correct assay operation and results interpretation as the assays are manually performed in decentralized settings. A smartphone-based software can automate interpretation of an LFA kit, but such algorithms typically require a very large number of images of assays tested with validated specimens, which is challenging to collect for different assay kits, especially for those released during a pandemic. Here, we present an approach – AutoAdapt LFA – that uses few-shot learning, an approach used in other applications such as computer vision and robotics, for accurate and automated interpretation of LFA kits that requires a small number of validated images for training. The approach consists of three components: extraction of membrane and zone areas from an image of the LFA kit, a self-supervised encoder that employs a feature extractor trained with edge-filtered patterns, and few-shot adaptation that enables generalization to new kits using limited validated images. From a base model pre-trained on a commercial LFA kit, we demonstrated the ability of adapted models to interpret results from five new COVID-19 LFA kits (three detecting antigens for diagnosing active infection, and two detecting antibodies for diagnosing past infection). Specifically, using just 10 to 20 images of each new kit, we achieved accuracies of 99% to 100% for each kit. The server-hosted algorithm has an execution time of approximately 4 seconds, which can potentially enable quality assurance and linkage to care for users operating new LFAs in decentralized settings.

Rapid Detection of SARS-CoV-2 Antigen from Serum in a Hospitalized Population (preprint)

SARS-CoV-2 viremia has been demonstrated in some patients using molecular assays. Here we demonstrate detection of SARS-CoV-2 antigen in a cohort of hospitalized patients using a rapid diagnostic test from Anhui Deepblue Medical Technology Co., Ltd. We detected antigen in serum from 11 of 13 patients at time points ranging from three to eighteen days from symptom onset and observed that the disappearance of an antigen signal was associated with seroconversion. These results demonstrate proof of principle use of a rapid antigen test with serum samples in a format compatible with point of care testing.

Development of a Rapid Point-Of-Care Test that Measures Neutralizing Antibodies to SARS-CoV-2 (preprint)

As increasing numbers of people recover from and are vaccinated against COVID-19, tests are needed to measure levels of protective, neutralizing antibodies longitudinally to help determine duration of immunity. We developed a lateral flow assay (LFA) that measures levels of neutralizing antibodies in plasma, serum or whole blood. The LFA is based on the principle that neutralizing antibodies inhibit binding of the spike protein receptor-binding domain (RBD) to angiotensin-converting enzyme 2 (ACE2). The test classifies high levels of neutralizing antibodies in sera that were titered using authentic SARS-CoV-2 and pseudotype neutralization assays with an accuracy of 98%. Sera obtained from patients with seasonal coronavirus did not prevent RBD from binding to ACE2. As a demonstration for convalescent plasma therapy, we measured conversion of non-immune plasma into strongly neutralizing plasma. This is the first report of a neutralizing antibody test that is rapid, highly portable and relatively inexpensive that might be useful in assessing COVID-19 vaccine immunity.

Retrospective Clinical Evaluation of Four Lateral Flow Assays for the Detection of SARS-CoV-2 Antibodies (preprint)

Coronavirus disease 2019 (COVID-19) is a potentially life-threatening respiratory infection caused by severe acute respiratory coronavirus 2 (SARS-CoV-2), for which numerous serologic assays are available. In a CLIA laboratory setting, we used a retrospective sample set (n = 457) to evaluate two lateral flow immunoassays (LFIAs; two iterations of Rapid Response COVID-19 Test Cassette, BTNX Inc.) and a subset of to evaluate SARS-COV-2 IgG/IgM Rapid Test, ACON Laboratories (n = 200); and Standard Q COVID-19 IgM/IgG Duo, SD BIOSENSOR (n = 155) for their capacity to detect of SARS-CoV-2 IgG. In a cohort of primarily hospitalized patients with RT-PCR confirmed COVID-19, the BTNX assays demonstrated 95% and 92% agreement with the Abbott SARS-CoV-2 IgG assay and sensitivity was highest at [≥] 14 days from symptom onset [BTNX kit 1, 95%; BTNX kit 2, 91%]. ACON and SD assays demonstrated 99% and 100% agreement with the Abbott assay at [≥] 14 days from symptom onset. Specificity was measured using 74 specimens collected prior to SARS-CoV-2 circulation in the United States and 31 "cross-reactivity challenge" specimens, including those from patients with a history of seasonal coronavirus infection and was 98% for BTNX kit 1 and ACON and 100% for BTNX kit 2 and SD. Taken with data from EUA assays, these results suggest that LFIAs may provide adequate results for rapid detection of SARS-CoV-2. Replicating these results in fingerstick blood in outpatient populations, would further support the possibility that LFIAs may be useful to increase access to serologic testing