Quantitative analysis.

We summarized participant characteristics by estimating mean, standard deviation, frequency and proportion for all categorical and continuous sociodemographic and health factor variables, respectively. Additionally, we estimated the mean, standard deviation, median, and interquartile range (IQR) for each TFA construct measure and also estimated the frequency and proportion of participants who stated that they strongly agreed or agreed with each statement. We calculated overall adherence by taking the number of submitted test results and dividing it by the 21 days of the study. We categorized adherence into three groups and estimated the number and proportion of participants per group: 100% (tested daily), 50%-99% (tested at least every other day), and <50% (tested less than every other day). Finally, we calculated the frequency and proportion of participants who stated they would prefer testing daily with a rapid antigen test versus their current PCR regimen if given the choice.

Qualitative analysis.

Interview transcripts were analyzed by a team of three researchers in Dedoose using Template Analysis, which is appropriate for studies using a priori themes and exploring perspectives of different groups in an organizational context [17]. We initiated data analysis with a top-down approach, using the seven constructs in the TFA to create an organizational template. One member of the research team created a preliminary codebook by reading through interview debrief forms and identifying subconstructs within each of the TFA constructs. With the seven TFA constructs listed as parent codes and the subconstructs as child codes, the initial codebook was applied to two interview transcripts in Dedoose. A second researcher then applied the preliminary codebook to two more interviews and the codebook was further discussed and refined. All three coders then re-coded the initial four interviews using the finalized codebook to maximize reliability across coders. The three coders then individually coded the remaining 11 interviews, meeting on an ad hoc basis to resolve coding questions, and using memos to note emerging themes. The coders generated code reports for each TFA construct, identified key quotes for each construct and subconstruct, and held team meetings to discuss emergent themes that conceptualized how each construct either directly or indirectly influenced acceptability and adherence. Lastly, we created quote matrices for each construct and subconstruct using the identified key quotes organized by emergent themes.

Adherence to daily testing regimen.

Mean adherence to the 21-day testing regimen was 88% (SD: 21%, Median 95%, IQR: 90%, 100%) with 43% of participants achieving 100% adherence, 48% achieving 50–99% adherence, and 10% achieving less than 50% adherence (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. COVID-19 testing adherence and preference.

https://doi.org/10.1371/journal.pone.0267766.t002

Affective attitude.

Participants reported overall positive feelings about the intervention though only 58% agreed with the statement “I like using the test every day” (see Table 3 for mean and median Likert scale scores). The overwhelming feeling reported by participants was comfort and peace of mind (Table 4, Affective attitude 1). Testing negative gave them needed reassurance to go about their daily lives and participants reported feeling “liberated” to engage in everyday activities without fear of infecting others (Affective attitude 2, 3). Participants also described using the tests after engaging in a “high risk” activity or potentially being exposed to COVID-19 (Affective attitude 3). Notably, a small number of participants described significant anxiety due to living with someone who was immunocompromised; for these participants, the ability to test daily or on demand provided significant mental health benefits (Affective attitude 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Quantitative measurement of Sekhon’s domains of acceptability of a COVID-19 daily antigen testing regimen in an employment context (N = 31).

https://doi.org/10.1371/journal.pone.0267766.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Acceptability of a COVID-19 daily antigen testing regime in an employment context according to Sekhon’s seven domains of acceptability.

https://doi.org/10.1371/journal.pone.0267766.t004

For some participants who struggled to incorporate testing into their daily routine or who had difficulty using the mobile app to capture and interpret their test result, daily testing was more burdensome. In these cases, participants reported feeling dread when they remembered they still had to test or frustration at being repeatedly unable to scan in a test result through the mobile app (Burden 6, Opportunity cost 5).

Burden.

More than half of participants disagreed with the statements “it is inconvenient to use the self- test every day” (61%) (Table 3). Most participants describing daily self-testing as very convenient compared to other testing options they experienced (e.g., large drive-through testing events early in March/April 2020 or mandatory testing offered by their employer). Even among participants who had access to weekly on-site testing at their workplace, most preferred the daily antigen test. These participants overwhelmingly described testing daily as fast and easy often noting that daily self-testing took less time overall than weekly PCR testing (Table 4, Burden 1); that they didn’t mind daily testing because they could test on their own schedule and get the results quickly (Burden 2) or incorporate self-testing into other established routines (e.g., timing the 15-minute waiting period with a shower) (Burden 3); and that they did not find the daily swabs to be invasive (Burden 4). While some participants described mild discomfort from the daily swabbing, this was not a major burden.

In the minority were participants who acknowledged that while daily at-home antigen testing was more convenient than weekly PCR testing, they still preferred testing using PCR because of the mental energy required to self-administer the test (Burden 5). Other participants described difficulty incorporating testing into their daily routine and reported testing fatigue (Burden 6).

Self-efficacy.

Participants reported that the self-test was easy to learn to use (90%) and results easy to understand (100%) (Table 3). However, only half of participants (55%) stated that they developed a habit of testing for COVID-19 in their daily routine, with experiences adhering to daily testing varying widely across participants. While some participants reported feeling initially overwhelmed by the testing steps, nearly all reported mastering the steps after a few days of testing (Table 4, Self-efficacy 1, 2). Factors that supported daily adherence included developing strategies to streamline the testing process (Self-efficacy 2); incorporating the test into their daily routine (Self-efficacy 3, 4); and placing the test kit in a prominent location (Self-efficacy 3). In contrast, participants reported that changes to their daily routine and weekends were the greatest barrier to daily adherence (Self-efficacy 5, 6).

Opportunity cost.

Most participants strongly agreed with the statement: “It is easy for me to find a time to use the test every day” (68%) (Table 3). Nonetheless, time was often described as the biggest opportunity cost to testing daily (Table 4, Opportunity cost 1, 2). Some participants attempted to minimize the time cost by multi-tasking (Burden 3) or spreading out the time commitment by setting up their test kit the night before (Self-efficacy 3). However, others reported that efforts to multi-task resulted in more time spent testing because they would lead to mistakes like missing the results window during and having to repeat the testing process again (Opportunity cost 1, 2, 3). For other participants, access to strong internet and/or a good light source were additional opportunity costs to daily testing as the mobile app required a specific level of light to accurately interpret the test result (Opportunity cost 4, 5). Participants who struggled to upload a picture of their test result consequently described testing as a “daily hassle” and “frustrating”, negatively impacting their affective attitude about the testing regimen.

Ethicality.

Most participants agreed that it is ethical for an employer to require its employees to test regularly for COVID-19 (87%) (Table 3). However, a small minority believed that regular testing should be a personal choice (10%) and a significant portion of participants believe that sharing test results should be a personal choice (42%). Many participants articulated how working long hours near other coworkers necessitated the need for routine testing and saw mandatory testing as a practical approach to keeping the workplace safe (Table 4, Ethicality 1, 2). Specifically, many used Utilitarian arguments to justify mandatory testing, saying that the small loss of privacy and autonomy was justified for “the greater good” of workplace safety (Ethicality 3, 4, 5). Others went further, stating that routine testing helped them fulfil their personal responsibility to keep their workplace safe, noting that they “didn’t want to be transmitters” and that giving someone else COVID would be “an unacceptable thing” (Ethicality 6). In contrast, a small number of participants believed mandatory workplace testing was not ethical as it infringed upon individual liberties. While these participants were personally willing to test regularly through their employer, they believed that coercive tactics should not be used to keep the workplace safe (Ethicality 7).

While participants tended to feel strongly that requiring regular COVID-19 tests was either ethical or unethical, participants’ feelings about who should have access to test results and what this health information should be used for was more nuanced. Many agreed that test results should be reported to their employer to prevent the spread of infection should an employee test positive (Ethicality 8); however, some had concerns about who within their company should have access to results (Ethicality 9). Participants also had mixed feelings regarding whether COVID-19 test results should be shared with larger health organizations, such as state health departments or the Centers for Disease Control and Prevention (CDC). Most participants believed these were trusted entities and saw the benefit of sharing COVID-19 results for contract tracing, surveillance, and allocation of health services (Ethicality 10); however, this feeling was not universal, and some reported not trusting the CDC or worrying that the health department would limit their freedom if they tested positive (Ethicality 11, 12). Additionally, a small number of participants believed test results should stay within their employer due to concerns about health information privacy, including fears that their health information could be sold or compromised (Ethicality 13, 14). At the same time, many participants liked the idea of their test results being automatically reported to their employer or doctor through a mobile app, as this would eliminate the need for a third party to handle test results (Ethicality 15).

Coherence.

The majority of participants believed that testing for COVID-19 daily made sense to prevent the spread of COVID-19 at and beyond the workplace (77%) (Table 3). Support for daily testing was particularly strong for times when COVID-19 rates were increasing or high; in this context, regular testing was considered an effective and necessary strategy for monitoring infection rates and preventing COVID-19 transmission (Table 4, Coherence 1). Specifically, many participants believed the unique needs of the entertainment industry, which relies on non-replaceable people to fill specific roles, further necessitated a stricter approach to COVID-19 control to ensure that work would not be disrupted (Coherence 2). Many participants said they wish they had access to daily testing during the original peak of the pandemic in March and April of 2020 (Coherence 3).

While support for regular testing was near universal during times when risk of COVID-19 infection was perceived to be high (i.e., before vaccines were available, when rates of infection were still increasing), there was less agreement about when routine testing was no longer needed. Many participants thought it made sense to continue regular COVID-19 testing until there was more research available on the transmissibility of emerging COVID-19 variants and effectiveness of the COVID-19 vaccines (Coherence 5). Others preferred to listen to scientific recommendations from public health agencies to decide whether frequent testing made sense (Coherence 6). Some participants expressed increasing doubt that regular COVID-19 testing was still necessary given that vaccines were widely available to all adults and rates of COVID-19 in the workplace and surrounding community were very low (Fig 1). One participant in a managerial position used a cost benefit analysis to argue that frequent testing did not make sense from a financial perspective given the low positivity rates (Coherence 7). Many believed testing did not make sense for vaccinated populations, but that it may be an alternative for those who did not wish to become vaccinated (Coherence 8).

Perceived effectiveness.

Participants had high confidence in the effectiveness of a daily testing intervention with the majority reporting that they trusted the results of the at-home antigen test (81%) and nearly all expressing confidence that testing daily will keep the people they work with safe from COVID-19 (90%) (Table 3). Whether participants perceived the daily testing intervention would keep the workplace safe depended on three factors: 1) whether they trusted the test results, 2) whether they believed their co-workers would be willing and able to test daily, and 3) whether they believed frequent testing would introduce any unintended consequences. Although the antigen test used in the study had not yet received FDA approval and was only for use in research settings, nearly all participants reported trusting the results, citing trust in Columbia University (which helped develop the test in collaboration with a biotechnology company), trust in Disney (their employer) to provide a safe and effective test, and trust in experts/scientists (Table 4, Perceived effectiveness 1, 2). Other participants drew on previous experience using similar at-home tests (e.g., home pregnancy tests, glucose monitoring, maintaining swimming pool pH) as reasons for trusting the results (Perceived effectiveness 3). One participant noted that they trusted the daily antigen test result more than weekly PCR because the antigen test result reflected current COVID-19 status while the PCR result reflected a status from 24–48 hours ago and could give a false sense of security (Perceived effectiveness 4). In contrast, a different participant reported trusting the antigen test result less than PCR because they had heard from family and friends with medical training that antigen tests were not as accurate as PCR (Perceived effectiveness 5).

While participants were generally confident in their own ability to successfully use the at-home antigen test, they were less confident about their co-workers’ abilities, which they believed could undermine the effectiveness of the daily testing intervention. These concerns included not believing that their coworkers would obtain a sufficient swab sample and properly self-administer the test at home (Perceived effectiveness 6), believing that a test administered by a medical professional would be more accurate (Perceived effectiveness 7), and worry that the identity of the test taker and the test result could not be confirmed in a home setting (Perceived effectiveness 8). Finally, some participants raised concerns that a frequent testing regimen would lead to employees increasing risky behavior that could result in more COVID-19 transmission (Opportunity cost 9, 10).