Mobilizing Community-Driven Public Health: Increasing Access to Diagnostic Testing for Underserved Individuals Through Lab-In-A-Van Partnerships

HIGHLIGHTS

• •Mobile testing is a feasible and accessible model that is highly accepted by the public.
• •A cargo van can be outfitted to provide high-complexity clinical diagnostic testing.
• •The lab-in-a-van model can increase access to diagnostic testing in low-resource settings.
• •Community-informed testing programs can help overcome barriers and increase health equity.

Abstract

Introduction

Underrepresented populations, who often have the highest need for accessible health care, frequently face barriers to accessing services. To address health inequities, researchers evaluated a community-informed lab-in-a-van model that was designed to increase access to COVID-19 testing for vulnerable populations.

Methods

Free saliva-based SARS-CoV-2 diagnostic testing was offered between June 2023 and July 2024 at 123 events in Connecticut, U.S., using a Clinical Laboratory Improvement Amendments–certified mobile laboratory. Approximately 100 local leaders and organizations informed the program’s design. After providing samples, participants completed an IRB-approved survey; responses were stored using REDCap.

Results

Offering on-site testing helped remove barriers and increased access for traditionally underserved communities. Overall, 1,428 individuals were tested, and 838 completed a testing experience survey. Of the respondents, 54% identified as Black, Indigenous, People of Color; 59% reported an annual household income <$25,000; and 31% were uninsured. Participants reported that the service was easy to access (74%) and comfortable to use (75%); 29% received their first COVID-19 test, and 48% were unaware of alternative testing options. Results were reported in an average of 3.1 hours, enabling many participants to receive their diagnostic results during the community event; 48 positive samples were identified.

Conclusions

This program evaluation demonstrates that community-informed mobile testing programs capable of delivering same-day diagnostic results can expand access to testing among underserved populations and may represent a scalable strategy for improving diagnostic access during future public health emergencies.

INTRODUCTION

Addressing health disparities in vulnerable communities remains a challenge, even as social awareness of disparities has increased.¹ The coronavirus disease 2019 (COVID-19) pandemic offered many examples of how public health response efforts attempted to address such challenges, with varying degrees of success. At 4% of the world's population, the U.S. accounted for 16% of the world’s COVID-19 deaths by the end of 2022.² This catastrophic event exposed fracture lines in the construct of the public healthcare system—specifically, a lack of consistently accessible, affordable, and accurate diagnostic and clinical management solutions​.³

Those most impacted by systemic failures were the vulnerable and uninsured communities; disparities emerged among socioeconomically disadvantaged groups, attributed to the lack of health insurance, access to transportation, and individual awareness of the disease severity​​.⁴ This highlighted the need for effective solutions that went beyond making COVID-19 tests available at a pharmacy or through the mail-order programs organized by the U.S. government. There needed to be improved testing, surveillance, and monitoring; data transparency; and targeting of public health interventions that engaged the community and proactively closed the gap. As such, the project team worked with local leaders from government, nonprofit, and academic backgrounds to explore whether a mobile, saliva-based testing program could overcome the barriers to testing in low-resource communities. The team included leadership from SalivaDirect, Inc., an independent nonprofit dedicated to enhancing public health by developing and deploying accessible testing strategies, and Yale Pathology Labs, a Clinical Laboratory Improvement Amendments–certified laboratory within the Yale School of Medicine able to offer high-complexity testing, the most stringent category of laboratory tests, in compliance with the federal Clinical Laboratory Improvement Amendments. This program focused on the Greater New Haven Area of Connecticut, U.S., where lower-income households and residents of color face significant barriers to accessing health care.⁵ For example, between 2020 and 2021, the mortality rate due to COVID-19 for Black residents in the New Haven area was more than twice the mortality rate for White residents.⁶

With this in mind, the study team aimed to use a full lab-in-a-van setup to perform tests on site in the Greater New Haven area and deliver diagnostic results directly to community residents with minimal turnaround time. By offering testing and same-day results at trusted community sites, researchers also aimed to show that the mobile testing model could provide easily accessible, noninvasive, reliable diagnostic testing to uninsured and low-income individuals. Saliva-based testing was offered because it is a noninvasive sample type, reducing aversion to testing; saliva can also be readily self-collected, reducing the risk of viral transmission to healthcare workers. Results from the same-day testing could then be used by participating individuals to make informed health decisions, such as seeking follow-up care or notifying their close contacts. In addition, this study served as an opportunity for evaluating whether vulnerable populations would be willing to use an on-site mobile testing option and, if so, understand the community’s impressions of mobile testing services. This information could then be used to gauge whether and how mobile programs that offer high-quality, same-day diagnostic testing services could be incorporated into future public health response efforts, looking beyond COVID-19 to explore additional applications, such as testing for sexually transmitted infections, offering broader respiratory panels, and more, where results delivered to participants within hours could shape efficient responses for future public health crises.

This program evaluation examined the implementation of a community-informed mobile diagnostic testing model designed to increase access to COVID-19 testing among underserved populations. Specifically, the study team evaluated the program’s reach, testing outcomes, and participant perceptions of the accessibility and acceptability of the lab-in-a-van model.

METHODS

This study used a community-engaged program design and a quantitative post-test survey to examine the reach, testing outcomes, participant accessibility, and acceptability of a mobile diagnostic testing program. In the months before mobile testing began, community leaders and organizations with direct connections to the target population were emailed Google Form surveys and asked to identify where, to whom, when, and how testing events should take place. This engagement process was used to establish program parameters, understand community needs, and inform testing logistics. Responses were exported into Google Sheets and reviewed to identify patterns and build an informed framework for deploying the mobile testing program (Figure 1). This draft framework was circulated back to survey respondents through email to confirm accuracy and solicit additional feedback, which was provided through email or phone conversations. This iterative design process continued with an in-person community kick-off meeting, where key stakeholders from local government, academia, social service organizations, and others met for a half-day workshop to refine the proposed model. Stakeholder presentations, small group breakout sessions, and collaborative conversations narrowed down the proposed list of test site locations and times; determined the formatting style, wording, and imagery used in engagement materials; informed the language support needs of the target community; and did more (Appendix 1, available online).

Figure 1.

Community-engaged mobile testing program design framework.

Note: This framework illustrates the 4-phase, community-engaged process used to design, implement, and refine the mobile testing program. Approximately 100 community leaders and organizations informed program development through surveys, a half-day community kick-off workshop, and ongoing feedback throughout the study period. As indicated by the dashed feedback loop, insights gathered during monitoring and adaptation continuously informed earlier phases, including the reassessment of community needs, the refinement of the program framework, and the adjustment of implementation strategies. An implementation toolkit informed by the learnings from this program is available to support the replication and adaptation of mobile community testing programs in other settings.

The evaluation focused on 4 domains: (1) program reach, measured by the number and characteristics of individuals tested; (2) implementation characteristics, including the number and location of testing events and turnaround time for results; (3) participant experience and accessibility, measured through post-test surveys; and (4) testing outcomes, including the number of positive cases identified.

Once on-site testing began, in situ observations were documented by project team members, including the number of people who approached the van, interacted with team members on site, or declined to participate. This information was tracked in an Excel document, as was feedback offered organically to the team members on site by testing event partners and participants. Throughout the study period, the project team reviewed the record of in situ observations, verbal feedback, and regular email correspondence with partners to identify themes and improve the mobile testing experience. In addition, after testing at the mobile site, each participant was asked to complete a formal survey to learn about their experience using the mobile van (e.g., comfort, accessibility, convenience) as well as to document their demographics, previous testing experiences, day-to-day access to health care, and others (e.g., insurance status, availability of transportation, knowledge of services). At the conclusion of the study, informal virtual interviews were conducted with representatives from 3 community partners to inform future program development; these insights were used to contextualize program implementation but were not analyzed as formal qualitative data.

Community leaders and organizations were engaged throughout the project. Input was sought from approximately 100 organizations and local leaders representing or actively working with low-resource neighborhoods in Connecticut, including local health departments, community resource distribution centers, food banks, religious institutions, and neighborhood associations. Iterative engagement through emails, phone calls, surveys, and in-depth conversations with key stakeholders was used to develop an initial approach, refine and expand outreach strategies, identify events or services to pair with testing events, and better understand barriers and evolving community health needs (example is provided in Appendix 1, available online). Responses from these community representatives were leveraged to inform testing locations and engagement strategies, lend their credibility to establish trust in the mobile testing program with target populations, solicit feedback from stakeholders at testing events, and encourage participation in mobile testing events. In consultation with community partners, including local public health departments, SalivaDirect, Inc. built and managed an online scheduling platform where testing events could be requested by community partners and the broader community through a Calendly software form. An Excel-based mobile testing deployment calendar was developed by SalivaDirect, Inc. and updated weekly to track upcoming events, confirm staffing, and summarize testing counts (e.g., number of tests administered at each event, total tests to date, number of positives identified, and others).

With support from the NIH Rapid Acceleration of Diagnostics - Underserved Populations (NIH RADx-UP),⁷ the interior of a cargo van was modified to provide clinical-level diagnostic testing (Video 1). The mobile testing unit itself, operated as a Clinical Laboratory Improvement Amendments–certified site under the Yale University School of Medicine’s Department of Pathology, was built out with the bench space needed to house compact DNA-extraction and PCR systems (Myra automated liquid handling system and Mic real-time PCR cycler, Bio Molecular Systems), effectively creating a satellite laboratory capable of rapid on-site testing. The mobile unit had appropriate specimen storage and handling facilities, biosafety cabinets, and the molecular biology machinery necessary to perform diagnostic testing. The laboratory team, with laboratory personnel from Yale Pathology Labs, was equipped with all necessary software to ensure proper consent documentation and participant intake into RELAY (Remote Electronic Labs at Yale), a Yale-developed Laboratory Information System that integrated directly with the study team’s primary Laboratory Information System, Co-Path, to ensure seamless data exchange and workflow coordination as well as accurate test result delivery and streamlined data sharing (example report is provided in Appendix 2, available online).

Free severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) testing was offered to all interested parties present at community-hosted events. Participants were tested following the SalivaDirect RT-qPCR protocol, an extraction-free method designed to be a cost-effective, simple test for laboratories that has been granted Emergency Use Authorization by the U.S. Food and Drug Administration (EUA 202097).⁸ Events included community festivals, back-to-school days, public health awareness/wellness fairs, food distribution sites, social service centers, and community health centers. Being on site offered the opportunity for in-person health education, such as conversations with community health workers and the distribution of informational fact sheets by the project team. It also served to generate familiarity with the project team and testing availability through introductions facilitated by community partners. This helped increase awareness of the on-site testing services and built rapport between the project team and prospective participants who might otherwise decline testing owing to discomfort or a lack of trust in the offered health services. Following an IRB-approved protocol (IRB Protocol Number 2000034551), individuals were electronically consented by technicians on site at the mobile van location. Consent forms were available in English and Spanish; additional language services were provided by on-site community partners. After obtaining consent, participants were given a sterile plastic tube and directed to provide a self-collected saliva sample. Samples were then returned to the laboratory technicians for immediate testing on the van using the SalivaDirect protocol.⁹ Participants were given the option to wait on site for their results, available as a printed report in as little as 2 hours; have test results sent to their provided email from the Health Insurance Portability and Accountability Act–compliant RELAY system; or have results mailed as a printed report to their provided address. During the hours the van was operational at testing sites, a community health worker or navigator, provided by the community partner hosting the event, was present as a resource for health-related questions. Health workers followed up with individuals who had positive test results, using a template email or call script; provided information about at-home care and precautions based on Centers for Disease Control and prevention guidelines; and aided in accessing local healthcare resources.

After sample provision, participants were asked to complete an IRB-approved survey about their mobile testing experience (e.g., ease of access, comfort, reason for choosing the mobile testing site), provide demographic information (e.g., age, income, education level, race), and respond to additional questions from the standardized NIH RADx Data Hub Tier 1 Common Data Elements (Appendix 3, available online, provides complete survey information). Participant responses were tracked digitally through Wi-Fi–enabled tablets on site and securely stored using REDCap survey technology and electronic data capture tools hosted at Yale University.^10,11 Survey responses were exported from REDCap as a CSV (Comma-Separated Values) file and then imported and analyzed in Excel. Incentives ($10 gift cards) were made available to every participant at all testing events held after October 2023; this was in response to feedback regarding the time required to participate, particularly the minutes needed to complete the accompanying testing experience survey questions. All study participants were eligible to complete the survey and were invited to do so, although not all elected to participate even when offered an incentive for their time. For participants aged <18 years, the respective consenting adult supported or facilitated survey responses. For those unable to read survey questions, on-site support was available from the project team, health workers, and/or community partners to offer additional on-site translation(s), assist with language comprehension, and navigate the survey’s digital interface.

RESULTS

Between June 2023 and July 2024, 123 community testing events were hosted across 9 ZIP code areas in 6 Connecticut cities—West Haven (56%, 69 of 123), New Haven (22%, 27 of 123), New London (15%, 19 of 123), Shelton (4%, 5 of 123), Ansonia (2%, 2 of 123), and Bridgeport (1%, 1 of 123) (Figure 2 and Appendix 4, available online). The engagement strategy as well as testing event identification was informed by input from approximately 100 organizations and local leaders representing or actively working with underserved neighborhoods in Connecticut. From these events, 1,428 SARS-CoV-2 PCR tests were administered. The results are derived primarily from quantitative survey data and program implementation metrics, supplemented by observations documented during testing events.

Figure 2.

Overview of Connecticut mobile laboratory project timeline and testing (February 2023 to July 2024).

Note: This timeline includes key study milestones (detailed below the year line), notably the lead time between the start of community engagement and the first testing event, the gaps between initial testing events, and the increasing frequency as recurring test sites were established. It also shows the increase in tests administered over time (blue) as $10 gift card incentives were introduced and recurring test sites were established and the increase in positive tests over time (red) that the van was able to identify.

Of the 1,428 participants tested at the mobile laboratory, 838 (59%) responded to the accompanying participant experience survey afterward; the extent of survey participation and subsequent completion varied among test takers, resulting in different denominators, because some questions had more respondents than others. The age of survey respondents ranged from 0 to 91 years (mean=43 years); 54% (302 of 564) identified as Black, Indigenous, or People of Color; 53% (285 of 536) identified as male; 41% (218 of 536) identified as female; 1% (7 of 536) identified as nonbinary; and 5% (26 of 536) preferred not to answer. A total of 59% (285 of 486) reported an annual household income <$25,000; 91% (442 of 486) reported <$50,000; and 92% (725 of 791) were residents of 9 Connecticut ZIP codes (Appendix 4, available online). Just over half of the participants had never been vaccinated against COVID-19 (55%, 294 of 540). Of those who responded to questions about symptoms at the time of testing, as described in the NIH RADx Data Hub Tier 1 Common Data Elements (Appendix 3, available online), varying numbers of participants responded to each symptom question: 3% (23 of 641) reported fever, 10% (69 of 664) reported cough, 6% (39 of 664) reported shortness of breath, and 2% (15 of 660) reported loss of taste.

Tests were available to all individuals present at testing events, irrespective of their insurance status. Of those who responded to the testing experience survey, 31% (251 of 808) reported being uninsured. Among the 557 respondents who reported having insurance, 295 provided additional detail regarding their type of coverage. Of these, 93% (275 of 295) reported public insurance (Medicare, Medicaid, or Tricare), and 7% (20 of 295) reported private insurance.

Attendance at host events often ranged from 50 to 100 individuals; however, testing participation was voluntary. The majority of tests (80%; 1,146 of 1,428) occurred at a single partner site that hosted weekly testing events; this site averaged 17.1 tests per event (range=0–48) over 67 separate events, greatly exceeding the average of 5.0 tests per event (range=0–22) for sites that hosted 1-time or infrequent events.

SARS-CoV-2 PCR results were made available to participants either printed on site or sent through secure email in 2–9.6 hours (mean=3.1 hours). Over the entire study period, 48 positive individuals were detected (3%; 48 of 1,428), and each received follow-up information from a healthcare worker. Among participants who elected to wait on site for results, diagnostic results were typically available within the duration of the community event, allowing individuals to receive results before leaving the testing site.

Survey participation and subsequent completion among test takers varied, resulting in some questions having more respondents than others. The time to complete the survey averaged 7.7 minutes (range=1.1–39.8 minutes). Survey data showed that 74% (597 of 808) of test takers agreed that it was easy to access the van and get a COVID-19 test; only 5% (42 of 808) disagreed with this statement, and 21% (169 of 808) were neutral. The majority of test takers felt comfortable using the van for testing (75%, 608 of 808); only 5% (37 of 808) disagreed, and 20% (163 of 808) were neutral. Fewer participants responded when asked why they chose to test at the mobile site; of those who did, 58% (272 of 470) said that it was due to easy access; 30% (143 of 470) said that it was because they felt comfort and familiarity with the location and/or community representatives involved; 7% (33 of 470) were referred by either their family, friend, or physician; only 5% (21 of 470) noted the incentive of gift cards as motivating (Tables 1 and 2 and Appendix 5, available online). For 29% (167 of 571) of the participants, testing at the van was their first COVID-19 test. The questionnaire documented that 48% (255 of 529) were unaware of alternate COVID-19 testing opportunities, 44% (218 of 493) reported difficulty in accessing health care, and 49% (241 of 492) identified transportation as a challenge. Most participants lived alone (56%), 35% (160 of 458) lived in a nuclear family, and 9% lived in a multigeneration household or shared accommodation with other adults.

Table 1.

Responses to Survey Questions, by Demographic Group, Regarding Opinion of the Comfort and Ease of Testing Access (Connecticut, U.S.; July 2023 to July 2024)^a

Demographic	Overall, you were comfortable using the van to get a Covid test today	Overall, you felt it was easy to access the van and get a COVID test today.
Age range, years (n)
0–17 (50)	Strongly agree	Strongly agree
18–34 (90)	Strongly agree	Strongly agree
35–44 (111)	Strongly agree	Strongly agree
45–54 (77)	Strongly agree	Strongly agree
55–64 (90)	Strongly agree	Strongly agree
65–74 (45)	Strongly agree	Strongly agree
≥75 (8)	Somewhat agree	Somewhat agree
Gender (n)
Woman (218)	Strongly agree	Strongly agree
Man (285)	Strongly agree	Strongly agree
Nonbinary (7)	Strongly agree/neutralb	Strongly agree/neutralb
Prefer not to answer (26)	Strongly agree	Strongly agree
Race (n)
American Indian or Alaska Native (30)	Somewhat agree	Neutral
Black or African American (210)	Strongly agree	Strongly agree
Asian (14)	Strongly agree	Strongly agree
Native Hawaiian or Other Pacific Islander (1)	Strongly agree	Strongly agree
White (262)	Strongly agree	Strongly agree
Some other race (47)	Strongly agree	Strongly agree
Prefer not to answer (39)	Strongly agree	Strongly agree
Annual household income, 2019 before taxes (n)
$0–$14,999 (167)	Strongly agree	Strongly agree
$15,000–$19,999 (77)	Strongly agree	Strongly agree
$20,000–$24,999 (41)	Somewhat agree/strongly agreeb	Somewhat agree
$25,000–$34,999 (71)	Somewhat agree	Somewhat agree
$35,000–$49,999 (86)	Strongly agree	Strongly agree
≥$50,000 (44)	Strongly agree	Strongly agree
Prefer not to answer (88)	Strongly agree	Strongly agree

^aThe extent of survey completion varied among test takers, resulting in rows having different denominators.

^bEqual response (n).

Table 2.

Responses to Survey Questions, by Demographic Group, Regarding the Primary Reason for Selecting the Van for Testing (Connecticut, U.S.; July 2023 to July 2024)^a

Demographic	Why did you choose this location for Covid testing?
Age range, years (n)
0–17 (37)	Access
18–34 (32)	Comfort/familiarity
35–44 (84)	Access
45–54 (35)	Access
55–64 (33)	Access
65–74 (21)	Access
≥75 (6)	Access
Gender (n)
Woman (151)	Access
Man (198)	Access
Nonbinary (5)	Access
Prefer not to answer (11)	Access
Race (n)
American Indian or Alaska Native (9)	Access
Black or African American (152)	Access
Asian (13)	Referral
Native Hawaiian or Other Pacific Islander (1)	Access
White (197)	Access
Some other race (29)	Access
Prefer not to answer (18)	Access
Annual household income, 2019 before taxes (n)
$0–$14,999 (123)	Access
$15,000–$19,999 (54)	Access
$20,000–$24,999 (33)	Access
$25,000–$34,999 (57)	Comfort/familiarity
$35,000–$49,999 (51)	Comfort/familiarity
≥$50,000 (39)	Access
Prefer not to answer (53)	Access

^aThe extent of survey completion varied among test takers, resulting in rows having different statistical denominators.

DISCUSSION

This program evaluation found that a community-informed mobile diagnostic testing model can expand access to same-day COVID-19 testing in underserved communities. Over 1,400 individuals were tested across 123 community events, with a substantial proportion of participants reporting low income, a lack of insurance, and limited awareness of alternative testing options. Participant surveys indicated that the mobile testing service was widely perceived as accessible and comfortable to use, suggesting that mobile diagnostic programs deployed through trusted community organizations may help reduce structural barriers to accessing testing services.

Diagnostic testing has remained a cornerstone in managing the global COVID-19 pandemic since early 2020. However, vulnerable populations, who often have the highest need for testing, often bear the brunt of its inequitable management.^12,13 Barriers are not just confined to the number of testing centers but also to various factors at societal and individual levels, such as lack of access to testing or reliable healthcare information and limited financial resources.¹⁴ Testing should be accessible and ideally be intertwined with people’s everyday routines, livelihoods, and interpersonal relationships. To combat health inequity, leaders must address unmet needs and remove barriers to accessing health services, such as limited transportation, knowledge, or insurance.

Understanding of the challenges and opportunities of a community testing program can be achieved by collaborating with community leaders and organizations that have an established relationship with the neighborhood and are aware of the demands, deficits, and strengths of existing resources. As such, the study team worked in partnership with the community to curate testing implementation strategies and deploy a community-engaged program that addressed barriers to testing. All sites were open to everyone in the Greater New Haven area of Connecticut and were free of charge regardless of insurance coverage or referral. Offering readily available, no-cost testing in marketing materials and implementing on-site engagement practices to more directly share information about the availability of getting results in as little as 2 hours helped to reach the target population. Of those who participated, 91% had an annual household income <$50,000, and 31% of test seekers were uninsured. Of those who were insured, 93% identified public insurance (Medicare, Medicaid, Tricare) as their primary healthcare plan. Both lack of insurance and uncertainty in reimbursement have frequently been mentioned as deterrents to seeking care.^15,16 This barrier may be why, even after testing has been available since 2020, 29% of participants had their first-ever COVID-19 test at the van, which was in operation from June 2023 to July 2024.

Barriers to testing also included logistical challenges such as transportation and accessibility. In the post-testing survey, 44% of participants reported difficulty accessing health care in general, and 48% were unaware of alternative COVID-19 testing options in their community. Considering that this study took place at a time when free tests could be requested online, from the federal government through the mail, or picked up at local pharmacies, even if there was awareness of testing options, with 49% identifying transportation as a persistent challenge, going to pick up a free test still presented a barrier.

These findings highlight the importance of bringing diagnostic services directly into community settings where individuals already gather for essential services. Aligning testing events with existing community resources—such as food distribution programs—allowed the mobile testing model to reach populations that may otherwise face barriers to accessing clinic-based testing. Among participants who waited on site for results, diagnostic reports were typically available within the duration of the community event, enabling individuals to receive their results before leaving the site.

This highlights the opportunity for mobile testing programs to meet community members at locations convenient to them and to offer timely care where vulnerable populations are likely to be rather than setting up testing at sites that lack a connection to existing services or aligned organizations. For this study, aligning events with existing resources or public health services and deploying local testing events in higher-density areas increased the accessibility of the same-day service. Among participants who chose to wait on site, results were typically delivered within a few hours, allowing individuals to receive their diagnostic status before leaving the event. Community partners reported that this rapid turnaround was particularly valuable for individuals with limited access to the internet, phone service, or mailing addresses. Testing was most successful at repeat, demographic-aligned locations; 58 testing events were held to coincide with a weekly food distribution event, where a notable portion of the 50–100 attendees at the event each week elected to test at the van. This level of uptake accounted for 80% of the tests administered throughout the study and >3 times the number of tests per event compared with other sites. These findings suggest that bringing diagnostic services directly into community settings may help reduce the structural barriers that limit participation in traditional clinic-based testing programs.

Participants and community partners also noted challenges in understanding how to access testing programs or interpret symptoms, such as those of COVID-19 versus allergies. In addition, barriers frequently cited by partners—including limited internet access, technological literacy, and stigma associated with testing positive—highlight the importance of designing public health communications that account for structural barriers and institutional distrust in underserved communities.^1,17 These findings reinforce prior research suggesting that awareness of testing systems and culturally appropriate messaging play important roles in testing uptake.¹⁶

In contrast, this model of offering same-day results from saliva-based testing at local community events was widely recognized by test takers as convenient, accessible, and easy to use, demonstrating that this type of mobile solution can effectively offer diagnostic testing in a manner that is well received in traditionally underserved communities. Furthermore, offering on-site testing with same-day results made a key difference for community members who lacked reliable access to the internet, phone service, or a mailing address or who would have required language support to understand their results. Without on-site testing, many participants in this community might have continued to face barriers to accessing COVID-19 testing. These findings suggest that similar on-site testing programs may represent a practical strategy for reaching vulnerable populations who face barriers to accessing health services.

Beyond survey findings, the implementation of the mobile program demonstrated the importance of partnerships with established community organizations. Leveraging existing relationships helped to identify appropriate testing locations, align testing events with ongoing community services, and build trust among potential participants. Maintaining clear communication and minimizing logistical demands on partners—such as providing promotional materials and ensuring that the mobile team arrived fully equipped—supported sustained collaboration throughout the program.

Ongoing engagement with partners expanded the network of participating organizations and informed more responsive testing deployment. Although the average number of tests per event was modest, participation was highest at recurring community sites that were aligned with existing services, suggesting that a consistent presence and trusted community partnerships may support sustained engagement. For this program, organizations and individuals with existing connections to the target community were asked for input on the program design and to recommend additional stakeholders, who were then invited to contribute and suggest further connections. Broadening the network in this way meant that the community cohort was far more diverse over time, leading to better-informed engagement plans and a study design that was more representative of the target community’s needs. For example, feedback from participants cited referrals as being key to their participation—knowing someone who had tested at the site before led to their trusting the service provided (Appendix 5, available online). Maintaining a routine testing schedule and offering small incentives for completing the accompanying survey were also important implementation lessons. Although few participants reported that incentives were their primary motivation for testing, the $10 gift cards helped compensate individuals for the time required to complete the survey and were associated with increased survey participation. Recurring events at consistent locations also improved operational efficiency and facilitated community referrals over time.

Continuous engagement with partners allowed the program to adapt to evolving community priorities. For example, when hosts reported declining interest in SARS-CoV-2 testing and concerns about survey completion time, the program introduced modest incentives to compensate participants for their time, which increased study enrollment. Partners also suggested that in settings where immediate results are less critical, a hybrid model—collecting samples on site while processing them in a nearby laboratory—could expand the program’s reach. This approach is currently under development. Feedback gathered throughout the study period reinforced the importance of maintaining an open, ongoing dialog with community organizations. Regular check-ins with partners—through emails, virtual meetings, and phone calls—helped to identify new testing locations, refine outreach materials, and respond to on-the-ground observations from testing events. For instance, on-site staff noted that without the gift card incentives, participants at multiple events would not have completed the accompanying survey, underscoring the practical role these incentives played in sustaining data collection rather than driving testing participation itself. Community partners also offered specific recommendations for future program development, including expanding the mobile model to additional diagnostic services, such as respiratory pathogen panels and sexually transmitted infection testing, where rapid on-site results could support timely clinical decision making. These recommendations, combined with participant feedback indicating high comfort and satisfaction with the mobile testing format, suggest that community-informed mobile diagnostic programs have the potential to serve as adaptable platforms for addressing a range of public health needs beyond COVID-19. An implementation toolkit informed by the learnings from this program has been developed to support the replication and adaptation of mobile community testing programs in other settings.¹⁸

Community testing programs, such as the mobile model described in this study, may help bridge gaps in access to diagnostic services by delivering testing directly to underserved populations. In addition to supporting individual diagnoses, these programs may contribute to improved surveillance and earlier identification of infectious disease cases. Successful programs should reflect the needs of the target populations, mitigate barriers to access, and incorporate ongoing input from community stakeholders. Future public health responses may benefit from integrating mobile diagnostic programs with trusted community organizations to expand their reach and improve equity in access to testing services.

Limitations

Limitations of this study include the population surveyed (e.g., sample size, demographics, recruitment level, test site locations, and others were limited to the region in which this study took place, the reach of community connections, and the availability of on-site team members), the potential influence of incentives on biasing participation and perception of testing experience, the response rate to the survey itself (participants did not arrive at events solely to partake in a research project, so they could not be expected to always complete the full survey), and reliance on nonstudy team members to translate for participants who did not speak English or Spanish. Because this evaluation was observational and lacked a comparison group, the study cannot determine the causal effects of the intervention on testing behavior.

CONCLUSIONS

This program evaluation demonstrates that community-informed mobile diagnostic testing is a feasible and acceptable strategy for expanding access to health services among underserved populations. The lab-in-a-van model enabled same-day SARS-CoV-2 testing at trusted community sites and reached populations with substantial barriers to healthcare access, including individuals who were uninsured, low income, or previously untested. Participant surveys indicated that the service was widely perceived as accessible and comfortable to use. These findings suggest that mobile diagnostic programs, particularly those developed through partnerships with community organizations, can help address structural barriers to healthcare access. Similar models may support future public health responses by bringing diagnostic services directly to communities with limited access to traditional healthcare infrastructure.

CRediT authorship contribution statement

Brittany L. Choate: Conceptualization, Data curation, Formal analysis, Funding acquisition, Methodology, Project administration, Supervision, Visualization, Writing – original draft. Ruhani Sardana: Formal analysis, Writing – original draft. Acsah Mathews: Data curation, Formal analysis, Investigation, Writing – review & editing. Stephanie Weirsman: Conceptualization, Data curation, Funding acquisition, Supervision, Writing – review & editing. Katherine Fajardo: Conceptualization, Writing – review & editing. Yasmine Ali: Supervision, Writing – review & editing. Chen Liu: Supervision, Writing – review & editing. Pei Hui: Supervision, Writing – review & editing. Kevin Schofield: Data curation, Writing – review & editing. Anne L. Wyllie: Conceptualization, Formal analysis, Funding acquisition, Methodology, Resources, Validation, Supervision, Writing – review & editing. Angelique W. Levi: Conceptualization, Formal analysis, Funding acquisition, Supervision, Writing – review & editing.

Appendix. Supplementary materials

Video 1. Mobile laboratory van tour.

Note: A cargo van was outfitted to provide high-complexity clinical diagnostic testing, including compact, multipathogen PCR systems (Myra automated liquid handling system and Mic real-time PCR cycler, Bio Molecular Systems), effectively creating a satellite laboratory capability for rapid on-site testing. The mobile unit has appropriate specimen storage and handling facilities, biosafety cabinets, and molecular biology machinery as well as all necessary software, including REDCap survey technology, to ensure accurate test result delivery and streamlined data sharing. The mobile unit was operated as a CLIA-certified site under Yale University School of Medicine's Department of Pathology. File: SD mobile testing van_tour.mov— https://drive.google.com/file/d/1UB1HigODtQg-aGA4mC32htBfYN8Z7MYO/view?usp=drive_link. CLIA, Clinical Laboratory Improvement Amendments.