Feasibility and acceptability of COVID-19 self-testing offered through community screening posts in Mozambique: a mixed-methods study

Abstract

Background

COVID-19 self-testing is a key strategy for point-of-care testing and COVID-19 outbreak management to identify people infected with SARS-CoV-2 in low- and middle-income settings. We assessed the feasibility and acceptability of healthcare worker-supervised COVID-19 self-tests at community bi-directional screening posts for COVID-19 and tuberculosis in urban and peri-urban areas in southern Mozambique.

Methods

Between May and August 2023, Persons presenting with symptoms suggestive of COVID-19 and those at risk of COVID-19 infection were eligible for self-testing. We collected and described a cascade of care and survey data from people who completed self-testing and conducted logistic regression to assess predictors of the likelihood of using the self-test again (likely/not likely). We invited a subset of clients and supervising healthcare workers to semi-structured in-depth interviews, and we analysed their transcripts using a thematic approach.

Results

We invited 382 people for self-testing, of whom 369 (96.6%) clients performed the self-test. 328 (88.9%) were willing to use the self-test again when experiencing COVID-19 symptoms, and 332 (90%) declared to accept its use as a screening measure for COVID-19 contacts. Regarding its affordability, clients perceived self-testing as an acceptable strategy following its accessibility, comfort of use, and quick results. The feasibility of self-testing requires instructional videos in local languages, and implementation should be done through community channels. All supervising healthcare workers perceived self-testing as an important tool for increasing testing accessibility, early diagnosis, reduction of overcrowding and COVID-19 infection risk in healthcare facilities.

Conclusions

COVID-19 self-testing is acceptable and feasible if provided with linguistically and culturally appropriate health education and instructions at an affordable price. Self-test delivery packages and systems should be developed that could be adapted to existing or future self-tests for infectious diseases or other non-communicable diseases.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-026-26889-1.

Background

COVID-19 testing facilitates early identification and isolation of persons with COVID-19, which prevents further transmission of SARS-CoV-2. From 20th of March 2020 to March 2024, Mozambique reported 233,731 people positive for COVID-19, with 2,250 deaths [1]. COVID-19 vaccination coverage in Mozambique was reported at 98.9% in March 2023 [2]. Although COVID-19 is currently not considered highly incident, with Mozambique reporting an average of less than 150 new cases per day in March 2023, measuring preparedness for a new influx of cases is warranted [3].

During the COVID-19 pandemic in low and middle-income countries, rapid identification of persons with SARS-COV-2 infection was hampered by barriers related to reverse transcription-polymerase chain reaction testing (RT-PCR) for COVID-19. RT-PCR testing required skilled laboratory technicians and the use of molecular biology laboratories, which in Mozambique were only available in provincial capitals [4]. COVID-19 rapid antigen testing is an effective and important intervention for rapid identification and subsequent isolation of persons with SARS-COV-2 infection [5, 6].

To reduce the overload at health facilities in the provision of COVID-19 diagnostic services, the Ministry of Health (MoH) launched the “Use of SARS-CoV-2 Antigen-detection rapid diagnostic tests (Ag-RDT) Guidance” in October 2020⁷. This guidance was aimed at increasing access to early COVID-19 diagnosis using Ag-RDTs administered in health facilities by healthcare workers [4]. Additionally, the National Institute of Health created diagnostic points at hotspots in Maputo City to increase accessibility to COVID-19 testing. Furthermore, Fundação AURUM implemented bi-directional testing of COVID-19 and tuberculosis in Community Posts, increasing testing accessibility to additional districts of southern Mozambique. This need was already highlighted in previous publications (Manhiça 2022), as it was found that COVID-19-related emergency measures potentially had a negative impact on tuberculosis notification [8].

In previous values and preferences studies, self-testing for COVID-19 was perceived as an acceptable tool for early detection of COVID-19 in multiple low-middle income countries (LMICs) [9–11], and its use was supported by the Africa Centres for Disease Control and Prevention [12]. One study showed that the general population of nine LMICs reported the likelihood of using the self-test when experiencing symptoms [13]. Similarly, healthcare workers and decision-makers supported using self-testing as an approach that they felt could assist with early case detection and improve access to testing [10]. However, as of March 2024, Mozambique has not implemented COVID-19 self-testing, and testing continues to be conducted only by professionals in health facilities or community posts. To ensure pandemic preparedness and prevent health services from collapsing when testing demand increases in a new surge of cases, it is important to study the feasibility of self-testing and identify facilitators and barriers to the implementation of this testing strategy. To inform future decision-making on the implementation and scale-up of COVID-19 self-testing, we assessed the acceptability and feasibility of self-testing strategies for COVID-19 in Mozambique.

Methods

Study design and setting

From May to August 2023, we conducted a mixed-method study with an explanatory sequential design (quantitative data collection followed by qualitative data collection) to assess the feasibility and acceptability of COVID-19 self-testing among individuals symptomatic for COVID-19 and those asymptomatic but at substantial risk of COVID-19 infection. The quantitative component of this study consisted of two surveys assessing the cascade of care for self-testing, the perception and experience with COVID-19 self-testing among clients who self-administered a COVID-19 self-test, and among healthcare workers (HCWs) and Community Health Workers (CHWs) who observed the COVID-19 self-administration. The qualitative component included semi-structured individual interviews, which provided an in-depth understanding of the acceptability and feasibility of COVID-19 self-testing among the same clients, HCWs and CHWs who participated in the quantitative survey.

We used a non-probabilistic sampling for convenience to select seven (7) community health posts located in community markets and open-air bus terminals in the urban and peri-urban areas of Maputo City, Maputo Province, Xai-Xai and Chókwè, in southern Mozambique to conduct the study. The survey instruments and interview guides were piloted in some community posts to ensure that the questions were understandable and culturally appropriate. The quantitative pilot included 20 clients and 5 healthcare workers, while the qualitative pilot involved in-depth interviews with 4 clients and 5 healthcare workers. The data collected during the pilot was included in the analysis.

Study procedures

All individuals aged 18 years and older seen in the community posts and presenting either with symptoms suggestive of COVID-19 or asymptomatic presentations but at elevated risk of COVID-19 were eligible for study participation. We excluded individuals who had tested positive for COVID-19 seven days before the self-test and were still within their isolation period, as well as those with medical conditions requiring prompt attention at a health facility.

Research staff approached eligible clients who were waiting in line for a COVID-19 test at a community post, provided them with study information and invited them to participate in the study (i.e. conducting a COVID-19 self-test and a subsequent survey). Participants were recruited using a sequential random sampling approach; the sample size was determined assuming conservatively that 50% of individuals offered a self-test would agree to self-administer an AgRDT. Under this assumption, 196 respondents were required to achieve a 95% confidence interval with a ± 7% margin of error. To capture perceptions across rural, urban, and peri-urban settings, a minimum of 294 participants was required; therefore, a total of 338 participants were included in the survey to account for variability across settings and 10% potential non-response. Reasons for non-participation were inquired about and reported by trained data collectors. Potential participants were provided with information on the purpose and objectives of the study, the survey, and the interview, including the types of questions, expected duration, and the request for an audio recording. After allowing sufficient time to consider participation, individuals were asked to provide written informed consent in person before the start of the survey and/or interview. All consent forms were labelled with a unique participant code, which was also recorded on the survey, interview notes, and transcripts. For illiterate participants, the information sheet was read aloud in the presence of an impartial witness, who signed the consent form to confirm whether the participant agreed or declined to take part in the study. After agreeing to the procedures, clients received a COVID-19 self-test kit, including written instructions and pictures for COVID-19 self-testing performance (see Additional file 1). For COVID-19 self-testing, we offered clients the World Health Organization and FIND-recommended and Food and Drug Administration-approved ACON Flowflex COVID-19 Antigen Home Test, manufactured by the Chinese company ACON [6, 14, 15]. Thirteen healthcare providers were present, consisting of one medical technician and 12 community healthcare workers, actively observing clients with their COVID-19 self-testing. Observers were instructed not to interfere with the self-testing process, even if a client failed to use the kit correctly, unless intervention was necessary to prevent harm to the client. Individuals who received a positive result in the self-test were promptly referred to a nearby health facility, ensuring appropriate follow-up and care. An additional file provides a detailed overview of study procedures (Additional file 2).

Data collection

Quantitative data collection and analysis

At the screening event, trained on-site data collectors collected an anonymous cascade of care data from clients [Additional file 3]. Upon completion of the COVID-19 self-test, on-site data collectors surveyed clients on their experience and ease of use (Additional file 3). Additionally, healthcare workers (HCWs) and community health workers (CHWs) who were involved in observing clients performing self-tests completed a survey on their experiences and observations after they had finished all their observations (Additional file 3).

The survey questionnaires were designed in English, translated into Portuguese and entered into a Research Electronic Data Capture (REDCap) [16, 17] mobile app. For general respondents, data collectors conducted the surveys in Portuguese or Changana, according to the client’s language preference. For HCWs respondents, the survey was conducted in Portuguese due to Portuguese being the official language and preferred among HCWs and approved by the local IRB on the data collection tools. To ensure data quality and consistency, data collectors received standardized training, the questionnaires were piloted before data collection, and routine supervision was conducted in the field. In addition, completed surveys were reviewed daily for completeness and accuracy. A back translation of the questionnaire to English was done for this manuscript and report purposes.

We used descriptive statistics to present the cascade of care and survey data, including clients’ characteristics, risk perceptions of COVID-19, experiences with self-testing, and envisioned future use and applicability of self-tests. Univariate descriptive analysis was performed in the statistical software package R Version 4.32 [18].

Qualitative data collection and analysis

We reported qualitative data collection and analysis in line with COREQ guidelines (see Additional file 4).

Authors IS, SI, and EMM developed two interview topic guides (See Additional file 5 A and 5B) in English and translated them into Portuguese to guide the semi-structured interviews with the two study populations. The interview topic guides included at least the following themes: knowledge and perceptions of COVID-19 in general, COVID-19 diagnosis; experience with COVID-19 self-testing; linkage to care following a positive self-test; and future implementation of COVID-19 self-testing. Senior qualitative researcher (authors IS) provided training of trainers (authors EMM, SI) trained interviewers (authors PGM, CCB, CB, AN (male: 2 and female: 2) and data collectors with previous experience in in-depth interviews on qualitative interview skills and topic guides.

Purposive sampling of respondents followed two different strategies for clients and HCWs/CHWs. For clients, trained (male and female) data collectors (Authors: CCB and AN) and trained data collectors randomly approached clients who completed COVID-19 self-testing face-to-face: after self-introduction, they explained the purpose and objective of the semi-structured interviews and invited them to participate as Key Informants (KIs). Due to the limited number of CHWs and HCWs observing the clients administering self-tests, authors PGM, CCB, and AN approached them all face-to-face to introduce themselves, explain the study and interview purpose and invite them to participate in the semi-structured interviews. All respondents provided written informed consent before the interview. For four participants, consent was obtained through an impartial witness in accordance with ethical requirements.

All interviews took place at a private and quiet location near the screening event, convenient for the interviewee. No reimbursement was given to the participants. Interviewers audio-recorded all interviews, which were subsequently transcribed verbatim from Changana or Portuguese and translated into Portuguese and English, respectively, by a local experienced person. After each interview, the interviewer wrote a report including field notes. Regular debriefing sessions were held with the research team to discuss emerging issues and ensure consistency in interviewing techniques. Transcripts were reviewed against audio recordings for accuracy, and translations were cross-checked to maintain fidelity to the original meaning.

All 30 invited clients, 13 invited CHWs and HCWs, 30 clients and 13 CHWs/HCWs participated in the semi-structured interviews. Authors EMM and SI evaluated after each five client interviews if data saturation had occurred, which was reached at 30 interviews. Data saturation among CHWs/HCWs interviews was not the aim due to the limited number of eligible. No repeated interviews were carried out. The interviews had an average duration of approximately 30 min. Table 1 describes the characteristics of the two populations interviewed.

Table 1.

Characteristics of respondents (Qualitative)

Number of interviews	Clients		Healthcare workers
	30		13
Age range
-Average	39	Years	34	Years
- Min.	20	Years	25	Years
- Max.	72	Years	55	Years
Sex
- Female	14	(47%)	10	(77%)
- Male	16	(53%)	3	(23%)
Level of education
- No Formal Education	4	(13%)	0	(0.0%)
- Primary School	11	(36%)	0	(0.0%)
- Secondary High School and Technical or Vocational School	14	(40%)	10	(61.5%)
- Higher Education	1	(3%)	3	(23.3%)

We used a thematic approach for qualitative data analysis, which included both deductive and inductive coding. After familiarization with transcripts, authors SN, EMM, and BT developed a primary codebook based on the interview topic guides and added subsequent newly emerging themes as codes to the codebook. The codebook was discussed between authors SN, BT, EMM and SI and after coding 3–5 interviews and adjusted when needed. Coding and data analysis were done using NVIVO 12 software [19]. Finally, authors SI, EMM, BT, SN and IS summarized findings into identified themes with supporting quotes using the informants’ own words.

Results

Sociodemographic characteristics

Out of the 382 persons who were invited to participate, 369 (96.6%) persons accepted to perform COVID-19 self-testing, of whom five (1.3%) persons had positive self-test results. Thirteen (3.4%) persons declined self-testing out of fear and anxiety about positive test results. Six (1.6%) persons obtained invalid self-test results, of whom five (83.3%) chose to repeat the self-test, all with negative test results. Figure 1 presents an overview of the COVID-19 self-test cascade of care. Table 2 presents characteristics of the study participants.

Fig. 1.

Overview of the COVID-19 self-test cascade of care

Table 2.

Sociodemographic characteristics of survey participants

Characteristics	Clients
	n (%)
Total participated	369 (100.0%)
Sex
Male	158 (42.8%)
Female	211 (57.2%)
Age (years)
18–29	160 (43.4%)
30–39	65 (17.6%)
40–49	66 (17.9%)
50–59	31 (8.4%)
≥ 60	47 (12.7%)
Level of education
Primary School	138 (37.4%)
Secondary School	134 (36.3%)
Technical or Vocational School	22 (6.0%)
Higher education	22 (6.0%)
No Formal Education	53 (14.3%)
Employment status
Employed	130 (35.2%)
Self-employed	115 (31.2%)
Retired/Pensioner	12 (3.3%)
Student	45 (12.2%)
Unemployed	67 (18.1%)
Geographic location (all)
Maputo City	168 (45.5%)
Chamanculo A Community Post	55 (32.7%)
Xipamanine Community Post	56 (33.3%)
Mafalala Community Post	57 (33.9%)
Maputo Province	56 (15.2%)
Matola Gare Community Post	56 (100%)
Gaza Province	145 (39.3%)
Praça Municipal Community Post	52 (35.9%)
Senta Baixo Community Post	38 (26.2%)
Carmelo Hospital	55 (37.9%)

Acceptability of COVID-19 self-testing

COVID-19 self-testing was acceptable among clients and HCWs. Most clients and HCWs perceived self-testing as (very) important for early diagnosis of COVID-19 (98.6% and 100%, respectively) and increasing accessibility to COVID-19 testing and diagnosis (98.1% and 100%, respectively) (Table 3). In interviews, most clients expressed a strong preference for COVID-19 self-testing over other test methods because of more comfort (no pain) and the fast availability of results. Additional benefits of COVID-19 self-testing mentioned by clients and HCWs included flexibility in terms of testing timing, elimination of the need for travel to health facilities, and enhanced testing comfort.

Table 3.

Acceptability of COVID-19 self-testing as perceived by clients and healthcare workers

	Clients	HCWs
	N (%)	N (%)
Total clients	369	13
Perceived risk of getting COVID-19
Elevated risk	67 (18.1%)	6 (46.1%)
Moderate risk	94 (25.5%)	4 (30.8%)
Mild risk	42 (11.4%)	1 (7.7%)
Minimal risk	135 (36.6%)	1 (7.7%)
No risk	31 (8.4%)	1 (7.7%)
Perceived importance of self-test for early diagnosis of COVID-19
Especially important	175 (47.4%)	11 (84.6%)
Important	189 (51.2%)	2 (15.4%)
Makes no difference	4 (1.1%)	0
Not important	1 (0.3%)	0
Perceived importance of self-test to increase accessibility to COVID-19 testing
Especially important	144 (39.0%)	7 (53.8%)
Important	218 (59.1%)	6 (46.2%)
Makes no difference	6 (1.6%)	0
Not important	1 (0.3%)	0
Perceived importance of self-test to relieve the burden on the healthcare workers
Especially important	161 (43.6%)	7 (53.8%)
Important	204 (55.3%)	6 (46.2%)
Makes no difference	3 (0.8%)	0
Not important	1 (0.3%)	0
Likeliness to use self-test again after developing symptoms suggestive of COVID-19
Very Likely	131 (35.5%)	n.a.
Likely	191 (51.8%)	n.a.
Neutral	6 (1.6%)	n.a.
Unlikely	6 (1.6%)	n.a.
Very unlikely	35 (9.5%)	n.a.
Likeliness to use self-test again after being a contact of a confirmed COVID-19 patient
Very likely	128 (34.7%)	n.a.
Likely	193 (52.3%)	n.a.
Neutral	11 (3.0%)	n.a.
Unlikely	9 (2.4%)	n.a.
Very unlikely	28 (7.6%)	n.a.

HCW Healthcare workers

“The self-test is simpler, it’s not scary. The other one [COVID-19 test] wasn’t scary, but it hurts a little compared to this one. This one you don’t feel anything” (Client, 307 − 0011).

COVID-19 self-testing was also perceived as a tool to relieve work burden for healthcare facilities by 98.9% of clients and 100% of HCWs (Table 3). In interviews, HCWs expressed their preference for self-testing over other testing methods because they believed it could also reduce the risks of transmission associated with overcrowding in health facilities.

“It would be very useful [implementing self-tests], as I believe many people know that in hospitals there is a lot of adhesion, there are long queues, so this self-test will reduce overcrowding in hospitals” (HCW, 302-0002).

Most clients reported the possibility of using the COVID-19 self-test again when experiencing COVID-19 symptoms or being in contact with a person with confirmed COVID-19 (88.9% and 90%, respectively) (Table 3). In interviews, clients expressed that they would recommend COVID-19 self-testing to friends and family members.

“Because the self-test, I felt it is better compared to the previous testing, that is why I will ask other people to come forward to do the test themselves. When I get home, I will tell everyone that if we have a cough or fever again, we will go to the hospital, there is a test that you can do by yourself” (Client, 307 − 0022).

Some HCWs expressed in interviews fear of a positive self-test result and anticipated stigma related to having COVID-19 could prevent people from taking the self-testing for COVID-19 and therewith hamper acceptability. Furthermore, HCWs expressed in interviews that as COVID-19 is no longer defined as a public health emergency of international concern, people may lack motivation and therefore have low willingness to test for COVID-19. That said, it may also reduce countries’ efforts in making COVID-19 self-tests available.

“My biggest concern is that COVID-19 has been declared by the WHO to no longer be an international health emergency, and this will make people think they no longer need to worry and will not want to get tested. Even the government may not buy many tests because the demand may have reduced” (HCW, 307-0001).

To increase acceptability for a successful implementation of COVID-19 self-tests, clients and HCWs indicated the need to use different media channels, including community radio, TV spots, and leaflets for communication and health education to ensure everyone understands the main messages. Furthermore, HCWs stressed that persons with a positive self-test should to be ensured with linkage to a health facility. They said this could strengthen the trust and acceptability of the self-test by the community and HCWs.

“Dissemination of information about the test itself, perhaps creating a situation of putting up posters, advertising spots, even sometimes in the morning small campaigns, small lectures about self-testing, perhaps in health units or testing places, would be very good for clients to accept it, let’s say, in a light way.” (HCW, 304-0002).

“I believe that the linkage that we establish between the client and the health facility after testing positive for COVID-19, helps people in the community to accept the self-testing. […] They know that they will always be well treated at the health facility when referred to [after a positive test].” (HCW, 302-0002).

As the conditions for using the self-test, clients and HCWs highlighted the importance of it being available through either health facilities or at lower levels (e.g. pharmacies), supported by community leaders and provided free of cost. If COVID-19 self-tests need to be charged for by health authorities, our clients and HCWs suggested a price of around 8 US dollars per self-test.

“For me, the most important first step to guarantee the acceptability of the self-test is its availability and ease access, it has to be in health facilities and be free for anyone to be able to acquire”. (HCW, 307-0001)

“I prefer this test because it’s free, I don’t have to worry about having money to do it and I don’t even have to go to the health facility” (Client, 307-0009).

Feasibility of COVID-19 self-testing

Overall, 92.3% of HCWs perceived the self-test to be feasible and usable for future implementation in daily practice. Among those who previously tested for COVID-19, 171 clients (46.3%) perceived self-testing as (very) convenient. Overall, 98.4% of clients and 76.9% of HCWs considered the self-test procedure easy to do (from the sample collection to the result interpretation) (Table 4). Some HCWs did express in interviews the need to intervene when procedures were misunderstood or executed incorrectly by clients.

Table 4.

Feasibility of COVID-19 self-testing perceived by clients and healthcare workers

	Clients	HCWs
	n (%)	n (%)
Total	369	13
Clarity of (written) self-test instructions
Yes, truly clear	196 (53.1%)	7 (53.8%)
Yes, it was clear and understandable	143 (38.8%)	6 (46.2%)
Yes, but some parts were difficult to understand	24 (6.5%)	0
No, not clear or understandable	6 (1.6%)	0
Not clear at all	0 (0.0%)	0
Usage of other information sources for instruction or guidance
No	283 (86%)	n.a.
Yes	86 (30.4%)
Website on Testing for COVID-19	61	n.a.
Video on how to take a test	0
Advice from another person	81
Information from a mobile app or health hotline	0
Pamphlet or informational flyer	3
Usage of other information sources for follow-up test after self-test
No	275 (X%)	9
Yes	94	4
Website on Testing or COVID-19	0	9
Advice from another person	75	0
Information from a mobile app or health hotline	0	1
Pamphlet or informational flyer	87	5
Likeliness to use mobile phone app for self-test instructions
Unlikely	139 (37.7%)
Likely	230 (62.3%)
Likeliness to use mobile phone app for reporting self-test results to healthcare facility
Likely	232 (62.9%)	2 (15.4%)
Unlikely	137 (37.1%)	11 (84.6%)
Was the self-test easy-to-use?
Easy	364 (98.6%)	10 (76.9%)
Difficult	5 (1.4%)	3 (23.1%)
Easiness to understand the information about the self-test
Easy	363 (98.4%)	13 (100%)
Difficult	6 (1.6%)	0
How easy was the sample collection for the self-test?
Easy	364 (98.6%)	13 (100%)
Difficult	5 (1.4%)	0
How easy was it to get the sample in the buffer tube of the self-test?
Easy	366 (99.2%)	13 (100%)
Difficult	3 (0.8%)	0
Easiness of transferring of the material to the self-test device
Easy	361 (97.8%)	13 (100%)
Difficult	8 (2.2%)	0
Easiness of reading and interpretating self-test results
Easy	348 (94.3%)	13 (100%)
Difficult	21 (5.7%)	0
Usage of a timing device to wait for the self-test result
Yes	344 (93.2%)	12 (92.3%)
No	24 (6.5%)	0
Do not know	1 (0.3%)	1 (7.7%)

“I think it was easy, although I came across with some patients who were not easy to interpret the information in the leaflet and put into practice the information contained in the information leaflet, but I think everything went very well, most patients were able to interpret, and they performed very well.” (HCW, 305-0001).

“Everyone understood the information. I gave the information leaflet to the patient to read all the procedures and after reading they had to interpret what they see in their information leaflet, if they cannot, that is where I had to intervene explaining that it can’t be done this way, it has to be done this way” (HCW, 305-0001).

Although written self-test instructions were perceived as (very) clear by most clients (339 (91.9%)) and all HCWs, 86 (30.4%) clients also consulted other information sources for instructions, such as the advice from another person or a website on testing for COVID-19 was most frequently reported. Clients recommended including instructions in the local language. During interviews, HCWs explained that testing instructions, provided in Portuguese and English, can be difficult to comprehend for persons living in rural areas and who speak local languages.

“The barriers for the community are that we know that most of the population does not speak Portuguese, does not speak English and the leaflet, the procedures are only in two languages [Portuguese and English], more work must be done in the process of educating the communities so that they understand all procedures in the local language” (HCW, 306-0001).

Additionally, some clients and HCWs expressed their concern for the ability of low (health) literate persons to understand written and/or pictorial COVID-19 self-test instructions. They suggested developing audio and/or video self-test instructions to accommodate this group of individuals.

“Self-testing may be very challenging for individuals who haven’t studied because there is a manual provided, and if a person cannot read, they cannot proceed. This also applies to those who are deaf; if a person cannot hear or speak, they face difficulties. Therefore, implementing some videos for assistance would be beneficial.” (Client 305-007)

Of the self-testing procedure, clients rated the interpretation of test results most frequently as difficult (n = 21 (5.7%)). During interviews, HSWs suggested some improvements regarding the instructions on the waiting time before results can be interpreted. HCWs explained that self-test results were visible after 3 min, which confused clients about the need to wait between 15 and 30 min before official interpretation of test results could be done.

In terms of the reliability of self-test results, all clients and HCWs mentioned that they trusted the self-test results. For clients, the main reason for relying on the result was the fact that they carried out the tests themselves and monitored the entire process.

“Because [reasons to trust the test] at the time of taking the sample it was done in person, following the instructions, where what should be done, how should I do in person following those instructions and also when the test is not done correctly the result ends up being invalid” (Client, 302-0002).

Discussion

We assessed the acceptability and feasibility of COVID-19 self-testing as part of a community-based bi-directional screening event for tuberculosis and COVID-19. Our study results show that COVID-19 self-testing is acceptable and feasible as a first screening tool, as it is easy and safe to perform, increases accessibility to COVID-19 testing and relieves healthcare facilities in a context of high demand for diagnostic services. To ensure successful implementation, COVID-19 self-testing should be accompanied by health education through different channels and audio-visual self-test instructions, made available through health facilities and lower levels (e.g. pharmacies) by engaging community structures, and provided free of charge or else at a maximum of eight dollars.

Acceptability of COVID-19 self-tests in our study was high and facilitated by their ease of use, comfort (no pain), and the fast availability of test results. Similarly, the perceived acceptability and likeliness of future use were also high in previous values and preference studies conducted in Peru and South Africa, but not in Kenya (< 50%) [9, 19, 20]. Another factor driving the acceptability of COVID-19 self-testing was perceived as an important tool to promote early diagnosis, alleviate the workload at the health facilities and prevent transmission within health facilities. Similar perceptions were also reported in studies in South Africa, Indonesia, Greece, and Cyprus [5, 11, 20]. In our study, clients expressed in interviews to prefer self-testing over directly assisted PCR-testing with different perceptions on the location of the self-testing (at health facilities or elsewhere). Similarly, in a study in Mozambique on HIV self-testing, some clients preferred directly assisted HIV self-testing, while others preferred to do self-testing at home and others at the health facility [21]. Another Mozambiquan study on self-sampling for cervical cancer screening found majority of clients to prefer self-collection of samples for screening [22]. These results suggest that self-testing irrespective of the disease is highly acceptable and by a considerable proportion preferred, but options for assisted self-testing at health facilities should be available.

Feasibility of self-testing is greatly determined by the ability of the population of interest to understand the procedures and execute them correctly to ensure high accuracy of the self-test result. Two previous Mozambiquan studies on HIV self-testing and COVID-19 nasal swabs showed high accuracy of self-tests among first time users [21, 23]. Nevertheless, de Schacht et al. (2024) showed procedural errors, including incorrect tube positioning, specimen collection and waiting time to result interpretation among almost half of their HIV self-testing population in Mozambique [24]. Our study results showed that only small proportions of the clients had difficulties with executing the self-test and needed assistance. Procedures that were sometimes misunderstood were waiting time for the result and interpretation of results. Another study conducted in Malaysia showed similar results [25]. Furthermore, concerns existed about the understandability of self-test instructions for persons from rural areas and persons with low (health) literacy [26, 27]. Marbán-Castro et al. (2024) provided a training session at study sites to demonstrate the use of the tests to participants, which likely contributed to the feasibility of COVID-19 self-testing in their setting [25]. Hence, we suggest that implementation of self-testing for any type of disease should be accompanied by audio-visual instructions in local languages, educational sessions at strategic places and support through social media and hotlines to ensure feasibility, including testing quality and accuracy. Such interventions could potentially also limit possible fear of sampling incorrectly or discomfort with their bodies [28].

We studied the feasibility and acceptability of COVID-19 self-testing during a time in which the number of people seeking COVID-19 diagnostic services had declined and COVID-19 is no longer defined as a public health emergency of concern. This may have influenced our results: the feasibility of COVID-19 self-testing in a time of fear for COVID-19 and an overwhelming health system may have been different. Furthermore, we could not evaluate the linkage to care due to a low number of clients with a positive self-test. This linkage to care was found to be one of the aspects that needed improvement in a feasibility study on COVID-19 self-testing in Malaysia [25]. Despite these limitations, our study results contribute to the scarce evidence-body of acceptability and feasibility of client-administered COVID-19 self-testing in LMICs. Furthermore, lessons learned from this study could also be of value for the implementation strategies of self-testing for other conditions.

Self-testing in Mozambique has been studied for COVID-19, HIV, and Human Papilloma Virus (HPV). Recently, the World Health Organization also pre-qualified a self-test for the Hepatitis C virus. Hence, self-testing for communicable (such as diabetes), and non-communicable diseases (such as COVID-19, HIV, HPV, Hepatitis C) are increasingly being developed and implemented. To enhance the efficiency of (future) implementation of self-tests -also in light of pandemic preparedness, self-testing delivery packages should be developed and/or optimized for different user segments [25]. Lessons learned should be taken from evidence on self-testing and existing structures. Furthermore, empowerment of the community is needed to understand and manage their self-diagnosis following self-testing, including a clear objective from the client, an efficiently working easy-to-use reporting system for self-test results and rapid communication with health professionals using telemedicine or another virtual testing for persons with a positive self-test [29, 30]. Linguistic and culturally appropriate (health) education and communication interventions should be developed using a community-based participatory approach, taking into account the lessons learned from other programs and implementation studies [24].

Between February and December 2023, we observed a nationwide decline in the number of individuals seeking COVID-19 diagnostic services. This trend presented a potential limitation regarding the achievement of our target sample size within the projected timeframe. To mitigate this, the study team implemented strategic demand generation activities at all study sites. These efforts focused on community education to communicate to the public that the study’s primary objective was the evaluation of the testing cascade and participant engagement, rather than the identification of positive cases. By emphasizing that, we were able to maintain recruitment despite the decrease in symptomatic presentations. Additionally, because the study was conducted in selected community posts during a period of low COVID-19 prevalence, reduced healthcare-seeking behaviour, and does not include in the study design inferential analysis, the findings may have limited generalizability to other settings or periods with higher disease burden or different healthcare utilization patterns.

Conclusions

We assessed the acceptability and feasibility of COVID-19 self-testing as part of a community-based bi-directional screening event for tuberculosis and COVID-19. Our study results show that COVID-19 self-testing is acceptable and feasible as a first screening tool, as it is easy and safe to perform, increases accessibility to COVID-19 testing and relieves healthcare facilities in a context of high demand for diagnostic services. To ensure successful implementation, COVID-19 self-testing should be accompanied by health education through different channels and audio-visual self-test instructions, made available through health facilities and lower levels (e.g. pharmacies) by engaging community structures, and provided free of charge or else at a maximum of eight dollars.

COVID-19 self-testing is acceptable and feasible if provided with linguistic and culturally appropriate health education and instructions at an affordable price. Self-test delivery packages and systems, including reporting systems and linkage to care, should be developed based on the current evidence-body using a community-based participatory approach that could be adapted to existing or future self-tests for infectious diseases or other non-communicable diseases.

Abbreviations

Ag-RDT

Antigen-detection rapid diagnostic tests

CHWs

Community Health Workers

HCWs

Healthcare workers

LMICs

Low-middle income countries

RT-PCR

reverse transcription-polymerase chain reaction testing

REDCap

Research Electronic Data Capture

Authors’ contributions

IS, CM, RRP, RP, NM, SI, and EMM conceptualized and designed the study. SI, EMM, CP, SN, IA, CP, UB, CB, AN, PM, CB and BT executed the study and collected the data. SI, BT, SK, IS EV, CP, and EMM performed data analysis and interpretation. SI, EMM, and IS drafted the manuscript. All authors reviewed the draft manuscripts and approved the final manuscript.

Funding

This project is part of the IMPAACT4C19 project, funded by UNITAID.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study obtained approval from multiple ethical review bodies, including the CERC-WHO (CERC.0134 − 18/05/2023), Institutional Bioethics Committee of the National Institute of Health in Mozambique (222/CIE-INS/2022-28/11/2022), and National Bioethics Committee of Mozambique (169/CNBS/2022- 09/03/2023). This study received additional approval from ANARME to authorize the importation and use of self-test device in research (48/380/ANARME /2023–11/05/2023). All participants provided written consent for participation in this study.

Competing interests

The authors declare no competing interests.