Abstract

Background

Blood and oral fluid-based HIV self-tests are important for reaching the undiagnosed living with HIV. The study objectives were to evaluate the oral fluid-based 翱谤补蚕耻颈肠办庐 HIV Self-Test (HIV-ST) performance in comparison to laboratory reference testing; determine if laypersons can correctly perform the HIV-ST; document if intended users can successfully interpret pre-made contrived positive, negative, and invalid results; and document if intended users can understand the key messages in the product labeling.

Methods

This prospective study enrolled consenting adult intended users of HIV self-testing from six community health centres in four Canadian provinces between June 2022 and January 2024. Positive and negative agreement was determined by comparing the results of the HIV self-tests with the results of the laboratory-based 鈥済old standard鈥� Abbott Alinity HIV Antigen/Antibody Combo test. Descriptive statistics were used to summarize usability self-test procedure steps.

Results

Overall, 951 participants were recruited and consented with 911 available for all analyses. With respect to sociodemographics: 84% of participants were between 18鈥�45听years of age, 73% had at least a college education, 48% were Cis-male, 45% were employed; and 26% identified as White, 23% as African, Caribbean or Black, 5% as Indigenous [First Nations, M茅tis or Inuit], 33% as Asian, and 6% as LatinX. Primary efficacy analysis on the 911 who completed HIV-ST revealed a single confirmed positive participant and a negative percent agreement of 100% (880/880, 95% CI: 99.9鈥�100%) with the comparator method. For usability determination, the average success rate for 鈥渃ritical鈥� steps for completing the test was 94.1%. Approximately 97% of participants found the instructions easy to follow and 98% of participants reported they would use the test again. Of the 465 participants who interpreted the strong positive, weak positive, negative, and invalid pre-made contrived results, the average of correct interpretations ranged from 59鈥�97%

Conclusions

A licensed oral fluid-based HIV self-test in Canada can present an accurate, easy-to-use, and less invasive alternative to blood-based HIV testing. The addition of an oral-fluid self-test along with the current licensed blood-based HIV self-test could help reach the undiagnosed with HIV in Canada and positively impact HIV testing rates overall by offering individuals a choice of self-testing devices.

Background

Across Canada, HIV diagnosis rates continue to rise. There were 2,434 new HIV diagnoses in 2023 (35.2% increase since 2022) with the national rate of new HIV diagnoses rising to 6.1 per 100,000 population in 2023 compared to 4.7 per 100,000 population in 2022. The highest rates were reported in the prairie provinces of Saskatchewan and Manitoba [1]. Although Canada has endorsed the Undetectable鈥�=鈥塙ntransmittable (U鈥�=鈥塙) declaration and the UNAIDS 95鈥�95-95 targets (95% diagnosed, 95% of those on treatment and in care, and 95% of those achieving viral suppression by 2030), much still needs to be done to reach those targets [2]. From the most recent published data as of the end of 2022, an estimated 65,270 people were living with HIV in Canada with an estimated 89% diagnosed. Of those diagnosed, 85% were estimated to be on treatment with 95% of those having a suppressed viral load [3]. Collectively, as of the end of 2022, there were still nearly 16,000 people who are undiagnosed, or who are diagnosed but not yet on antiretroviral therapy in Canada and therefore not engaged in the continuum of HIV care. These gaps in diagnosis and connection to care and treatment are drivers of the epidemic in Canada and need to be addressed.

The acceptance and usability of oral fluid and blood-based HIV self-tests by key populations in global settings have been well established [4,5,6]. In Canada, recent novel testing approaches using finger stick blood-based HIV self-testing to reach the undiagnosed, such as I鈥檓 Ready and GetaKit, have shown high uptake, particularly amongst key and underserved populations such as gay and bisexual men who have sex with men, African, Caribbean, and Black populations, people who use substances, women, youth and Indigenous peoples. Self-testing programs can also provide information and 鈥渄o-it-yourself鈥� pathways for support, care, and treatment management, and have been shown to be broadly acceptable and effective in reaching the undiagnosed in Canada and linking them to care [7,8,9]. Furthermore, studies of HIV self-test implementation in Ontario, Canada found that programs providing low-barrier access to self-testing effectively reached key and underserved populations, identified persons living with undiagnosed HIV, and linked them to care and treatment [10,11,12]. A fundamental step towards ending the HIV epidemic by 2030 is reaching the undiagnosed which remains a critically important first step for people to know their status and get engaged and connected to treatment, care and prevention as appropriate [13, 14].

Following the World Health Organization (WHO) recommendation in 2016 for the use of HIV self-testing as an alternative to conventional facility-based testing [15], the WHO reported in 2024 on a wide array of countries adopting policies that include having both blood or oral fluid-based HIV self-testing as a vital part of differentiated HIV testing approaches [16]. This followed the UNITAID report in 2018 on the importance of the then-nascent HIV self-test development for use by key and underserved populations who may not choose to access facility-based testing alternatives [17].

In reviews published by Pai and colleagues on HIV self-test use, including oral fluid-based self-tests, in populations with high and low risk of acquiring HIV it was shown that HIV self-testing was broadly acceptable, and also that implementation of technology tools such as a mobile phone application can help increase testing uptake and facilitate linkages to care [18, 19]. In qualitative surveys of Canadian healthcare and community-based service providers, respondents reported that HIV self-testing is desired by Canadians and organizations were ready to support HIV self-testing. Yet, gaps remain in self-test accessibility and linkage-to-care pathways that need to be addressed in order for HIV self-testing programs to be fully established [20]听(Penny L, Lo Hog Tian J, McBain K, Galli R, Gahagan J, Worthington C, et al: HIV Self-testing and Mobile Health: A National Survey of Community-Based HIV Service Providers in Canada, submitted).

The first blood-based HIV self-test was approved for use in Canada by Health Canada on November 3, 2020 [21, 22], with uptake and scale-up of self-testing continuing across the country through targeted intervention programs [10,11,12, 23]. In a scoping review of global studies based on the actual use of blood and oral fluid-based HIV self-testing, covering a period from 2011鈥�2022, Adepoju et. al found a preference for oral fluid HIV self-testing over blood-based self-testing [24]. Since Canada currently has only a single, blood-based HIV self-test approved for use, the over-arching goal of this study was to provide performance outcome data on the 翱谤补蚕耻颈肠办庐 HIV Self-Test (OraSure Technologies Inc., Bethlehem, Pennsylvania, USA) that are necessary for regulatory approval of an oral fluid-based HIV self-test in Canada which would provide self-testers with a choice of self-test devices to best meet their preferences for testing.

The specific study objectives were to evaluate the device performance, i.e., positive and negative percent agreement with laboratory reference testing (4th generation HIV antigen/antibody test); document if intended users (i.e., laypersons) can successfully perform the procedural steps of the 翱谤补蚕耻颈肠办庐 HIV Self-Test without any training or demonstration of the test (accuracy/usability); document if intended users can correctly interpret pre-made contrived strong positive, weak positive, negative, and a range of invalid results in addition to their own results; and document if intended users can understand the key messages in the labelling (label comprehension).

The design and objectives of this study are aligned with the necessary Health Canada requirements for the evaluation of the performance and effectiveness of the 翱谤补蚕耻颈肠办庐 HIV Self-Test. The Health Canada Guidance for Manufacturers of Human Immunodeficiency Virus (HIV) Rapid Diagnostic Tests (RDTs) for use at the Point of Care (POC) or for Self-Testing provides performance targets that must be met for clinical sensitivity and specificity: 鈥淓vidence that the RDT intended for use at the POC or for self-testing has a minimum sensitivity and specificity of鈥夆墺鈥�99% for HIV antibody detection should be provided鈥� [25].

Methods

Study design

This prospective study enrolled consenting adults from six community health centres and medical clinics across Ontario, Qu茅bec, Manitoba, and British Columbia between June 2022 and January 2024.

Inclusion criteria

The study was open to the general public, either English or French-speaking, aged 18 or older who met the inclusion criteria for enrolment and did not meet any of the exclusion criteria listed in Appendix A. Although individuals under 18听years of age were not eligible for enrolment, no participant was excluded based on any other sociodemographic characteristic. A participant compensation of $50 CAD was provided for their initial study site enrollment visit and an additional $25 CAD was provided on their second visit when they returned to receive their laboratory testing results.

Study population and sample size

Recruitment of participants into the study occurred directly at the study sites during routine clinic visits using brochures, posters, and word of mouth. Participant鈥檚 self-reported HIV risk activity from the previous 12听months was collected at the time of study enrolment. The target sample size for the study was a minimum of 900 individuals with unknown HIV status including at least 400 individuals determined to be at risk of HIV infection, as specified by Health Canada for HIV self-test prospective studies. Because this study was designed to meet regulatory requirements for the number of participants based on the published guidelines [25], no sample size calculations were conducted.

Study protocol

Upon obtaining informed consent from each participant, one venous blood sample was collected for testing at respective provincial central laboratories by the Comparator Method (CM). Each participant then undertook the self-test process at the study site which required use of the test stick device to self-swab the gums for oral fluid collection onto an integrated absorbent pad. Participants then performed the 翱谤补蚕耻颈肠办庐 HIV Self-Test on the collected fluid and interpreted the results using only the manufacturer鈥檚 instructions for use as a guide. All steps in the self-test performance were directly observed by a study Observer who was a trained healthcare provider at the study site and who did not instruct, demonstrate the test, intervene or otherwise interact in any way with the participant conducting the 翱谤补蚕耻颈肠办庐 HIV Self-Test. Observations about the participant鈥檚 test experiences were noted throughout the self-testing process. The results of the rapid, lateral flow 翱谤补蚕耻颈肠办庐 HIV Self-Test performed and interpreted by the study participants were recorded as Non-Reactive, Reactive or Invalid and were subsequently compared with the results of a licensed 4th generation, chemiluminescent microparticle immunoassay (CMIA)-based HIV-1/HIV-2 combination antigen/antibody (Ag/Ab) test CM (Abbott Alinity HIV Ag/Ab Combo test, Abbott Laboratories, Abbott Park, Illinois, USA) [26] that is in routine use at the central public health laboratories of each of the participating provinces. Alinity HIV Ag/Ab Combo test results are reported as Non-Reactive (considered negative for HIV-1 p24 antigen and antibodies to HIV-1 and HIV-2) or Reactive (to undergo supplemental confirmatory testing). The Geenius鈩� HIV 1/2 Confirmatory Assay (Bio-Rad Laboratories, Marnes-la-Coquette, France), which is an immunochromatographic test for the detection and differentiation of individual antibodies to HIV-1 and HIV-2, was used for positive confirmation where necessary. Results are interpreted as Positive, Negative, or Indeterminate for HIV antibodies based on criteria indicated in the package insert [27].

For usability determinations, there were 18 items listed in the survey which were similar to those utilized in an evaluation of several blood and oral fluid-based HIV self-test devices in South Africa [28]. These survey items were designed to capture information on participant鈥檚 use of the test kit instructions, preparations for the self-test, self-collection of the oral fluid sample, completion of the self-test procedure, and related post-test procedures. Of these 18 items, 10 were considered 鈥渃ritical鈥� for correct usability (identified with * in Table听5). Immediately after the participant completed the self-test, they indicated their result interpretation to the Observer, who also interpreted the HIV self-test result and recorded the result independently. In addition, to comply with Health Canada requirements, a minimum of 400 volunteer participants were provided with blind-coded pre-made contrived result devices in random order and asked to interpret the test results.

The CM testing on serum from venous blood samples collected from study participants was conducted by the provincial public health laboratories and the results were provided back to the sites. Results were also uploaded to the central database at St. Michael鈥檚 Hospital, Toronto. The central laboratory test results were used as the reference result for the primary efficacy evaluation of the 翱谤补蚕耻颈肠办庐 HIV Self-Test. To allow for adequate turnaround time for laboratory results to be provided back to the clinic, study participants were advised to return to the clinic approximately two weeks after their initial visit to receive their laboratory testing results. All HIV-related clinical care and prevention services provided to each participant were based only on the results of the standard HIV testing that is in normal use at each site, and not on the self-test result.

Electronic data capture

Participant data were entered directly by the study Observers into tablets pre-loaded with study-specific surveys developed in Qualtrics (Qualtrics, Provo, UT, USA; ). This collected data was saved and stored on local secure and encrypted servers in Toronto at St. Michael鈥檚 Hospital for further data analysis. Only designated research team members were provided access to the password-protected data from the Qualtrics servers. To maintain privacy and confidentiality, participant survey data was linked only with their unique, anonymized participant ID. The full survey instruments that were developed for this study are provided in Appendix C.

Ethics approval

This study was approved by respective Research Ethics Boards (REB) for the Ontario, Qu茅bec, Manitoba, and British Columbia sites participating in the study: St. Michael鈥檚 Hospital REB, University of Toronto REB, Veritas IRB Qu茅bec and British Columbia, and University of Manitoba HREB. This study was conducted under a Health Canada-approved Investigational Testing Authorization (ITA), application ITA# 336,702, issued on 11 May 2022 and is listed in ClinicalTrials.gov ID# NCT05020353.

Data analysis

The study data analysis was conducted using IBM SPSS Statistics version 24. Sociodemographic variables and usability characteristics were summarized using descriptive statistics. The device performance (Primary Efficacy) was analyzed through the calculation of positive and negative percent agreement between study participants鈥� self-interpreted results of the 翱谤补蚕耻颈肠办庐 HIV Self-Test versus the 4th generation Abbott Alinity test results received from the provincial laboratories. For data analysis purposes, each study participant鈥檚 鈥渢rue鈥� HIV status was determined upon completion of the CM and confirmatory testing if necessary, and the participant鈥檚 interpretation of their self-test result then determined whether their self-test results were included in either the positive percent agreement (PPA) population or the negative percent agreement (NPA) population. All invalid self-test results, those with 鈥渋nconclusive鈥� laboratory results and responses indicated as 鈥渄o not know鈥� were excluded from the calculation of PPA and NPA.

For PPA and NPA calculations, the overall 95% confidence interval was determined where:

• PPA鈥�=鈥塠TP / (TP鈥�+鈥塅N)]鈥壝椻��100, where TP (true positive) is positive self-test in agreement with positive Alinity test, and FN (false negative) is negative self-test discordant with positive Alinity test; and,

• NPA鈥�=鈥塠TN / (TN鈥�+鈥塅P)]鈥壝椻��100, where TN (true negative) is negative self-test in agreement with negative Alinity test, and FP (false positive) is positive self-test discordant with negative Alinity test.

  In addition, the study participants鈥� self-interpreted results were compared with the Observers鈥� interpretation of the self-test results.

Results

Study participants

A total of 951 participants were enrolled over the study period. Of these, 40 participants were excluded from all analyses for reasons including failure to provide a venous blood sample or failure to provide a self-test result (see Appendix B for details on excluded participants). Figure听1 illustrates the CONSORT diagram for the number of participants who proceeded to enrolment, allocation, follow-up, and analysis within the study.

Fig.听1

Participant enrolment CONSORT flow diagram

Table 1 presents the sociodemographic profile of the 911 participants who were included in the primary efficacy analysis. Overall, the sample recruited across all sites had the following sociodemographic characteristics: 84% of the sample was between the ages of 18 and 45 years of age; 50% identified as 鈥淐is-women鈥�, 48% identified as 鈥淐is-men鈥� and 3% identified as other (11 identified as 鈥淣on-binary鈥�, 7 indicated Trans woman, 4 indicated 鈥淕enderqueer鈥� i.e. gender identity that is fluid, changing, 2 indicated Trans man and 2 who self-described gender). There were also 7 participants who indicated that they were 鈥淭wo Spirit鈥� (a person who embodies both a masculine and feminine spirit) and 5 participants who indicated that they were pregnant. Regarding ethno racial composition, of the 911 who provided survey responses, approximately 33% (n=307) identified as 鈥淎sian鈥� descent, 26% (n=235) identified as 鈥淲hite鈥�, 23% (n=211) as 鈥淎frican/Caribbean/Black鈥�, 5% (n=46) as 鈥淚ndigenous鈥�, 6% (n=54) as 鈥淟atinX鈥�, 3% (n=25) as 鈥淢iddle East/North African鈥� and the remaining 4% (n=33) as 鈥渙ther鈥�. Of the 908 participants who provided their educational achievement, approximately 46% (n=420) had a university degree or higher, 27% (n=242) completed College, 25% (n=226) completed High School, and 2% (n=20) completed primary school. Only 4% (n=39) of the 905 participants who responded reported having any kind of reading or writing impairment, and 33% (n=301) indicated visual impairment. Of the 904 participants who responded to the question on employment, 45% (n=405) were employed, 29% (n=261) indicated that they were students, 24% (n=215) were unemployed, and 3% (n=23) were retired.

Table 1 Participant demographics (n鈥�=鈥�911)

A total of 143 participants self-reported medical conditions, including hypertension (16%), diabetes (15%); and recent or existing sexually transmitted diseases (9%); approximately 60% disclosed 鈥渙ther鈥� medical conditions (e.g., mental illness, pre-diabetes, Hepatitis C, asthma, and cancer were the most common). Regarding the 911 who self-reported HIV status, 47% (n=426) indicated that their status was unknown or had never been tested, while 53% (n=485) were negative indicating prior experience with HIV testing.

When evaluating the self-reported HIV risk category, 57% of the 911 participants (n鈥�=鈥�520) reported at least one behaviour characteristic that would be considered 鈥渉igh risk鈥�. In terms of specific self-reported high-risk characteristics: 31% identified as having condomless sex with men, while 21% had unprotected sex with women. Approximately 19% indicated having multiple sexual partners, 44 participants (5%) indicated having used injection drugs, 17 participants (2%) self-identified as sex workers, and 9 individuals (1%) indicated that their sexual partner was HIV positive.

Primary efficacy (Performance) analysis

The primary efficacy analysis on the 911 who completed the 翱谤补蚕耻颈肠办庐 HIV Self-Test showed a positive percent agreement of 100% (1/1) and a negative percent agreement of 100% (880/880, 95% CI: 99.9鈥�100%) when comparing participant鈥檚 self-test results to the Abbott Alinity (see Table听2 for a full comparison of participant and laboratory test results and Table听3 for PPA and NPA calculations). The study encountered one previously undiagnosed participant living with HIV that was confirmed by the Abbott Alinity and Geenius鈩� comparator methods. The overall percent agreement between participant and observer for the self-test results was 97.6% (884/906) (see Table听4 for a full comparison of participant and observer interpretations of HIV-ST results). The overall 鈥渋nvalid鈥� rate for the HIV Self-Test was 2.2% (n鈥�=鈥�20) and an additional 1.0% (n鈥�=鈥�9) of participants indicated they were uncertain (鈥渄o not know鈥� or 鈥渦nsure鈥�) as to how to interpret the results.

Table 2 Self-Test Interpretation by the Participant Compared with Lab Result

Table 3 Positive and Negative Percent Agreement with Laboratory Results based on 911 completed HIV Self-Tests

Table 4 Participant and observer interpretations

Usability assessment

A total of 911 participants were included in the usability assessment surveys conducted by the study Observers. Across the 18 usability assessment questions in the survey, the average for all expected outcomes was 94.1%. For the 10 items identified as critical, the average of expected outcomes was also 94.1% (see Table听5 for a full breakdown of the Observers鈥� usability assessment). 83.2% of participants were able to correctly collect the oral-fluid sample, 5% were observed to incorrectly touch the absorbent pad during the collection procedure and 98% were able to correctly place the test device into the developer vial after sample collection.

Table 5 Observer usability assessment

Following the oral fluid sample collection, there were two categories of grouped procedure steps carried out as part of the test process: (1) performing the test procedure steps in order (88% did so correctly), and (2) waiting at least 20听min before interpreting the results (91% waited the correct time to interpret their results). Approximately 5% (n鈥�=鈥�46) missed one or more steps but continued the self-test process despite this, and only one participant withdrew before the test was completed.

Participants鈥� self-test usability survey results, shown in Table 6, were compiled from participant responses. When asked about the instructions for use, 97% of participants (n=878) found them easy to follow. Regarding use of the device, 100% of responding participants indicated that they were confident with performing the test on their own, 99% (n=895) found the device easy to use, 98% of participants (n=882) indicated that they would use the test again and 99% (n=890) would recommend this test to a sexual partner or someone in their social network. In terms of preference in self-testing locations, 77% (n=697) indicated that they would prefer to do the test at home, while 23% (n=205) indicated that they would prefer to do the self-test at a clinic.

Table 6 Participant usability assessment

Pre-made contrived test results interpretation

A total of 465 participants agreed to complete the pre-made contrived test result interpretation exercise. The six study sites were provided with 2 panels each, with each panel containing 5 blind-coded 翱谤补蚕耻颈肠办庐 cartridges with the following pre-made contrived results:

• 1. Strong positive (control line with highly visible test line)

• 2. Weak positive (control line with weak but visible test line)

• 3. Negative (control line but no test line)

• 4. Invalid (no control line but a highly visible test line)

• 5. Invalid (no control line and no test line)

Participants were presented with the pre-made contrived results cartridges in singular random order and were asked to interpret the results for each. As shown in Table听7, the rate of correct interpretations varied across the different types of pre-made contrived results cartridges. For 3 of the 5 result types (strong positive, negative, and invalid with no control line and no test line), the rate of correct interpretations was鈥夆墺鈥�90%. For the 鈥渨eak positive,鈥� the correct interpretation mean was 70% (range 58鈥�75%) which indicates that the majority of participants were able to see the line despite the weak intensity. The lowest rate of correct interpretations was obtained for the 鈥渋nvalid with a visible test line but no control line,鈥� with a mean of 59% (range 49鈥�80%), indicating that participants had some difficulties interpreting this type of invalid result. In follow-up analyses, most incorrect interpretations of the weak positive were recorded as negative, and for the invalid with no control line but a strongly visible test line most incorrect interpretations were reported as positive. There were no participant sociodemographic characteristics that were broadly associated with incorrect interpretations of the pre-made contrived results.

Table 7 Total pre-made contrived device result interpretations

Discussion

This study was successful in demonstrating the excellent performance of the 翱谤补蚕耻颈肠办庐 HIV self-test overall in meeting the expected results for the study objectives. The study outcome for negative percent agreement was comparable to specificity data from large-scale studies conducted in the US. The US data, summarized in the FDA Summary of Safety and Effectiveness also showed acceptable sensitivity [29], which led to FDA approval of the 翱谤补蚕耻颈肠办庐 HIV Self-Test. Despite only encountering a single confirmed HIV positive in this study, the positive percent agreement and negative percent agreement with 4th generation HIV reference tests were 100%. The single positive was a previously undiagnosed individual who was living with HIV and was subsequently linked to care and treatment.

In terms of usability, there was a 97.6% concordance in the self-test results interpretation between the self-test study participants and the trained observers indicating that intended users of the 翱谤补蚕耻颈肠办庐 HIV Self-Test can interpret their self-test results correctly compared to health care provider interpretations. Most discordant results between observer and participant were in the uncertainty (鈥渄o not know/not sure鈥�) response category although the overall number of 鈥渄o not know/not sure鈥� results reported by participants was low (n鈥�=鈥�9). A total of 20 participants (2.2% of the sample) interpreted their self-test results as invalid. Of these, 17 were also interpreted as invalid by the Observers. The rate of invalid results obtained by intended users of a self-test device is often an indication of ease of use and if some users may be experiencing difficulties in proper sample collection and/or performing the test procedures correctly. Previous Canadian studies of a blood-based HIV self-test had shown that rates of invalid results with self-testing by intended users could vary according to user experience with HIV testing and other contributing sociodemographic factors, with invalid rates of up to 20% reported [21, 30]. The low rate of invalid results observed in this study indicates that participants had few difficulties with performing the test procedures correctly on their own or in seeing a visible control line.

The pre-made contrived results interpretations for the 465 participants who volunteered for this portion of the study showed the majority were concordant with the expected results. There was high percent agreement, as expected, with 3 of the 5 pre-made contrived results: for the 鈥渟trong positive,鈥� 鈥渘egative鈥� and the 鈥渋nvalid, with no control or test line present.鈥� However, results indicated that the 鈥渨eak positive鈥� and 鈥渋nvalid with a visible test line鈥� can be somewhat more challenging to interpret correctly for some individuals. For the weak positive, it is to be expected that a test line with low intensity may not be consistently visualized. This outcome (70% correct interpretations) is similar to that reported by Majam et al. in their study of the performance of seven HIV-ST, including pre-made contrived results, in South Africa [28]. This does indicate a potential risk for a false negative interpretation based on visual acuity for discerning a test line of weak intensity, which is a limitation of all diagnostic devices that rely entirely on visual interpretation. For the invalid with no control line but a strongly visible test line (59% correct interpretations), this is unlikely a concern in real-life testing situations as this would rarely occur and would usually be an indication of a manufacturing flaw in a particular device, which had escaped the rigid quality control measures in place during manufacture. Although demographics such as visual impairment that was reported by 33% of participants in this study were not associated with higher rates of incorrect interpretations of pre-made contrived results, some mitigations may be considered to improve the instructions for interpreting the range of self-test results that users may experience, particularly with more challenging results such as a weak positive or invalid result with no control line but a visible test line. It would be important in the instructions to emphasize that the absence of a control line indicates an invalid result regardless of whether a test line is visible or not. An added instruction to conduct self-testing in a well-lit area would be beneficial for the user as well. Additional suggested modifications could include slight revisions to the package insert to provide clearer instructions and/or visual illustrations on the proper interpretation, as well as alerting the self-tester that even a faint line intensity is considered valid.

In general, participants in this study presented a broad diversity of sociodemographics that would be representative of the intended users of HIV self-tests in Canada. Participants expressed a high level of satisfaction with the HIV-ST experience as the vast majority of participants (97% to 99%) found the instructions for use easy to follow, found the self-test device easy to use, would use the 翱谤补蚕耻颈肠办庐 HIV Self-Test again, and would recommend its use to a sexual partner or friend which aligns with other studies among a variety of populations in different global settings [10, 11, 14] and can lead to increases in HIV testing rates overall [6, 31, 32]. Moreover, results from a large-scale oral fluid self-test implementation program in the US indicated that a high percentage of persons in the program had never previously received testing for HIV or had not received testing within the previous year, especially in younger populations under 25听years of age [33].听In this current Canadian study, 47% of participants indicated that they were unaware of their HIV status or had never tested before, despite having access to free, licensed blood-based HIV self-tests. Oral fluid remains the most common sample substrate for HIV self-testing globally and oral fluid HIV self-tests are broadly utilized in various distribution strategies [32, 34, 35].

Overall, 17% of study participants were observed to have had some difficulties in the correct use of the swab for oral fluid collection, indicating that additional instructions might be necessary in some situations. Of interest, approximately 77% of participants indicated a preference for using the self-test in a home environment while 23% would prefer to use the test at a clinic or health care facility, suggesting that both assisted (supervised) and unassisted self-testing strategies could potentially be implemented to maximize 鈥渞eal world鈥� access and uptake. Collectively, these outcomes indicate that this diverse, representative study population, with no prior experience with this oral fluid-based self-test, had a high degree of comfort and satisfaction with their self-testing experience.

Studies have shown that preferences for oral fluid or blood-based HIV self-testing devices vary depending on user perceptions of accuracy, sample type integrity, and ease of collection of the sample type [36,37,38,39,40]. However, self-test uptake overall could be enhanced by having both types of HIV self-tests licensed for use in Canada and offering users a choice of methods through the implementation of sustainable self-testing programs.

Despite the overall favourable outcomes from this study, there were some limitations: the study encountered only a single HIV-positive participant in the study population, including the 57% (n鈥�=鈥�520) who reported at least one behaviour characteristic that would be considered 鈥渉igh risk鈥�, therefore meaningful determination of the device sensitivity for accurate detection of HIV antibodies in the hands of intended users could not be measured. Survey responses from participants were voluntary and response rates varied across the individual questions, yielding variable denominators for determining outcomes for study participant sociodemographics and usability analyses. Lastly, a total of 73% of participants indicated they had a college or university education, which may be a potential bias and not truly representative of a broader and more educationally diverse intended user population for the self-test.

Conclusions

We found the oral fluid-based 翱谤补蚕耻颈肠办庐 HIV Self-Test to be highly accurate and usable by the diverse study participants. Since Canada has only a single licensed HIV self-test which is blood-based, having access to both blood-based and oral fluid-based HIV self-testing has the potential to substantially increase HIV testing rates. An approved, accessible oral fluid HIV self-test in Canada can offer an accurate, non-invasive, and easy-to-use alternative to facility-based HIV testing and to blood-based HIV self-testing for key/priority populations.