Evaluation of saliva self-collection devices for SARS-CoV-2 diagnostics

Abstract

Background: There is an urgent need to expand testing for SARS-CoV-2 and other respiratory pathogens as the global community struggles to control the COVID-19 pandemic. Current diagnostic methods can be affected by supply chain bottlenecks and require the assistance of medical professionals, impeding the implementation of large-scale testing. Self-collection of saliva may solve these problems, as it can be completed without specialized training and uses generic materials.

Methods: We observed 30 individuals who self-collected saliva using four different collection devices and analyzed their feedback. Two of these devices, a funnel and bulb pipette, were used to evaluate at-home saliva collection by 60 individuals. SARS-CoV-2-spiked saliva samples were subjected to temperature cycles designed to simulate the conditions the samples might be exposed to during the summer and winter seasons and sensitivity of detection was evaluated.

Results: All devices enabled the safe, unsupervised self-collection of saliva. The quantity and quality of the samples received were acceptable for SARS-CoV-2 diagnostic testing, as determined by human RNase P detection. There was no significant difference in SARS-CoV-2 nucleocapsid gene (N1) detection between the freshly spiked samples and those incubated with the summer and winter profiles.

Conclusion: We demonstrate inexpensive, generic, buffer free collection devices suitable for unsupervised and home saliva self-collection.

Fig. 1

Collection devices are inexpensive, easy to use, and yield testable samples. Survey responses were reported from strongly disagree to strongly agree. a The four collection devices tested are inexpensive and provide users with a range of features to choose from. Prices at time of publication are shown in US dollars. b Participants reported being self-sufficient and confident in their ability to correctly collect saliva samples (from Additional file 2: Fig. S2). The questions are displayed above the corresponding graphs. The percentage response value for each device is shown above each bar. Two sets of participant responses were excluded because one participant did not provide a response for all four devices and one did not understand the response scale. P pipette tip, C collection aid, F funnel, B bulb pipette

Fig. 2

At-home saliva collection kit components suitable for sample collection. a Each of the participants were sent an at-home collection kit comprised of either a funnel (i) or bulb pipette (ii) with a labeled screw-cap tube (iii), patient identifier sticker (iv), biohazard collection bag with absorbent sheet (v), FedEx UN 3373 Pak (vi), an alcohol pad (vii), and box for return shipment (viii). b Participant confidence in at-home self-collection of saliva when using either a funnel or bulb pipette (from Additional file 4: Fig. S4). Survey responses were reported on a scale of 1 (strongly disagree) to 5 (strongly agree). Overall, there was no significant difference between the collection devices in relation to the participant’s confidence and ability to use either device. The questions are displayed above the corresponding graphs. F funnel, B bulb pipette

Fig. 3

The quality of the samples was adequate for testing with a PCR-based assay. Laboratory survey questions pertaining to the quality of the samples are shown on the x-axis (from Additional files 3 and 5). Data points represent the mean response, green dots represent samples collected from the pilot study and blue dots represent samples collected from the at-home collection kit. Survey responses were reported on a scale of 1 (strongly disagree) to 5 (strongly agree). Samples with less favorable responses are highlighted in red. Mean and standard deviation (st. dev.) are shown in black. The graph on the right shows the cycle threshold (Ct) values for the internal control RNAse P (RP) from each of the saliva samples submitted. The dots represent Ct value per participant. Ct values over 35 are considered invalid and are highlighted in gray. P-value is shown using one-way Mann–Whitney. Mean and standard deviation (st. dev.) are shown in black

Fig. 4

SARS-CoV-2 detection in saliva remains stable following summer and winter shipping conditions. Spiked saliva samples diluted twofold were incubated under summer or winter profiles in triplicate (from Additional file 8: Table S3). Summer profile consisted of 40 °C for 8 h, room temperature for 4 h, 40 °C for 2 h, 28 °C for 36 h, and 40 °C for 6 h. Winter profile consisted of − 20 °C for 8 h, room temperature for 4 h, − 20 °C for 2 h, 4 °C for 36 h, and − 20 °C for 6 h. Samples with invalid Ct values are highlighted in red. a No significant differences were found in N1 Ct values when comparing the profiles to the fresh (control) samples (Kruskal–Wallis; P > 0.05). b The Ct values for RP of samples for the winter profile were significantly higher than fresh samples (Kruskal–Wallis; P < 0.03). The horizontal bars indicate the median Ct value