Article Text

Original research
Evaluation of a novel university-based testing platform to increase access to SARS-CoV-2 testing during the COVID-19 pandemic in a cohort study
  1. Julia Catherine Bennett1,
  2. Jessica O’Hanlon1,
  3. Zachary Acker2,
  4. Peter D Han2,3,
  5. Devon McDonald1,
  6. Tessa Wright1,
  7. Kyle G Luiten1,
  8. Lani Regelbrugge2,
  9. Kathryn M McCaffrey2,
  10. Brian Pfau2,
  11. Caitlin R Wolf1,
  12. Geoffrey S Gottlieb1,4,5,
  13. James P Hughes6,
  14. Marco Carone6,
  15. Lea M Starita2,3,
  16. Helen Y Chu1,
  17. Ana A Weil1
  1. 1 Department of Medicine, University of Washington, Seattle, Washington, USA
  2. 2 Brotman Baty Institute, Seattle, Washington, USA
  3. 3 Department of Genome Sciences, University of Washington, Seattle, Washington, USA
  4. 4 Department of Global Health, University of Washington, Seattle, Washington, USA
  5. 5 Environmental Health & Safety Department, University of Washington, Seattle, WA, USA
  6. 6 Department of Biostatistics, University of Washington, Seattle, Washington, USA
  1. Correspondence to Julia Catherine Bennett; jubenn94{at}


Objective We aimed to evaluate the feasibility and utility of an unsupervised testing mechanism, in which participants pick up a swab kit, self-test (unsupervised) and return the kit to an on-campus drop box, as compared with supervised self-testing at staffed locations.

Design University SARS-CoV-2 testing cohort.

Setting Husky Coronavirus Testing provided voluntary SARS-CoV-2 testing at a university in Seattle, USA.

Outcome measures We computed descriptive statistics to describe the characteristics of the study sample. Adjusted logistic regression implemented via generalised estimating equations was used to estimate the odds of a self-swab being conducted through unsupervised versus supervised testing mechanisms by participant characteristics, including year of study enrolment, pre-Omicron versus post-Omicron time period, age, sex, race, ethnicity, affiliation and symptom status.

Results From September 2021 to July 2022, we received 92 499 supervised and 26 800 unsupervised self-swabs. Among swabs received by the laboratory, the overall error rate for supervised versus unsupervised swabs was 0.3% vs 4%, although this declined to 2% for unsupervised swabs by the spring of the academic year. Results were returned for 92 407 supervised (5% positive) and 25 836 unsupervised (4%) swabs from 26 359 participants. The majority were students (79%), 61% were female and most identified as white (49%) or Asian (34%). The use of unsupervised testing increased during the Omicron wave when testing demand was high and stayed constant in spring 2022 even when testing demand fell. We estimated the odds of using unsupervised versus supervised testing to be significantly greater among those <25 years of age (p<0.001), for Hispanic versus non-Hispanic individuals (OR 1.2, 95% CI 1.0 to 1.3, p=0.01) and lower among individuals symptomatic versus asymptomatic or presymptomatic (0.9, 95% CI 0.8 to 0.9, p<0.001).

Conclusions Unsupervised swab collection permitted increased testing when demand was high, allowed for access to a broader proportion of the university community and was not associated with a substantial increase in testing errors.

  • COVID-19
  • Feasibility Studies
  • Observational Study

Data availability statement

Data are available on reasonable request.

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • The Husky Coronavirus Testing Study was a large, university-wide, voluntary SARS-CoV-2 testing cohort that captured demographic, health and SARS-CoV-2 testing data on over 25 000 university affiliates over time.

  • These data allowed us to evaluate the feasibility and utility of SARS-CoV-2 testing mechanisms in a routine use setting.

  • We were unable to assess the exact reasons for individuals choosing to use unsupervised versus supervised testing because we did not collect data on participant’s preference for each testing mechanism.


University campuses have been sites of multiple COVID-19 outbreaks1–7 and have led to COVID-19 transmission into surrounding communities.8 9 University COVID-19 testing programmes are one of several strategies for reducing and tracking campus SARS-CoV-2 transmission.10–12 Testing programmes that use molecular tests, rather than rapid antigen tests, are primarily conducted through provider obtained or supervised self-swabs, which require substantial human resources.1 3 6 11 13–15 Less resource-intensive strategies include pooled testing,15–17 unsupervised self-swabs returned to laboratories by mail,2 18 unsupervised self-swabs or saliva collection returned to ‘drop box’ locations on university campuses1 17 or use of rapid antigen tests.3 Although many communities have transitioned to primarily using rapid antigen tests, molecular testing is more sensitive, is recommended in certain circumstances and may be the first test available for new or emerging pathogens.19 20 Several universities use an unsupervised testing option for SARS-CoV-2 testing, however, to our knowledge, this mechanism has yet to be systematically described and evaluated in comparison to a more resource-intensive supervised collection method.1 17 21–23

The University of Washington (UW) Husky Coronavirus Testing (HCT) research study provided SARS-CoV-2 testing at a large, urban university in Seattle, Washington, USA to university affiliates enrolled in the study.1 2 From September 2020 to August 2021, testing was available through supervised self-swabs collected at staffed testing locations on-campus and unsupervised self-swabs returned to the laboratory by courier.2 18 In September 2021, an unsupervised testing option was added in which participants pick up a swab kit from a staffed testing station, self-test (unsupervised) and return the kit to an on-campus drop box. Unsupervised testing was implemented to increase the convenience of testing, accommodate temporary increases in testing demand (eg, during return to campus after academic breaks or during large campus COVID-19 outbreaks) and expand testing access to individuals not using other testing options. Here, we evaluate the feasibility and utility of the unsupervised testing mechanism for a university COVID-19 testing programme compared with a supervised self-swab method.


Study design

HCT was a voluntary research study. The study design and methods have been previously described.1 2 Briefly, English-speaking university affiliates, including faculty, staff and students, above the age of thirteen years were eligible to enrol. Demographics, risk behaviours, symptoms and vaccination status were reported through online questionnaires. Participants were invited to test for SARS-CoV-2 after reporting new symptoms, exposure to known SARS-CoV-2 cases or recent out-of-state travel on electronic daily attestations received by email or text message. In addition, participants were invited to test if they were members of a campus group experiencing an outbreak and walk-in testing was available and could be used for any reason. Data were collected using Project REDCap.24 25

Patient and public involvement

Patients and the public were not involved in the design, conduct, reporting, or dissemination plans of this research.

Swab collection

During the 2021–2022 academic year, COVID-19 testing was available through three testing mechanisms: (1) supervised self-swabs at staffed testing locations, (2) unsupervised self-swabs returned to an on-campus drop box and (3) unsupervised self-swabs returned to the laboratory by courier for participants under quarantine measures or with mobility limitations. The self-testing protocol used in this study was previously evaluated in a home-based respiratory virus surveillance study.18 Here, we describe methods for and compare (1) supervised self-swabs and (2) unsupervised self-swabs returned to on-campus drop boxes only and do not include methods or data for (3) unsupervised self-swabs returned by courier as very few swabs were collected using that mechanism.

To test at supervised testing locations, participants checked in to confirm eligibility, study staff activated a self-swab kit and participants completed the self-swab supervised by study staff. Supervised self-swab samples were transported to the laboratory twice a day. For unsupervised self-swabs, participants confirmed eligibility and picked up self-swab kits at staffed drop box locations on campus. Participants were advised to activate their kit by entering their full name and collection barcode in an online form, self-swab outside or in a private setting, and return the completed kit to an on-campus drop box the same day as swabbing. Drop boxes were secure parcel drop containers labelled with kit return instructions located at several high-traffic areas on campus (online supplemental figure 1). A leaflet with instructions for registering the kit, self-swabbing and returning the sample was included with the self-swab kit (online supplemental figure 2). Samples returned to drop boxes that had not been activated or were incorrectly activated online were discarded. Samples were picked up from drop boxes and transported to the laboratory at least once a day. RHINOstic Automated Nasal Swabs were used for all swabs, except during supply chain shortages when US Cotton #3 swabs were used for some supervised swabs. Both supervised and unsupervised testing locations were closed on university holidays and on 4 days during the analysis period due to inclement weather. Unsupervised swab kit design and assembly are described in online supplemental figure 3.

Supplemental material

Unsupervised testing was used to increase testing capacity during periods of increased testing demand. At the request of university fraternity and sorority organisations (ie, student social organisations with chapter housing with upwards of 50 members and large, frequent social events) and Housing and Food Services (ie, department that manages university dormitories and apartments), unsupervised testing kits were distributed to fraternity and sorority houses and dormitories during move-in week in fall 2021 to reduce wait times at supervised test sites and accommodate return-to-campus testing demand. Staffing hours at both supervised and unsupervised testing locations were extended during periods of increased testing demand, including during return from academic holiday breaks (ie, Thanksgiving (25 November 2021–26 November 2021) and winter (18 December 2021–2 January 2022) and spring (19 March 2022–27 Macrch 2022) breaks and during the Omicron wave (approximately December 2021–February 2022).

Laboratory methods

Collected swabs were entered into a laboratory inventory management system and examined in a class II biosafety cabinet for viability. Swabs that were expired (past 2-day sample viability window), contaminated (excess mucus or blood on the swab, swab damaged or swab improperly packaged), or that were not activated or were not correctly activated by the participant were discarded. Swabs were stored dry (ie, stored in a sample tube without any media or preservative) and eluted with 300 µL Tris-EDTA for RHINOstic swabs or 1 mL Tris-EDTA for US Cotton #3 swabs. A 50 µL of eluate was treated with proteinase K and heat and swabs were tested for SARS-CoV-2 using RT-qPCR, an extraction-free method previously described.26 The RT-qPCR assay uses probe sets for SARS-CoV-2 Orf1b and S-gene that are multiplexed with a probe set from human Rnase P.26 RT-qPCR results were internally verified and transferred to the state department of health, university public health and participants.

Statistical analysis

Descriptive statistics were computed to describe the characteristics of our study sample. We used logistic regression to estimate the odds of a self-swab being conducted through the unsupervised versus the supervised testing mechanism by participant characteristics, including year of study enrolment, pre-Omicron versus post-Omicron time period (1 October 2021–10 December 2021 vs 11 December 2021–20 July 2022), age, sex, race, ethnicity, affiliation, symptom status and among students only, fraternity or sorority membership and being and on-campus resident.27 This regression model was fitted using generalised estimating equations to account for correlation possibly induced by the fact that some individuals provided multiple swabs. An independent working correlation structure was used and Wald-based inference was conducted using robust standard errors. Our analysis excluded swabs collected from 10 September 2021 to 30 September 2021, as during this time unsupervised testing was primarily available to students only (and not faculty and staff) and a ‘return to campus’ testing campaign directly distributed unsupervised swab kits to students. Given the relatively low level of missingness in relevant variables (except vaccination status), our analysis was based on complete-case data alone. All hypothesis tests conducted were two sided and calibrated to a significance level 0.05. All analyses were conducted in R (V.4.1.1, R Core Team, 2021).


From 10 September 2021 to 10 July 2022, results were returned for 119 671 total swabs collected from 26 695 individuals. We compared supervised self-swabs collected at staffed testing sites to unsupervised self-swabs returned to on-campus drop boxes for this analysis; <2% of unsupervised swabs were returned to the laboratory by courier and these tests were not included. During the fall academic quarter from 10 September 2021 to 20 September 2021, unsupervised testing was only available to students. Test kits were delivered directly to students living in residence halls and fraternity and sorority houses and as a consequence, 59% of all student swabs during this time period were unsupervised tests and returned to drop boxes (figure 1). Unsupervised testing became available to faculty and staff on 21 September 2022 and from this time until 1 December 2021, 24% of swabs from students and only 10% of swabs from faculty and staff were unsupervised tests. The use of unsupervised testing increased throughout the academic year. The proportion of all swabs that were unsupervised increased to 24% for faculty and staff during the Omicron wave, from 10 December 2021 to February 2022 but did not increase overall among students (23% of all swabs were unsupervised among students during this time). As testing demand fell in spring 2022, the proportion of swabs that were unsupervised remained constant for both students (22%) and faculty and staff (24%) from March to June 2022, presumably as participants that had become familiar with unsupervised testing continued to use this testing option.

Figure 1

Supervised and unsupervised self-swabs collected from 10 September 2021 to 10 June 2022. Campus events, holidays, breaks in courses, and periods of online instruction that impacted university populations are shown. Testing demand was reduced on weekends and operations were paused for holidays, inclement weather, and campus closures (represented by gaps in testing).

Of the 26 359 unique participants included in this analysis, about half (55%) used supervised testing only, 19% used unsupervised testing only and 25% used both testing mechanisms. Most study participants were students (79%), the median age at enrolment was 22 years (IQR 20–28), 61% were female and about half (49%) reported their race as white (table 1). Individuals reported return from out-of-state travel within 7 days prior to testing for 15.3% of all swabs and being symptomatic at testing for 32%, among those responding to online questionnaires (table 2 and online supplemental figure 4). Among symptomatic individuals, median time from symptom onset to testing was 2 days and 35% reported cough, 52% sore throat, 15% feeling feverish or sweats and 1% loss of smell or taste. Among those who reported COVID-19 vaccination status at the time of testing, 97% had received the primary series or the primary series and a booster dose. The median number of self-swabs completed per participant during the study period was 3 (range: 1–66, IQR: 1–6). Relative to those who completed five or fewer swabs, a greater proportion of individuals who completed six or more swabs were faculty and staff, female, non-Hispanic, white and aged ≥60 years (online supplemental figure 5 and online supplemental table 1).

Table 1

Demographic data on unique participants who tested for COVID-19 through either supervised and/or unsupervised testing mechanisms from 10 September 2021 to 10 June 2022

Table 2

Characteristics of COVID-19 testing events through supervised and unsupervised testing mechanisms from 10 September 2021 to 10 June 2022

During the study period, we distributed 36 508 unsupervised and 92 554 supervised self-swab kits, and of these, the laboratory received 26 800 (73%) and 92 499 (99%) of unsupervised and supervised swabs for testing, respectively (figure 2). Very few (0.1%) supervised swabs received by the laboratory could not be tested, but 4% of unsupervised swabs were not tested because they were not activated or were incorrectly activated (1%) or were expired (3%). Few swabs (0.1% of both supervised and unsupervised swabs) were not tested due to contamination (excess mucus or blood on the swab, swab damaged or swab improperly packaged). The proportion of unsupervised swabs received by the laboratory that could not be tested was highest during fall 2021 (5% overall during September–November 2021) and at the start of the Omicron wave (4% overall from 10 December 2021 to January 2022) and decreased in Spring 2022 (2% overall in March–June 2022 (online supplemental figure 6). Less than 1% of all swabs tested for SARS-CoV-2 resulted in PCR molecular failure (human internal control, RNase P, was not detected). Results were returned for 92 263 supervised and 25 711 unsupervised self-swabs from 26 359 unique participants. Overall test positivity rates, median cycle threshold values for positive samples, and median RNase P values, a proxy for sample quality, were similar between unsupervised and supervised swabs, respectively (online supplemental figure 7 and table 2).

Figure 2

Flow chart of kit distribution to return of results for supervised and unsupervised self-swabs from 10 September 2021 to 10 June 2022. (1) Past 2-day sample viability window. (2) Excess mucus or blood on the swab, swab damaged or swab improperly packaged. (3) Human internal control (RNase P) not detected.

Our adjusted analysis of participant characteristics associated with unsupervised testing included 103 383 self-swabs with complete participant data (96.5% of all swabs collected from 1 October 2021 to 10 July 2022, online supplemental table 2). We estimated the odds of using unsupervised testing to be significantly lower prior to versus during the Omicron wave (ie, before vs after 10 December 2021) (estimated OR 0.5, 95% CI 0.5 to 0.5, p<0.001); among individuals who enrolled during the first versus second year of the study (ie, before vs after 10 September 2021) (0.3, 95% CI 0.4 to 0.4, p<0.001); and among individuals symptomatic versus asymptomatic or presymptomatic at the time of testing (0.9, 95% CI 0.8 to 0.9, p<0.001). The odds of using unsupervised testing were estimated to be greater among individuals of Hispanic versus non-Hispanic ethnicity (1.2, 95% CI 1.0 to 1.3, p=0.01). The odds of unsupervised testing also differed significantly by age group (p<0.001), with individuals 25–29 years (0.8, 95% CI 0.7 to 0.9), 30–59 (0.9, 95% CI 0.9 to 1.1) and ≥60 years (0.8, 95% CI 0.6 to 0.9) estimated to have a lower odds compared with individuals 15–24 years of age. The odds of unsupervised testing differed significantly by affiliation (p<0.001), with faculty (1.6, 95% CI 1.3 to 1.9) and staff (1.5, 95% CI 1.3 to 1.7) estimated to have higher odds compared with students overall, despite very low utilisation of unsupervised testing among faculty and staff prior to the Omicron wave. Among students only, odds of using unsupervised testing were estimated to be 32% lower (0.7, 95% CI 0.6 to 0.8, p<0.001) among fraternity and sorority members relative to non-member students, and 2.5 times greater (95% CI 2.3 to 2.6, p<0.001) among on-campus residents relative to non-residents. Participant COVID-19 vaccination status was missing at the time of testing for 23% of self-swabs and thus was not included vaccination status in the adjusted analysis. In an unadjusted analysis, unvaccinated individuals (2.1, 95% CI 1.7 to 2.6) and individuals who had received a primary series and a booster dose (1.5, 95% CI 1.4 to 1.5) had greater estimated odds of using unsupervised testing relative to individuals who had received the primary series only.


We evaluated the use of unsupervised self-swabbing as part of a university COVID-19 testing programme to increase accessibility and to enable rapid scale up of testing. Unsupervised testing allowed us to capture a larger and more diverse proportion of the campus community in our COVID-19 testing programme, and the flexibility to accommodate increased testing volume proved critical following academic breaks and during the Omicron wave of the SARS-CoV-2 pandemic. Using unsupervised testing, testing capacity could be rapidly increased on short notice relative to supervised testing at staffed locations, primarily because operating unsupervised testing with return to drop boxes did not require additional staff or a clinical space for observed swabbing.

Extreme variation in COVID-19 testing demand is challenging to accommodate using hired staff due to expense, training and lack of need during the majority of time. Unsupervised testing allowed our COVID-19 testing programme to accommodate temporal fluctuations in testing demand using a convenient testing method, and this is critical for engaging populations with fluctuating testing interests. At the start of the Omicron wave, unsupervised testing increased markedly when capacity was strained at on-campus staffed testing locations and community-based testing locations. Even after testing demand declined at the end of the Omicron wave in early spring 2022, unsupervised testing remained higher than prior to the Omicron wave. This sustained use of unsupervised testing was particularly apparent among faculty and staff, who used unsupervised testing methods only minimally prior to the Omicron wave. This suggested that some participants began using unsupervised testing at times of high testing demand, and after becoming familiar with this mode of testing, may have been more likely to use unsupervised testing during subsequent times of lower testing demand.

Unsupervised testing also captured a unique population of individuals. Younger, Hispanic and unvaccinated individuals, populations at higher risk for COVD-19 infection,28–30 had increased odds of using unsupervised versus supervised testing. Although the use of unsupervised testing among faculty and staff was low prior to the Omicron wave, overall odds of using unsupervised versus supervised testing were greater for faculty and staff relative to students. Although we were not able to assess testing preferences of individuals or convenience of testing site locations relative to individual’s homes and work or study locations on campus, differences in demographic factors that we identified between unsupervised and supervised testing suggest that the unsupervised option allowed our testing programme to capture a broader and more representative proportion of the university community. This allowed more accuracy in our university COVID-19 testing programme, which enables tracking of transmission in the campus community, informs public health response and provides epidemiological data on younger-than-average healthy adults who rarely develop severe disease and are less often studied than more vulnerable groups.1 28 In addition, the findings of this study are consistent with other research showing that self-testing is an acceptable, feasible and reliable SARS-CoV-2 testing mechanism.31 32

Among swabs that were correctly activated and received by the laboratory, a greater proportion of unsupervised swabs could not be tested because they were expired (past 2-day sample viability window). Although swabs were picked up from both supervised testing locations and drop box locations at least once per day, about 3% of all unsupervised swabs expired on laboratory receipt, likely due to individuals self-testing and dropping off the sample one or more days later. In contrast, supervised samples used for testing were always picked up on the same day. In addition, a very small proportion of swabs were contaminated (0.1% for both supervised and unsupervised) or resulted in molecular failure (0.2% and 0.5% for supervised and unsupervised swabs, respectively). Greater molecular failure among unsupervised swabs was presumed to be due to incorrect swabbing technique compared with supervised swabs. Despite these findings, unsupervised testing still resulted in minimal testing failures and the correct use of unsupervised swabs returned to drop boxes increased with continued use of this method over the academic year.

When unsupervised testing was first made available to students prior to the start of the fall academic quarter, large numbers of unsupervised test kits were distributed directly to students. A relatively high number of these kits were not tested due to kits that were registered but not returned or returned but not registered or incorrectly registered. Likewise, a relatively high number of unsupervised swab kits could not be tested at the start of the omicron wave. We believe that unfamiliarity with this testing process (ie, the need to register the kit prior to returning the swab kit) contributed to these failure rates, especially among individuals using unsupervised testing for the first time. Incorrect use of unsupervised tests declined in spring 2022, presumably as the campus community became more familiar with this testing method. Clear and easy-to-follow instructions for unsupervised testing that communicate the importance of returning the swab kit on the same day as swabbing will be important for the success of any future unsupervised testing programmes, particularly for individuals using unsupervised testing for the first time.

This study has several limitations. First, we did not collect data on participant’s preference for each testing mechanism and are unable to assess the exact reasons for individuals choosing to use unsupervised versus supervised testing. In particular, during the 2020–2021 academic year, testing was only available at staffed self-testing locations and participants that enrolled during the 2020–2021 year, vs the 2021–2022 academic year, had greater odds of using supervised testing, likely due to familiarity with that testing mechanism. In addition, drop boxes and supervised testing locations were located in different areas on campus and we did not assess the degree that convenience of a testing location drove participant preference. For example, one of the drop box locations was in a dormitory building and we observed that on-campus residents had greater odds of using unsupervised testing relative to other students. Unsupervised testing also relied on the participant’s ability to complete online forms and self-swab without assistance, and we did not investigate if participants who were unwilling or unable to do this preferred supervised over unsupervised testing. Second, travel and other data collected at the time of testing relied on participants’ self-reports using online questionnaires. Individuals using unsupervised testing were prompted to answer daily attestation questions regarding travel when registering their kit online and a greater proportion individuals using unsupervised testing responded to these questions compared with individuals using supervised testing, where the questions were optional. This limited our ability to directly compare the likelihood of obtaining unsupervised vs supervised testing after travel, a potentially high-risk period for infection. Third, free rapid antigen self-tests became widely available by home delivery through both the federal and Washington state governments during the study period in January 2022 and we were unable to assess how this may have impacted participant decisions to take a PCR self-test through the study. Fourth, participant vaccination status was unknown at the time of testing for over 20% of self-swabs and we were not able to include vaccination status in our adjusted analysis comparing odds of unsupervised versus supervised testing.

Our study captured demographic, behavioural and clinical information about university affiliates seeking SARS-CoV-2 testing on a university campus with two testing methods during periods of fluctuating testing demand. Our findings suggest that unsupervised testing is an acceptable and resource-efficient way for universities to conduct high-throughput routine molecular COVID-19 testing. Unsupervised self-testing with return of samples to drop boxes also allowed us to accommodate massive variations in testing demand and had a low rate of testing failure. Testing programmes that can adjust to and meet changing testing demand are of value as universities transition away from emergency COVID-19 response and towards longer-term, routine public health efforts and surveillance of respiratory infections.

Data availability statement

Data are available on reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and the University of Washington IRB approved this study (#00011148). Participants gave informed consent to participate in the study before taking part.


We would like to thank the study participants. We also thank the UW Environmental Health and Safety COVID-19 Prevention and Response team including, Katia Harb and Julie Skene, UW Administration and Incident Command Leadership Group (Margaret Shepherd, Josh Gana, Pamela Schreiber, Jack Martin), the Chu Lab, the Brotman Baty Institute, Dr. Jan Englund, and Dr. Timothy Uyeki. Finally, we thank HCT study research assistants for their work running and managing testing sites: Alina Cordova, Alysse Daniels, Anna Elias-Warren, Ashley Subijanto, Aurora Mattson-Hughes, Brigitte Neuville, Casey Honz, Christian Gombio, Cody Krutzsch, Daniel Nguyen, Ella Howard, Hadar Dolev, Hana Arega, Hannah Diaz, Helen Nguyen, Izabel Stohel, Jennifer Khuc, Katie Bezaitis, Kevin Cai, Kino Watanabe, Lauren Shrestha, Liem Nguyen, Luke Johnson, Madeline McMannis, Maria Matlick, Molly Holmes, Monique Samodien, Nathalie Castro, Niya Carrera, Odalys Nuno, Quyen Pham, Rino Watanabe, Ryan Im, Sela Wein, Shailah Fakhri and Shiuli Pemmaraju.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • X @jucbennett

  • Contributors Acquired the data: JO'H, ZA, PDH, DM, TW, KGL, LR, KMM, BP and CRW. Conceptualised and designed the analysis: JCB, MC, HYC and AAW. Analysed the data: JCB, JO'H and ZA. Interpreted results: JCB, JO'H, ZA and AAW. Wrote the initial draft of the article: JCB, JO'H and ZA. Revised the article: JCB, GG, JPH, MC, LS, HYC and AAW. Acquired financial support for the project: HYC. Guarantor: HYC.

  • Funding This work was supported by the CARES act, University of Washington Grant # 624611.

  • Competing interests HYC reports consulting with Ellume, Pfizer, The Bill & Melinda Gates Foundation, Glaxo Smith Kline and Merck. HYC received research funding from Gates Ventures, Sanofi Pasteur and support and reagents from Ellume and Cepheid outside of the submitted work. GG received research grants and research support from the US National Institutes of Health, the University of Washington, the Bill & Melinda Gates Foundation, Gilead Sciences, Alere Technologies, Merck & Co., Janssen Pharmaceutica, Cerus Corporation, ViiV Healthcare, Bristol-Myers Squibb, Roche Molecular Systems, Abbott Molecular Diagnostics and THERA Technologies/TaiMed Biologics, all outside of the submitted work.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.