Diagnosing HIV infection in patients presenting with glandular fever-like illness in primary care: are we missing primary HIV infection?

Authors


Correspondence: Dr William Tong, Department of Infectious Diseases, St Thomas' Hospital, Westminster Bridge Road, London SE1 7EH, UK. Tel: 02071883147; fax: 02071883146; e-mail: william.tong@gstt.nhs.uk

Abstract

Objectives

The aim of the study was to examine the prevalence of HIV infection in patients presenting in primary care with glandular fever (GF)-like illness.

Methods

Samples from primary care submitted for a GF screen between April 2009 and June 2010 were identified. Samples without an HIV request were anonymized and retrospectively tested using a 4th-generation HIV antigen/antibody screening test. Reactive samples were further confirmed by an HIV antibody only test, with or without a p24 antigen assay. Antibody avidity testing based on the Recent HIV Infection Testing Algorithm (RITA) was used to identify individuals with evidence of recent acquisition (within 4–5 months).

Results

Of 1046 GF screening requests, concomitant HIV requests were made in 119 patients. Excluding one known positive patient, 2.5% (three of 118) tested HIV positive. Forty-five (4.3%) had a subsequent HIV test through another consultation within 1 year; of these, 4.4% (two of 45) tested positive. Of the remaining 882 patients, 694 (78.7%) had samples available for unlinked anonymous HIV testing, of which six (0.9%) tested positive. The overall HIV prevalence was 1.3% (11 of 857), with 72.7% (eight of 11) of cases missed at initial primary care presentation. Four of the nine (44.4%) available positive samples had evidence of recent acquisition, with three (75.0%) missed at initial primary care presentation.

Conclusion

Low levels of HIV testing in patients presenting in primary care with GF-like illness are resulting in a significant number of missed HIV and seroconversion diagnoses. Local policy should consider adopting an opt-out strategy to include HIV testing routinely within the GF-screening investigation panel.

Introduction

Primary HIV infection (PHI) or seroconversion illness is a self-resolving syndrome that occurs typically 2 to 4 weeks after infection in approximately 80% of individuals [1]. This symptomatic period usually lasts 2 to 3 weeks and often represents the only clinical manifestation of HIV infection before more advanced immunosuppression many years later.

Characterized by a combination of nonspecific symptoms, including fever, myalgia, headache and rash, it is well recognized that individuals with PHI often present with a clinical picture of a glandular fever (GF)-like illness. In addition, patients with advanced HIV disease often present with nonspecific illness such as lethargy and lymphadenopathy, which may also be thought to be GF-like. However, HIV testing is not always requested as part of the GF investigation panel [2].

UK HIV testing guidelines state that an HIV test should be considered in the investigation of patients with a mononucleosis syndrome [1]. Despite these guidelines, recent research highlights missed opportunities for offering HIV testing outside traditional genitourinary medicine (GUM) and antenatal settings [3-5]. Rates of general practitioner (GP)-offered testing, in particular, remain low [6, 7]. Missed opportunities for diagnosis of PHI are well recognized [5, 8]. Despite an increased awareness among the medical profession, a recent 24-year retrospective study found no significant improvement in the time delay in diagnosis of PHI [9].

The objective of this study was to examine the prevalence of HIV infection in patients presenting in primary care with GF-like illness to inform local health policy in incorporating HIV in the routine GF testing algorithm.

Methods

Guy's and St Thomas' Pathology Laboratories (GSTS) are part of the Guy's and St Thomas' Hospitals NHS Foundation Trust (GSTT), and provide virological testing for two teaching hospitals as well as primary care services within the inner London boroughs of Lambeth and Southwark Primary Care Trust.

Following local research ethics committee approval, samples from primary care submitted to GSTT for a GF screen between April 2009 and June 2010, with and without a concomitant HIV request, were identified. Samples without an HIV request were anonymized and retrospectively tested using a 4th-generation HIV antigen/antibody screening test. Reactive samples were further confirmed by an HIV antibody only test, with or without a p24 antigen assay. In conjunction with the HIV Reference laboratory at the Centre for Infection, Health Protection Agency, Colindale, antibody avidity testing based on the Recent HIV Infection Testing Algorithm (RITA) was used to identify individuals with evidence of recent acquisition (within 4–5 months).

Results

A total of 72 GP practices submitted GF screening requests during the study period. The average number of GF screen requests per practice was 15, with a median of 9 (and a range of 1–85). Thirty-two practices submitted 10 or more requests, and 18 practices submitted 20 or more requests.

Of 1046 primary care patients with GF screening requests, a concomitant HIV request was made in 119 patients (Fig. 1). One patient was known to be positive at the time of request and was excluded from the study results. Of the remaining 118 (118 of 1045; 11.3%), 2.5% (three of 118) were HIV positive. A further 45 (4.3%) had an HIV test requested subsequently through another consultation within 1 year; of these, 4.4% (two of 45) were HIV positive, and both were diagnosed through routine antenatal screening 6 and 8 months after the initial GF investigation.

Figure 1.

Chart showing the number of patients investigated for glandular fever-like illness, the number tested for HIV and the number of new HIV diagnosis.

Of the 882 patients with unknown HIV status, 694 (78.7%) had available samples for unlinked anonymous HIV testing; and from these, six further HIV-positive patients (0.9%) were identified. Thus, the overall prevalence of HIV infection in this patient group was 1.3% (11 of 857), with 72.7% (eight of 11) cases missed at the initial GP consultation.

Excluding the two patients found to be HIV positive following subsequent antenatal screening, four of the remaining nine patients (44.4%) were found to have evidence of recent acquisition based on the RITA testing algorithm, with three (75.0%) of these infections missed at the initial GP presentation. One further sample had an ‘invalid’ result because antibody levels were too low for the avidity test.

Discussion

Results indicate low levels of HIV testing in patients presenting in primary care with GF-like illness. Only 11.3% of patients presenting within our study period who received a GF screen also had a concomitant HIV test. As our study has demonstrated, this leads to a significant number of missed HIV diagnoses.

It is estimated that 24% of people living with HIV in the UK remained undiagnosed in 2010 [10]. With a diagnosed prevalence in Lambeth and Southwark of 1.39 and 1.13%, respectively [11], the undiagnosed prevalence in the two local authorities can be estimated as 0.4%. The overall positivity of 1.3% in our group presenting with GF-like symptoms is substantially higher than the estimated undiagnosed prevalence in the local population.

The prevalence of recent infections within our cohort (0.5%; four of 855) suggests a high prevalence of PHI within patients presenting with GF-like illness. The patient with an invalid RITA result because of low levels of antibody may represent a case of very recent acquisition. Diagnosis in a significant proportion of patients with evidence of recent acquisition (75.0%) was missed at what, for most, may be the only symptomatic presentation to healthcare services before more advanced disease years later.

Our study had several limitations. In our anonymized study we could not verify whether the 694 samples without concomitant HIV test requests were known HIV positives as all identifying laboratory information was removed as a condition for ethics approval. However, as almost half of the cases had symptoms and laboratory results consistent with PHI, the contribution of previous known positive cases is unlikely to be significant. Furthermore, we do not have data on the number of individuals who declined the offer of an HIV test. Local experience suggests that this is a relatively rare occurrence. Recent studies conducted by the Department of Health found that the uptake rate by patients is generally high – between 75 and 91% in London [12] and Brighton [13]. Lack of patient demographic data meant we could not identify groups with particularly high HIV prevalence, or particularly low rates of primary care requested HIV tests. Despite these limitations, we believe that data from this study make a compelling case for the exclusion of HIV infection in all patients who present with GF-like illness.

Late diagnosis means higher risk of poor response to treatment and increased mortality [14]. Onward transmission of HIV from infected individuals is more likely if the infected individual is unaware of their own infection [15]. The public health and clinical benefits are particularly relevant for diagnosis during PHI where viral load and thus infectivity are highest. Early diagnosis also provides an opportunity for maximizing the impact of recent partner notification.

We believe our results provide a strong economic case for including HIV in the standard GF screening tests. In our study, we carried out 694 additional HIV tests, and found three seropositive patients with evidence of recent acquisition (PHI). Assuming each test costs £10, the cost per diagnosis of PHI is £2310. The lifetime treatment cost of one patient is estimated to be around £280 000 to £360 000 [10]. Diagnosis of PHI represents a compelling economic argument for universal HIV testing in people presenting with GF-like illness. Formal cost-effectiveness studies have been conducted in the USA and France. In the USA, universal HIV testing is considered cost-effective if the positivity rate is greater than 1/1000 [16]. In France, a once-a-lifetime HIV test in the general population, and annual HIV tests in high-risk populations are considered cost-effective [17]. The UK national guidelines also recommend screening if diagnosed HIV prevalence exceeds 2 per 1000 population [18]. A prevalence of 1.3% in our GF cohort is well above the recommended threshold for routine screening.

Local policy should consider adopting the same opt-out strategy as in antenatal screening and include an HIV test routinely within the GF screening investigation panel.

Acknowledgements

We are grateful to Gary Murphy at the HIV Reference laboratory in the Centre for Infection, Health Protection Agency, Colindale for help with the RITA analysis.

Footnotes

  1. *Both diagnosed through Antenatal screening (6 and 8 months after initial primary care consultation)

Ancillary