Article Text

Download PDFPDF

Original article
Supervised blood-based self-sample collection and rapid test performance: a valuable alternative to the use of saliva by HIV testing programmes with no medical or nursing staff
  1. María J Belza1,2,
  2. M Elena Rosales-Statkus3,
  3. Juan Hoyos1,2,
  4. Pilar Segura4,
  5. Eva Ferreras2,5,
  6. Rebeca Sánchez5,
  7. Gemma Molist2,5,
  8. Luis de la Fuente2,5,
  9. the Madrid Rapid HIV testing Group*
  1. 1Escuela Nacional de Sanidad. Instituto de Salud Carlos III, Madrid, Spain
  2. 2CIBER Epidemiología y Salud Pública (CIBERESP), Spain
  3. 3Servicio de Medicina Preventiva. Hospital Central de la Cruz Roja San José y Santa Adela, Madrid, Spain
  4. 4Servicio de Medicina Preventiva. Hospital General Universitario de Ciudad Real, Ciudad Real, Spain
  5. 5Centro Nacional de Epidemiología. Instituto de Salud Carlos III, Madrid, Spain
  1. Correspondence to Juan Hoyos, Escuela Nacional de Sanidad. Instituto de Salud Carlos III. C/Sinesio Delgado, 8. 28029 Madrid, Spain; jhoyos{at}


Objectives Some saliva-based HIV testing programmes have resulted in an unacceptable percentage of false positives. Many countries require blood-based testing programmes to have doctors/nurses. The authors evaluate whether, after brief training and under the supervision of a skilled counsellor, blood-based self-sample collection and rapid test performance could be a valuable alternative.

Methods 208 Spanish-speaking attendees at a street-based HIV testing programme in Madrid participated in the study. Participants were tested twice, first in the study and then in the programme, using the same finger-stick whole-blood rapid test (Determine HIV-1/2 Ag/Ab Combo®). Based on previously adapted instructions, the study counsellor explained the procedure to follow throughout the test. Participants then performed the test under the guidance of the counsellor. Demographic and risk behaviour data were collected by a self-administered questionnaire. The test results in the programme and the study were read by the study counsellor.

Results 99.0% (95% CI 96.6% to 99.9%) of participants had a valid result in the study test, the same percentage as in the programme test conducted by the doctor/nurse. Two persons had invalid test results in both the study and the programme, but they were not the same persons.

Conclusion The study provides clear evidence that this methodology is a valuable alternative to saliva for HIV testing programmes when medical or nursing staff required to take blood samples is not available.

  • HIV testing
  • rapid test
  • testing services
  • HIV
  • public health
  • epidemiology (general)
  • programme evaluation
  • outreach services

Statistics from

Request Permissions

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


The use of rapid HIV testing to reduce the time it takes for infected persons to learn their serostatus is being consolidated as an important strategy in recommendations recently approved in different developed countries.1–3 The rapid tests currently available require that samples of saliva or blood be obtained. In practice, the saliva-based tests are less sensitive than those that are blood based.4 Moreover, some saliva-based programmes have resulted in an unacceptable percentage of false positives, although the reason for this is not clear.5 6 For these reasons, the Centers for Disease Control and Prevention has reported that the ‘strategy of following up reactive oral fluid test results with an immediate finger-stick whole-blood test reduced the number of apparent false-positive oral fluid test results and might be a useful strategy in other settings and locations’.5

One of the most well-known settings for the use of rapid tests are programmes in non-clinical settings aimed at capturing especially vulnerable populations that face barriers in accessing more standard diagnostic services. These programmes are often conducted by organisations that do not have medical or nursing staff. In many countries, however,7–9 blood-based testing programmes are required to have doctors or nurses on staff. Other countries have changed their legislation so non-health professionals could receive special training to be allowed to perform these tests.10 11 For those who have not adopted these reforms, one solution that would allow these organisations to carry out blood-based tests would be for their clients to obtain the blood sample—since the process is very similar to that used for blood glucose self-testing—and for a non-medical counsellor to read the test results and provide counselling.

In this feasibility study, we evaluate whether persons seeking a rapid HIV test are able to obtain and properly deposit a capillary blood sample after a brief training of 2–3 min under the supervision of a skilled counsellor. We estimate the percentage of valid tests obtained under the real-life stressful conditions of persons seeking a rapid HIV test.


Setting, sampling and recruitment

Between the cold rainy months of November 2009 and January 2010, participants were selected from persons requesting a rapid HIV test in a programme operated by the non-governmental organisation ‘Madrid Positivo’. This programme was carried out in a mobile unit located in three places in the city of Madrid: a central street in a neighbourhood with a large number of establishments catering to the gay community and two campuses of the Complutense University. After signing the informed consent necessary to perform HIV testing and while awaiting the test, a study investigator briefly explained the characteristics of the study to those clients who said they were Spanish speaking and had not been enrolled in the study before and asked if they would be willing to participate.

Those who accepted were sent to tents where the study was performed and received by a doctor or nurse who could conduct both the counselling tasks for the study and the medical tasks for the programme, in accordance with the requirements of Spanish legislation.

Self-blood sample collection and rapid test performance

The counsellor again explained the study and requested that participants sign to indicate their informed consent to the study. Then, the counsellor explained the proper way to obtain two drops of capillary blood by finger prick and to deposit them in the correct place on the reactive strip. This explanation was made following visual instructions developed ad-hoc and designed by the investigators after repeated simulation with volunteers using a non-reactive test. Instructions consisted of clear images of the main steps of the process supported by short sentences on how to do it and how not to do it. Participants then performed this process under the observation and guidance of the counsellor but without the counsellor's direct intervention. The doctor/nurse counsellor subsequently performed the test again, this time obtaining the blood sample directly, under the same conditions as used for everyone receiving the rapid HIV test in this programme.

The test selected for the study was the same as the one used in the programme (Determine HIV-1/2 Ag/Ab Combo®). Testing in the programme was carried out following the manufacturer's instructions, but two modifications were introduced for the study. First, we dispensed with the capillary tube, since the experience of the study of Lee et al,12 and the qualitative pilot study conducted in the initial design stages showed that a large percentage of persons were incapable of using it correctly. It was decided instead to ask participants to deposit ‘at least two good drops’ of blood directly in the appropriate place on the reactive strip, although the doctor–nurse counsellor performed the test using the capillary tube. The second modification was to mount the reactive strip on a support, indicating the exact place where the blood should be deposited, for easier identification.

The test results for both the programme and the study were read by a doctor or nurse and were conveyed in the same way in the post-counselling session.

Data collection and analysis

The counsellor noted on a brief questionnaire his/her observations on how the participant performed each of these steps. While waiting for the test results, participants filled a self-administered questionnaire intended to collect demographic and risk behaviour data, as well as previous HIV test experience.

A descriptive analysis was performed. The statistical significance of differences was evaluated using the χ2 test for proportions.


Of the 267 participants who were offered the opportunity to participate in the study, 208 (77.9%) accepted. Of these, 64.4% were recruited in the gay neighbourhood, 42.9% were under age 25, 16.1% were immigrants, 57.6% had university studies, 90.7% were single, 28.4% usually lived outside the city and 2.0% earned most of their income from illegal activities. Some 37.6% were men who have sex with men (MSM), 21.5% were heterosexual men and 41.0% were women; 56.9% had not previously been tested for HIV (25.4% of MSM, 71.4% of heterosexual men and 77.8% of women) and 51.1% had not gone expressly to be tested but attended because they were passing by and saw the unit. There were no significant differences in any of these variables between those who accepted and those who refused to participate (table 1).

Table 1

Demographic, behavioural characteristics and HIV testing experience of participants and those who refused to participate in a supervised blood-based self-sample collection and rapid test performance study

Some 99.0% (95% CI 96.6% to 99.9%) (206/208) of participants had a valid result on the study test, and this percentage was the same for those tested by a doctor/nurse in the programme test. In both the study and programme test, two persons had invalid test results, but they were not the same persons. In the two cases of invalid study tests, the counsellor indicated that the blood sample had been correctly obtained and that a sufficient amount of blood had been deposited.

Four persons, 1.9% (95% CI 0.5% to 4.9%), were reactive for antibodies in both the study test and the programme test, but none was reactive to P24 antigen. All four were subsequently confirmed as positive.


The study offers clear evidence that people seeking rapid HIV testing can perform a finger stick blood test under the direction of a counsellor trained in guiding the process and supported by good instructions developed ad-hoc for this purpose.

Only 1% of participants would have had to repeat the test if only the sample obtained by the client had been available, exactly the same percentage as when the sample was obtained by a doctor or nurse. This proportion of invalid tests is lower than that found in another study similar to ours but in which the counsellor (not a doctor or nurse) conducted almost the whole process,13 and it was much lower than in another study using the same test, in which the whole-blood samples were obtained by health professionals.4 Moreover, unlike other feasibility studies conducted in populations that already know their serostatus, this study was conducted in persons undergoing the emotional stress of finding out whether or not they are infected with HIV—the same as would occur if this technique were applied in regular testing programmes. It should also be noted that the study was carried out under conditions much less favourable than would exist if this technique was employed in diagnostic programmes. However, a large proportion of participants were highly educated and only Spanish speakers were selected since counsellor guidance and manual instructions content were in this language. Although our results may not apply for non-Spanish speakers such as some recent immigrants, the latter would represent only a small percentage of the vulnerable population. Most of our immigrants are from Latin America and are included in our study.14 Results may not apply either to less educated groups, such as drug injectors. Manual instructions written in other languages or using only images or videos could be tested in the future.

Dispensing with the capillary tube appears to have had no effect, given the percentage of valid tests obtained. Using the capillary tube would have meant greater certainty that a sufficient quantity of blood had been deposited, but its use had already been shown to require training,12 and this could not be done in diagnostic programmes.

Doctors or nurses were used as counsellors in the study because of the need to have the same person conduct the entire process so as not to prolong the stress of those who had requested a ‘rapid test’. The study test had to be performed before the programme test to avoid any learning process, which would not be available in future uses of this technique in a programme. However, considering the simplicity of the process, we believe the results obtained with trained counsellors without medical or nursing staff qualifications would have been very similar.

In light of the study results, it is obvious that the methodology used successfully in this study could be assumed by NGO programmes lacking authorised health personnel to obtain blood samples in countries that require these qualifications. It could also be employed as a complementary test in cases of positive or uncertain results in other similar programmes that use saliva samples,5 without excluding the need for subsequent confirmation. Thus, a new option is available in these countries for diagnostic programmes in non-clinical settings—one which makes it possible to reduce the cost of some programmes and to employ peers that the social groups most vulnerable to HIV perceive as closer to them (eg, MSM or ethnic minorities), without giving up the higher sensitivity and predictive value of blood-based tests.

Key messages

  • Attendees of a street-based HIV testing programme are able to perform their own finger stick blood test under the direction of a counsellor, who could interpret the results.

  • This methodology is an alternative for programmes in community settings where doctors/nurses are not the best option because of cost, availability and social skills.

  • This methodology is also an alternative for programmes where working with peers could help attract vulnerable populations.



  • * Madrid Rapid HIV testing Group: Jorge Álvarez, Óscar Ayala, María J Bravo, Jorge Gutiérrez, Mónica Ruiz, José Pulido, Sara Santos, Sonia Fernández-Balbuena.

  • Funding This study was funded by Fondo de Investigación Sanitaria (FIS:PI09/9074) and Fundación para la Investigación y Prevención del Sida en España (FIPSE/36634/07).

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval The Ethics approval was provided by the institutional review board of the Instituto de Salud Carlos III (approval number: CEI PI 16_2009).

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