Feasibility of SARS-CoV-2 virus detection from consumer-grade cotton swabs

To control the growing COVID-19 pandemic, increased testing and containment is essential, yet clinical-grade sampling supplies are expensive and rapidly being depleted. We demonstrate the feasibility of using alternative consumer-grade swabs stored in 95% ethanol rather than viral transport media to detect SARS-CoV-2 from ten hospitalized persons and hospital rooms.


Background 31
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) the causative agent of coronavirus disease 32 19  has spread to 185 countries resulting in 278,892 deaths and 4,006,257 total confirmed cases 33 as of May 11, 2020 [1-2]. Large-scale testing remains key for controlling viral spread, but sample collection 34 supplies including swabs, viral transport media (VTM) and personal protective equipment (PPE) are being 35 depleted in developed nations like the United States, and are in even shorter supply in low-and middle-36 income countries [3]. Validation of alternatives strategies such as self-administered testing using consumer-37 grade materials is urgently needed. for Disease Control (CDC) guidance for testing [4]. Similarly, the CDC-recommended VTM requires 43 specialized ingredients and contains antimicrobials likely to interfere with downstream assessment of the 44 microbial context of SARS-CoV-2 that may enable new insights into viral susceptibility and resistance [5]. 45 VTM also maintains viral viability and therefore requires processing in a facility with more stringent 46 biosafety practices compared to inactivating collection methods. Using inactivating sample collection 47 solutions could increase the number of testing laboratories and ameliorate the risks associated with sample 48 transport and processing. Here we demonstrate the feasibility of detecting SARS-CoV-2 RNA from patient 49 and built-environment samples using viral-inactivating storage solutions and alternative medical-grade and 50 consumer-grade swabs. 51 52

Methods 53
Swab feasibility testing 54 Six swab types were compared and processed following the standard SARS-CoV-2 protocol provided by 55 the CDC [6] (Supplementary Methods). These swab types are: sterile polyester-head, plastic-shaft ('PE'), 56 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(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Results: 69
We compared the efficiency of extracting and processing RNA from 200 µL of VTM eluent surrounding 70 polyester-tipped plastic-shafted nasopharyngeal (NP) swabs (CDC protocol) to more accessible nares 71 samples collected using PE swabs stored in viral-inactivating alcohol. RNA extraction efficiency was 72 significantly reduced from samples stored in 95% EtOH when extracted from the eluent (n=22), but similar 73 when extracted from the swab head itself (n=18) compared to the CDC protocol (n=39) (Figure 1a). 74 Extracting from the swab head also provided significantly higher SARS-CoV-2 viral load compared to the 75 EtOH eluent of the same patient samples (n=7) (Figure 1b, p=0.032). We separately compared the use of 76 95% EtOH vs 91% isopropanol as the storage media with human RNA and found no impact on extraction 77 efficiency (32.1% and 35.3% recovery respectively) (ANOVA, P>0.05) (Figure 1c). However, in the 78 presence of abundant RNase, 95% EtOH protected both human RNA and SARS-CoV-2 RNA better than 79 91% isopropanol (Figure 1d). 80 81 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 May 18, 2020. . https://doi.org/10.1101/2020 We next tested RNA recovery from a range of medical-and consumer-grade swabs (Supplementary 82 Methods). The yield was highest from swab heads compared to eluent regardless of the swab type and 83 whether stored in 95% EtOH (P<0.0001, U=37, Mann-Whitney) or 91% isopropanol (P<0.0001, U=28, 84 Mann-Whitney) (Figure 1e 2). Swab type did not impact the ability to detect a linear decrease of positive control human and SARS-91 CoV-2 RNA when extracting directly from either CDC-recommended PE swabs or CGp swabs 92 (Supplemental Figure 3). 93

94
As a clinical proof-of-concept, we collected samples from ten participants admitted to the hospital for 95 COVID-19 using TMI and/or CGp swabs alongside the recommended PE swabs, and performed RT-qPCR 96 per CDC guidelines. All three swab types successfully detected positive control SARS-CoV-2 RNA except 97 for one false negative CGp, and no false positives (Supplemental Figure 4). For this comparison, samples 98 inconclusive solely for the SARS-CoV-2 N2 amplicon (I-N2 in Figure 1g) were considered positive for the 99 presence of SARS-CoV-2 RNA based on observed and reported concerns with this primer set [7]. PE swabs 100 were 100% concordant with NP results when extracting from the swab head compared to 67% concordance 101 for eluent from the same samples. TMI swabs were 85% and 57% concordant while CGp were 70% and 102 56% concordant for swab head and eluent respectively ( Figure 1g). To evaluate contamination of SARS-103 CoV-2 RNA on environmental surfaces, we collected and compared swabs from the inside floor and bedrail 104 of the same participants' rooms using the same swab types [8]. Only 3/10 bedrails had detectable SARS-105 CoV-2 with any swab type, while floor samples tested positive for SARS-CoV-2 RNA for at least one swab 106 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.

Discussion: 110
We provide evidence that nasal samples collected using more widely-available consumer-grade cotton-111 tipped swabs can be stored in viral-inactivating alcohol without compromising the ability to detect SARS-112 CoV-2. The sensitivity for detection of SARS-CoV-2 RNA was comparable between the hospital NP swabs 113 (CDC protocol) and both TMI and CGp nasal swabs when extracting from the swab head. Negative NP 114 results for previously positive participants may be due to viral clearance, the timing of sampling during the 115 course of infection or inconsistencies among the standard NP swabs. Of note, wooden-shafted swabs 116 performed poorly only when extracting from the eluent, suggesting that RNA adsorption onto the shaft 117 rather than RT-qPCR inhibitors may be the source of interference with current eluent-based testing 118

methods. 119
Cotton-tipped swabs and alcohol-based solutions are compatible with standard microbiome and 120 metabolome analyses prohibited by VTM, and could enable more widespread assessment for SARS-CoV-121 2 RNA in human and environmental samples. SARS-CoV-2 was only detected on 30% of participants' 122 bedrails, which may have been due to routine cleaning measures and/or minimal interaction with the 123 surfaces from heavily-sedated or intubated patients. This may be present a challenge for monitoring shared 124 surfaces between the healthcare worker and patients. In contrast, the detection of SARS-CoV-2 RNA from 125 the floor samples demonstrates a potentially important reservoir for viral exposure, as shoe covers are note 126 currently recommended by the CDC. However, additional testing is needed to determine whether viable 127 virus remains on these surfaces. 128 In summary, our results suggest detection of SARS-CoV-2 RNA could be performed using less expensive, 129 consumer-grade materials. We add to the emerging body of literature supporting nasal sampling as opposed 130 to NP sampling [10][11][12][13][14]. It is conceivable that patients could collect samples at home, thus reducing risk 131 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.  (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.

Figure Legends
(multiple t-test) using either the human Rp or SARS-CoV-2 N1 primer set. e) Comparison human RNA 206 recovery across six swab types (PE=polyester 'commercial',BDF=BD foam 'commercial',TMI=BD 207 TMI 'commercial',CGp=plastic 'consumer grade',Pu=Puritan 'commercial',CGw=wood 'consumer 208 grade'), extracted from 200µL eluent (blank bar) or the swab head. Recovery for each swab type is 209 normalized to the CDC recommended method (eluent from PE swab). A '2' would indicate there was 2x 210 more RNA recovered whereas a 0.5 would indicate a 50% reduction in RNA recovery. f) Total RNA 211 copies per extraction for all samples which are grouped by sample-type (eluent or swab head) and storage 212 buffer (95% EtOH or 91% isopropanol). Pairwise comparisons performed within sample-type (not 213 significant) and across sample-type controlling for storage buffer (Mann-Whitney, U=test statistic). g) 214 Demonstration of consumer-grade CGp and bulk TMI swab congruence compared to clinical-grade 215 hospital tests using polyester-tipped plastic shafted NP swabs across ten patient rooms. Samples positive 216 (+, dark blue background) or negative (-, orange background) for N1, N2, and Rp by CDC guidelines. 217 Samples positive for N1 and Rp, but negative for N2 are labeled inconclusive for N2 (I-N2, blue) but 218 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. All study patients were diagnosed and hospitalized with COVID-19 with approval of the UC San Diego 243 Institutional Review Board under protocols #150275 and #200613. Both nasal samples and hospital 244 surfaces were collected using three unmoistened swab types (PE, TMI, CGp; see Table 1). Nasal samples 245 were collected by inserting the swab into one nostril to the depth of approximately 2-3 cm and rotated for 246 5-10 seconds. Hospital surfaces sampled included the floor inside the patient's room (approximately 1x1 247 square foot area) and the patient's bedrail. All swabs were immediately placed in a collection tube 248 containing 0.5-1 mL 95% ethanol and stored on dry ice and processed for RNA or total nucleic acid 249 extraction (Supplementary Methods). 250 251 RT-qPCR for VTM and 95% EtOH comparison using polyester-tipped plastic swabs 252 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.