Article Text

Short Report
The performance of the Alere HIV combo point-of-care test on stored serum samples; useful for detection of early HIV-1 infections?
  1. Carla van Tienen1,
  2. Sharona Rugebregt1,
  3. Sandra Scherbeijn1,
  4. Hannelore Götz2,3,
  5. Corine GeurtsvanKessel1
  1. 1Department of Viroscience, Erasmus Medical Center, Rotterdam, The Netherlands
  2. 2Department of STI Control and Sexual Health, Rotterdam-Rijnmond Public Health Service, Rotterdam, The Netherlands
  3. 3Department of Public Health, Erasmus Medical Center, Rotterdam, The Netherlands
  1. Correspondence to Dr Carla van Tienen, Department of Viroscience, Erasmus Medical Center, Wytemaweg 80, Nb1052, Rotterdam 3015 CN, The Netherlands; c.vantienen{at};


Introduction The Alere HIV-1/2 Antigen/Antibody Combo point-of-care test is a commercially available 4th-generation rapid test for the diagnosis of HIV infection, including acute infection. We evaluated the sensitivity of this test in samples from patients with acute, recent or chronic HIV-1 infection.

Methods A validation of the test was performed using 89 HIV-positive serum samples collected in 2008–2016, that were stored at −20°C. Twenty-three samples were only p24-positive (acute infection); 49 samples were antibody-positive and p24-positive (recent infection); 17 samples were only antibody-positive (chronic infection). HIV infection was confirmed by standard-of-care assays and PCR. Samples came from patients attending an outpatient clinic for STDs at the Public Health Department and from patients within the Erasmus Medical Center, Rotterdam, the Netherlands.

Results The overall sensitivity of the test for diagnosing HIV infection based on detection of p24 antigen and/or antibodies was 92% (95% CI 86% to 98%) (82/89). In acute sera with only p24 antigen positivity, the sensitivity of the test decreased to 65% (95% CI 46% to 85%) (15/23). When both antibody and antigen testing were positive, the p24 sensitivity was only 24% (95% CI 12% to 36%) (12/49), but in these sera the final test result was positive in all sera (49/49) due to the positive antibody component.

Conclusions In a laboratory setting, this test has an overall sensitivity of 92% to detect any stage of HIV-1 infection using sera specimens. It performs relatively well in detecting early HIV and may be beneficial as an initial screening in patients with a recent exposure to HIV. Additional testing in a laboratory setting remains mandatory as a proportion of acute HIV-1 infections are missed with this test.

  • P24

Statistics from


Rapid tests or point-of-care tests (POCTs) for detection of HIV-1 are often used in clinical practice in persons who have a high risk of HIV infection because they give a quick result, can improve the uptake of HIV testing, enable rapid follow-up testing and are relatively cheap. They are used at our Public Health STI clinic, in persons at high risk for HIV, specifically men who have sex with men or those notified for HIV. The results of the POCT are always confirmed by a second routine fourth-generation laboratory test. During acute HIV-1 infection, p24 antigen can be detected in serum while antibodies titres are still undetectable. Acute HIV-1 infections are important to diagnose because an early start of treatment improves the outcome of the infection in the individual and because it prevents transmission to others.1 The latest (fourth) generation POCTs can detect and differentiate antibodies and p24 antigen in sera and thereby aim to reduce the window period. The Alere HIV-1/2 Ag/Ab Combo rapid test (Combo RT, Alere Medical) is a novel fourth-generation test which can be used on serum, plasma and whole blood specimens and has a turnaround time of 20 min. A recent paper by Fitzgerald et al in this journal describes a high sensitivity (88%) of this test in sera with detectable p24 antigen. Unfortunately it does not report specifically on the sensitivity in case antibodies are present besides the p24 antigen, which is more often the case in screening.2

The previous fourth-generation HIV POCT, the Alere HIV-1/2 Ag/Ab Determine, showed a varying sensitivity for acute infections with only p24 antigen. Reported sensitivities varied from 0% (0/2 samples;3 0/8,4), 0% (0/13,5), 3% (1/34,6), 23% (13/57,7), 50% (3/6,8), 54% (14/269) to 67% (26/3910).

We investigated whether the new Alere Combo POCT is able to detect p24 antigen with or without antibodies (as found in acute or recent HIV infection) in HIV-1 infections in samples from the Public Health STI Clinic and the Erasmus Medical Center (MC) in Rotterdam.


Study design

The validation of Combo POCT was performed on retrospectively collected samples.

Study participants

Eighty-nine anonymised samples were tested from patients attending an STI clinic at the Public Health Centre and from patients within the Erasmus MC in Rotterdam, the Netherlands.


The sera were collected in 2008–2016 and stored at −20°C. Twenty-three samples were only p24-positive; 49 samples were both antibody-positive and p24-positive and 17 samples were only antibody-positive. Follow-up samples were available from 14 patients.

Samples were first screened by an Antibody/Antigen non-discriminatory test by the Architect (Abbott) (until 2013) or by an Antibody/Antigen discriminatory test by the LiaisonXL, (Diasorin) (chemiluminescence immunoassay). In case of positivity in the Architect, quantitative detection of p24 antigen was performed using enzyme linked fluorescent assay (VIDAS, Biomerieux) and a line immune assay (Inno-Lia, Inogenetics) for antibody confirmation. In case of positivity in the LiaisionXL, p24 was confirmed by PCR (Roche Molecular Systems, commercial PCR) and/or antibodies were confirmed by the Inno-Lia. When the confirmation test by the VIDAS or the Inno-Lia was negative, PCR was performed to confirm or rule out HIV infection. The Combo POCT was performed following the manufacturer's instruction. Clinical information was not known, but the results of the previous HIV tests were known to the performer of the Combo POCT.


Sensitivity was expressed in percentages with a 95% CI. The χ2 test was used to compare median levels of p24 antigen.


The overall sensitivity for diagnosing HIV-1 infection of the Combo POCT was 92% (95% CI 86% to 98%) (82/89 HIV-1 p24 and/or antibody-positive sera).

Acute HIV-1 infection

In the panel of patient sera with acute HIV-1 infection (p24 only), the sensitivity of the Combo POCT of detecting p24 only was 65% (95% CI 46% to 85%) (15/23) (table 1). One sample did not give a positive p24 antigen signal, but a positive antibody response in the Combo POCT. The eight sera that were p24-negative in the Combo POCT had a lower level of p24 (median=9 pg/mL) as compared with the sera that were tested p24-positive (median=168 pg/mL) (p<0.05; χ²).

Table 1

Comparison of the Alere fourth-generation HIV Combo POCT to PCR and p24 antigen in acute HIV-1 infection

Recent HIV-1 infection

For the panel of recent HIV-1 infection (both p24-positive and antibody-positive), the sensitivity for detecting p24 with the Combo POCT was 24% (12/49), but the antibody component was positive in all sera increasing the total result to 100% (49/49). Like in the sera with only the p24 antigen present, the sera in which the p24 component was detected by Combo POCT had a higher p24 level (median=98 pg/mL) than the sera that were p24-negative (median=5 pg/mL) (p<0.05; χ²).

Chronic HIV-1 infection

In the sera of patients with a chronic HIV-1 infection (antibody-positive only), the sensitivity of the Combo POCT for detection of antibodies was 100% (17/17).

From the 14 patients of whom a follow-up serum was obtained, 6 patients presented with only p24 antigen in their first serum (positive in the Combo POCT). Of these six patients, all the sequential sera showed seroconversion by both Combo POCT and confirmation assay.


We validated the latest fourth-generation Combo POCT by Alere to evaluate its performance in detecting different stages of HIV infection. The sensitivity to detect acute HIV infection with only p24 antigen was 65% (15/23). Two studies have reported on the sensitivity of this Combo POCT, which was 100% (6/6) for p24 only sera in the study by Ottiger and Huber8 and 88% (30/34) for p24 with or without antibodies (as this is not specifically stated in the paper) by Fitzgerald et al.2 Our lower sensitivity might be explained by the fact that the study by Ottiger and Huber8 only included samples with a high HIV RNA (all higher than 106 copies/mL) and high p24 antigen (lowest value: 49.1 pg/mL). The higher sensitivity by Fitzgerald et al might as well be related to the higher viraemia in the samples, as the lowest value of the p24 antigen measured was 8.3 pg/mL while in our samples this was 1.1 pg/mL; another or additional explanation is that the selected samples were mostly antibody-negative (not specifically reported in the paper). HIV RNA levels were not reported in this previous study.2 It has to be noted that the Combo POCT insert reports an analytical sensitivity of 2 IU/mL, which is approximately equivalent to 12.5 pg/mL. Thus, the Combo POCT did detect all but one of the p24-positive samples above this level.

Compared with previous studies of the fourth-generation Determine POCT, the sensitivity for p24 seems to have increased, although the sensitivities reported vary (0–67%).3 ,4 ,6–9 These differences could be caused by different sample sizes, different patient populations, different selection of samples and differences in HIV-1 subtypes. For example, the study by Beelaert and Fransen10 reported a relatively high sensitivity (67%) of the Determine POCT in seroconversion sera obtained from patients attending an STI clinic in Belgium. The level of the p24 antigen in the samples was not reported in this study nor was PCR performed, so a selection of samples with a high level of p24 might explain the higher sensitivity.

The sensitivity for recent HIV-1 infections (both p24-positive and antibody–positive), was lower than the sensitivity for p24 only sera. This could be due to the problem of the antigen-antibody complex formation leading to false negativity of the p24 antigen. The manufacturer claims to have improved the test by trying to disrupt antigen-antibody complexes.8 Based on our results, this issue has not been resolved. It is important to note however, that the antibody component was picked up in all 49 samples, so the Combo POCT would not have missed any recent HIV infection.

A potential limitation of this study is that we performed the Combo POCT on serum samples only. The insert of the test claims to have the same analytical sensitivity for serum, plasma and whole blood specimens, but a direct comparison would be useful to confirm this.

Overall, the sensitivity of detecting any type of HIV-1 infection with the Alere Combo POCT was 92% and in acute HIV-1 infection with only p24 antigen the sensitivity was 65%. The p24 sensitivity decreased to 24% in samples in which both p24 and antibodies are positive. Nevertheless, the antibody component of the Combo POCT performed very well in the recent HIV-1 patients (100%). It may be beneficial as a tool for screening patients with a recent exposure to HIV but clinical validation is required and has been started in our STI Public Health Department. Additional testing in a laboratory setting remains mandatory as a proportion of acute HIV-1 infections are missed with the Combo POCT. However, once a p24-positive result is seen, urgent confirmatory testing is done to confirm acute infection, followed by direct referral to an HIV treatment centre.


The authors thank Professor Charles Boucher for his valuable input and discussions and Dr Annemiek van der Eijk for her comments on the manuscript. The authors also thank Alere for kindly providing the tests for this validation, free of charge.


View Abstract


  • HG and CGvK contributed equally.

  • Handling editor Jackie A Cassell

  • Twitter Follow Carla van Tienen @carlavantienen

  • Contributors All authors have seen and approved the content of this manuscript. SR and SS tested and located the samples of this study. SR helped in analysing the data. SS has overseen the standard-of-care HIV laboratory testing in the past years and provided important information on these tests. HG and CGvK conceived and planned this study. CvT analysed the data and wrote the draft of the manuscript. HG and CGvK edited the manuscript.

  • Competing interests None declared.

  • Ethics approval Medical Ethical Committee Erasmus Medical Center (MEC-2015-306).

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

  • Data sharing statement Previous work on a few samples was done using the previous version of the POCT, the Determine. These results are stored at our department and can be shared with persons who have a specific (research) interest in this field upon personal correspondence with the corresponding author.

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.