A substantial proportion of individuals infected with HIV have experienced significant immunodeficiency or disease progression before they are diagnosed . Such individuals have a poor clinical prognosis [2,3] and may experience worse immunological, virological or both responses when they start HAART [4,5]. Earlier diagnosis of those infected with HIV may, therefore, lead to further reductions in morbidity and mortality, as well as a possible impact on onward transmission of HIV.
Globally, the majority of patients who initiate HAART do so at CD4 cell counts lower than recommended by current treatment guidelines , largely due to delayed diagnosis. At a European level, it has been estimated that anywhere between 10 and 45% of those diagnosed with HIV are diagnosed ‘late’ [3,7–11]. However, the definition of ‘late’ varies considerably from study to study, hampering attempts to make comparisons.
In the developed world where initiation of ART is recommended at a CD4 cell count below 350 cells/μl, it can reasonably be argued that presentation with a CD4 cell count below this, resulting in a delay in treatment initiation, is too late. However, people presenting with a CD4 cell count in the 300–350 cells/μl range, for example, are only moderately immunosuppressed and, if they remain untreated, do not have a high short-term risk of HIV-related death (<0.05% per year ), so there is a clear need for a definition which captures presentation at a stage when there is a substantial risk of death. We refer to this as presentation with ‘advanced HIV disease’. Definitions which have attempted to capture this have included use of CD4 cell count thresholds of 200 and 50 cells/μl [3,7,11,13] and/or whether there is presence of a clinical AIDS event, either at the time of diagnosis or within a short period after HIV diagnosis [14,15]. The aim of this study was to investigate the use of several common definitions of presentation with advanced HIV disease (which many have previously referred to as late presentation) with the aim of identifying a definition that can reliably identify a high proportion of individuals who will die shortly after HIV diagnosis.
Patients and methods
The UK Collaborative HIV Cohort (CHIC) study is a collaboration of some of the largest HIV clinics in the UK . Participating centres provide routinely collected data on all patients aged at least 16 years attending for care since 1996. The data collected includes demographic information, AIDS events, deaths, antiretroviral use, CD4 cell counts and HIV RNA levels; the current dataset includes information on 29 055 patients seen at 11 clinical centres up to the end of 2006. To ensure complete ascertainment of deaths even among patients lost to follow-up, UK CHIC data are linked anonymously to information from national UK death registers.
We identified all individuals who were seen for the first time at a participating clinic from 1 January 1996 to 31 December 2006. Individuals were excluded if there was any evidence of an earlier HIV diagnosis at another centre or if their first CD4 cell count was measured more than 1 year after HIV diagnosis. We compared two immunological (CD4 cell count < 200 cells/μl, CD4 cell count < 50 cells/μl) and two clinical (AIDS or severe/moderate AIDS at or within a month of HIV diagnosis) definitions of presentation with severe immunodeficiency and combinations of these. As our aim was to mimic analyses that are commonly presented in the literature, for all definitions that included immunological criteria, patients with missing CD4 cell counts were assumed not to have a count below the threshold of interest. AIDS events were classified as mild, moderate or serious as described by Mocroft et al. . The predictive ability [described by the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV)] of each definition for identifying individuals who died in the first 3 months after HIV diagnosis was assessed. Of note, 5103 (32.4%) of included patients did start antiretroviral treatment in the first 3 months after diagnosis, but these analyses ignored the use of HAART, as our aim was to identify patients at high risk of death regardless of treatment uptake or response.
A total of 15 774 patients were included (Table 1). These patients were followed in the UK CHIC study for a median (interquartile range) of 2.6 (0.7–5.4) years. Overall, 1495 (9.5%) of patients had a CD4 cell count below 50 cells/μl at diagnosis, with a further 2736 (17.3%) having a CD4 cell count from 50 to 199 cells/μl [giving 4231 (26.8%) with a CD4 cell count < 200 cells/μl]. Of note, CD4 cell counts at diagnosis were unavailable for 14.4% patients. A total of 1523 (9.7%) patients had AIDS at diagnosis, with 379 (2.4%) of these patients having severe or moderate AIDS. There was substantial overlap with the different criteria; for example, out of the 1523 patients with AIDS at diagnosis, 662 (43.4%) had a CD4 cell count below 50 cells/μl and 1185 (77.8%) had a CD4 count below 200 cells/μl. Similarly, of the 379 patients with severe/moderate AIDS, 174 (45.9%) and 299 (78.9%) had CD4 cell counts that were below 50 cells/μl or below 200 cells/μl, respectively.
Two hundred and six (1.3%) patients died in the first 3 months after diagnosis (with a further 23 patients experiencing a new serious AIDS event), 104 (2.5%), 67 (4.5%), 89 (5.8%) and 37 (9.8%) of those with a CD4 cell count below 200 cells/μl, a CD4 cell count below 50 cells/μl, AIDS or severe/moderate AIDS, respectively (Table 2, PPV). Importantly, 92 (44.7%) patients who died did not have a CD4 measurement prior to death, resulting in low sensitivities for definitions based on immunological criteria of 32.5% (CD4 cell count < 50 cells/μl) and 50.5% (CD4 cell count < 200 cells/μl). Definitions based on clinical criteria were not greatly better (AIDS, 43.2%; severe/moderate AIDS, 18.0%). As expected, given the low prevalence of clinical progression, PPVs were very low (2.5–9.8%) and NPVs were very high (all around 99%).
Definitions based on combinations of clinical and immunological criteria which classified patients as late presenters if either of the criteria were met, generally resulted in increased sensitivity than definitions based on each criterion alone, but reduced specificity. Sensitivity analyses, in which we considered clinical progression (death or a new serious AIDS event) in the first 3 months, reached similar conclusions (data not shown).
A common definition of late presentation would be advantageous for several reasons. First, such a definition would facilitate cross-country comparisons and would allow an ecological assessment of the potential impact of different public health interventions to encourage earlier HIV diagnosis. Second, a common definition would allow studies to identify risk factors for late presentation in a common way, and thus identify whether risk factors vary in different settings. Finally, a common definition would simplify analyses that monitor changes in the rate of late presentation over time.
Most studies of late presentation have focussed on people in a state where there is high risk of clinical progression and death. In almost all cases, this has led to definitions of late presentation which represent a much more advanced stage of infection than that at which initiation of ART is generally recommended. We recommend that the designation ‘late’ is reserved for presentation beyond the time that treatment is indicated, and thus propose that a late presenter is defined as any individual who presents with a CD4 cell count below 350 cells/μl or with a clinical AIDS event at, or within a month after, HIV diagnosis. However, in addition to this, there remains a need for a common definition which captures presentation at a time when the risk of death is substantial. Given our findings, we propose that any individual who presents with either a CD4 cell count below 200 cells/μl or a clinical AIDS event is defined as presenting with advanced HIV disease. The two definitions are complementary, address different needs and should be quoted when appropriate. Clearly, to apply either of these two definitions, studies and surveillance operations will be required to collect two pieces of information at the time of diagnosis: the CD4 cell count of the patient and whether an AIDS-defining event was present.
The choice of definition has a major impact on the proportion of individuals who are identified as presenting with advanced HIV disease. We compared several potential definitions for this, with the proportion of newly diagnosed individuals who are identified as such ranging from 2.4%, when the definition is based on the presence of a severe or moderate AIDS event, to 29.0% when a definition is based on the combination of an AIDS event and a CD4 cell count below 200 cells/μl. We found that in some situations, a definition that is based on immunological criteria alone may fail to identify all patients at highest mortality risk, simply because these patients may present and die without having a CD4 cell count measured. As the majority of these patients present with very advanced clinical disease consistent with an extremely low CD4 cell count, exclusion of such patients would lead to an underestimate of the proportion of patients who present late and a biased assessment of the association with mortality risk. More often, studies tend to include these patients in the denominator of any calculations but make the assumption that in the absence of a measured low CD4 cell count, these patients have counts that are not below the threshold of interest. We have tried to mimic this situation in our analyses. From a clinical perspective, there is a need to diagnose as many individuals with advanced HIV disease as possible so that appropriate care can be provided rapidly. Thus, a definition which has a high sensitivity is believed to be more appropriate in this setting. However, high sensitivity is often achieved at the expense of high specificity; the specificity for the proposed definition, though reasonably high (72%), was lower than that for definitions based on immunological or clinical criterion alone.
For our analyses, we have focused on clinical events that occur over the first 3 months after diagnosis. This timepoint was chosen to ensure that the outcome reflected the health status of the individual at the time of diagnosis, rather than capturing the effect of any treatment or management decisions that were made subsequently. Mortality risk is highest in the first 3 months after diagnosis; thereafter, deaths that occur generally reflect failures in patient care, poor adherence to HAART or both.
It should be noted that although the UK CHIC study is generally representative of patients seen for HIV care in the UK, our results may not necessarily be generalizable to all settings and, as such, validation of our findings in other studies would be desirable.
The UK CHIC study is funded by the Medical Research Council, UK (grant #G0000199 and G0600337). The views expressed in this manuscript are those of the researchers and not necessarily those of the Medical Research Council.
Writing committee: Caroline A. Sabin, Achim Schwenk, Margaret A. Johnson, Brian Gazzard, Martin Fisher, John Walsh, Chloe Orkin, Teresa Hill, Richard Gilson, Kholoud Porter, Philippa Easterbrook, Valerie Delpech, Loveleen Bansi, Clifford Leen, Mark Gompels, Jane Anderson and Andrew N. Phillips.
UK CHIC Steering Committee: Jonathan Ainsworth, Jane Anderson, Abdel Babiker, Valerie Delpech, David Dunn, Philippa Easterbrook, Martin Fisher, Brian Gazzard (Chair), Richard Gilson, Mark Gompels, Teresa Hill, Margaret Johnson, Clifford Leen, Chloe Orkin, Andrew Phillips, Deenan Pillay, Kholoud Porter, Caroline Sabin, Achim Schwenk and John Walsh.
Central co-ordination: Research Department of Infection and Population Health, UCL Medical School, London (Loveleen Bansi, Teresa Hill, Andrew Phillips and Caroline Sabin); Medical Research Council Clinical Trials Unit (MRC CTU), London (David Dunn, Adam Glabay and Kholoud Porter).
Participating centres: Barts and The London NHS Trust, London (Chloe Orkin, Kevin Jones and Rachel Thomas); Brighton and Sussex University Hospitals NHS Trust (Martin Fisher, Nicky Perry, Anthony Pullin and Duncan Churchill); Chelsea and Westminster NHS Trust, London (Brian Gazzard, Steve Bulbeck, Sundhiya Mandalia and Jemima Clarke); Health Protection Agency Centre for Infections London (Valerie Delpech); Homerton University Hospital NHS Trust, London (Jane Anderson and Sajid Munshi); King's College Hospital, London (Philippa Easterbrook, Frank Post, Yasar Khan, Paragi Patel, Fatimah Karim and Stephen Duffell); Medical Research Council Clinical Trials Unit (MRC CTU), London (Abdel Babiker, David Dunn, Adam Glabay and Kholoud Porter); UCL Medical School and The Mortimer Market Centre, London (Richard Gilson, Shuk-Li Man and Ian Williams); North Bristol NHS Trust (Mark Gompels and Debbie Dooley); North Middlesex University Hospital NHS Trust, London (Achim Schwenk and Jonathan Ainsworth); Royal Free NHS Trust and Department of Infection & Population Health, UCL Medical School, London (Margaret Johnson, Mike Youle, Fiona Lampe, Colette Smith, Helen Grabowska, Clinton Chaloner, Dewi Ismajani Puradiredja, Loveleen Bansi, Teresa Hill, Andrew Phillips and Caroline Sabin); Imperial College Healthcare NHS Trust, London (John Walsh, Nicky Mackie, Alan Winston, Jonathan Weber, Christian Kemble and Mark Carder); The Lothian University Hospitals NHS Trust, Edinburgh (Clifford Leen and Alan Wilson).
1. Fisher M. Late diagnosis of HIV infection: major consequences and missed opportunities. Curr Opin Infect Dis 2008; 21:1–3.
2. Egger M, May M, Chene G, Phillips AN, Ledergerber B, Dabis F, et al
. Prognosis of HIV-1-infected patients starting antiretroviral therapy: a collaborative analysis of prospective studies. Lancet 2002; 360:119–129.
3. Chadborn TR, Baster K, Delpech VC, Sabin CA, Sinka K, Rice BD, et al
. No time to wait: how many HIV-infected homosexual men are diagnosed late and consequently die? (England and Wales, 1993–2002). AIDS 2005; 19:513–520.
4. Moore RD, Keruly JC. CD4+
cell count 6 years after commencement of highly active antiretroviral therapy in persons with sustained virologic suppression. Clin Infect Dis 2007; 44:441–446.
5. Phillips AN, Staszewski S, Weber R, Kirk O, Francioli P, Miller V, et al
. HIV viral load response to antiretroviral therapy according to the baseline CD4 cell count and viral load. JAMA 2001; 286:2560–2567.
6. Egger M. Outcomes of ART in resource-limited and industrialized countries
. 14th Conference on Retroviruses and Opportunistic Infections
; 25–28 February 2007; Los Angeles, USA.
7. Wolbers M, Bucher HC, Furrer H, Rickenbach M, Cavassini M, Weber R, et al
. Delayed diagnosis of HIV infection and late initiation of antiretroviral therapy in the Swiss HIV Cohort Study. HIV Med 2008; 9:397–405.
8. Borghi V, Girardi E, Bellelli S, Angeletti C, Mussini C, Porter K, et al
. Late presenters in an HIV surveillance system in Italy during the period 1992–2006. J Acquir Imm Defic Syndr 2008; 49:282–286.
9. Lanoy E, Mary-Krause M, Tattevin P, Perbost I, Poizot-Martin I, Dupont C, et al
. Frequency, determinants and consequences of delayed access to care for HIV infection in France. Antivir Ther 2007; 12:89–96.
10. Brannstrom J, Akerlund B, Arneborn M, Blaxhult A, Giesecke J. Patients unaware of their HIV infection until AIDS diagnosis in Sweden 1996-2002: a remaining problem in the highly active antiretroviral therapy era. Int J STD AIDS 2005; 16:702–706.
11. Sabin CA, Smith CJ, Gumley H, Murphy G, Fiona FC, Phillips AN, et al
. Late presenters in the era of highly active antiretroviral therapy: uptake of and responses to highly active antiretroviral therapy. AIDS 2004; 18:2145–2151.
12. Phillips AN, Pezzotti P, CASCADE Collaboration. Short-term risk of AIDS according to current CD4 cell count and viral load in antiretroviral drug-naive individuals and those treated in the monotherapy era
13. Sullivan AK, Curtis H, Sabin CA, Johnson MA. National review of newly diagnosed HIV infections. BMJ 2005; 330:1301–1302.
14. Delpierre C, Lauwers-Cances V, Pugliese P, Poizot-Martin I, Billaud E, Duvivier C, et al
. Characteristics trends, mortality and morbidity in persons newly diagnosed HIV positive during the last decade: the profile of new HIV diagnosed people. Eur J Pub Health 2007; 18:345–347.
15. Torrone EA, Thomas JC, Leone PA, Hightow-Weidman LB. Late diagnosis of HIV in young men in North Carolina. Sex Transm Dis 2007; 34:846–848.
16. The UK Collaborative HIV Cohort Steering Committee. The creation of a large UK-based multicentre cohort of HIV-infected individuals: the UK Collaborative HIV Cohort (UK CHIC) study
. HIV Med
17. Mocroft A, Sterne JA, Egger M, May M, Grabar S, Furrer H, et al
. Variable impact on mortality of AIDS-defining events diagnosed during combination antiretroviral therapy: not all AIDS-defining conditions are created equal. Clin Infect Dis 2009; 48:1138–1151.