Abstract
Delayed diagnosis of HIV infection has negative clinical, economic and public health implications. The study primary aim was to identify factors associated with late HIV presentation (Late Presenters [LPS], CD4 cell count < 350 cells/mm3). A secondary aim was to identify changing trends of late HIV presentation from 2002 to 2014 at our centre. A retrospective cohort study was performed. Demographic data and CD4 cell count of new HIV diagnoses presenting to our ambulatory HIV service over four time-periods from 2002 to 2014 were recorded. Proportion of LPS and factors associated with late presentation were compared using Graphpad Instat. In 2014, of 231 new patients attending for HIV care, 75 (32.6%) were late presenters versus 146 (66.4%) in 2002. This indicates a decreasing proportion of LPS from 2002 to 2014. However, the proportion of those with CD4 cell counts <200 on presentation at these two time intervals remain unchanged. The overall proportion of male LPS has increased over time and the proportion of LPS in the men who have sex with men (MSM) cohort has decreased over time, reflecting increased frequency of both HIV testing and diagnoses in MSM in recent years. The proportion of heterosexual LPS has not changed significantly in the same time period and LPS were older in 2014 versus 2002. The proportion of LPS defined by CD4 cell count remains higher than is justifiable in an era of increased HIV testing and awareness. Further targets for HIV testing to decrease rates of LPS include non-traditional risk groups including heterosexual and older patient cohorts. LPS rates are lower than rates found internationally, and it is possible that consensus definition of LPS needs to be revised.
Keywords
The WHO estimates that worldwide, 37 million people are living with HIV, and now recommend anti-retroviral therapy for all those with HIV infection. It is estimated that currently worldwide, only 54% of people are aware of their infection status. 1
Late HIV presentation can be defined as those presenting for HIV care with a CD4 cell count of less than 350 cells/mm3. 2 Late presentation with HIV infection has a negative impact on outcome for the individual. Those who are unaware of their positive HIV status cannot benefit from widely available treatment options. As a result, late presentation for HIV care is associated with higher morbidity and mortality, even after treatment with antiretroviral therapy.3,4 Late presenters also carry a lower chance of recovery of CD4+ T lymphocytes (CD4) 5 following treatment, and without treatment carry a risk of HIV transmission while unaware of their HIV status. Furthermore, costly inpatient hospital admissions may be avoided with appropriate management of HIV at an earlier point.
In 2014, 377 new HIV diagnoses were notified in Ireland, giving a crude notification rate of 8.2 per 100,000 population. This is an increase of 11% compared with 2013 and can be accounted for by an increasing number of HIV notifications among men who have sex with men (MSM) and people who inject drugs (PWID). Between 2010 and 2013, the annual rate of new HIV diagnoses had been relatively stable in Ireland, ranging from 7.0 to 7.5 per 100,000 population.
The highest number of new diagnoses ever reported in MSM was in 2014, comprising 49% of new HIV diagnoses in Ireland. In the 10 years since 2005, the number of new diagnoses among MSM has increased threefold (from 60 to 183) and the median age at diagnosis has dropped from 37 to 31 years. Furthermore, in 2014, there were 27 (7%) new diagnoses among PWID, the highest number reported in this risk group since 2009.
In Ireland, the Health Protection and Surveillance Centre (HPSC) report that 49% of new HIV diagnoses in 2014 were late presenters, with CD4 less than 350 cells/μl or an AIDS defining illness at diagnosis. This number reflects similar proportions over recent years. Late presentation was less common among MSM (38%) and PWID (44%) than among heterosexuals (56% in females and 71% in males). Regular testing among MSM and PWID is likely to be a major reason for this as both groups had a much higher proportion reporting ever having had a previous negative test (58% in MSM and 66% in PWID versus 20% in heterosexuals). However, the lower proportion of late diagnoses among MSM and PWID may also be a result of more recently acquired infections in these population groups. Based on previous HIV testing history, 27% of infections in MSM, and 41% in PWID were acquired in the previous 2 years. 6
The primary aim of this study was to identify factors associated with late HIV presentation (LPS, CD4 cell count < 350 cells/mm3) and to further examine differences between those with moderate immunodeficiency (MI, CD4 200–350 cells/mm3) and advanced immunodeficiency (CD4 < 200 cells/mm3). We sought to determine opportunities missed to diagnose HIV earlier in this patient cohort. A secondary aim was to identify changing trends of late HIV presentation from 2002 to 2014 at our centre.
Methods
Study design and population
This was a retrospective cohort study. Appropriate in-hospital approval for research activity was obtained (ref 2015/109). Retrospective electronic chart review was undertaken. Patients with a new diagnosis of HIV infection who presented for care at our HIV centre over 2014, 2012, 2007 and 2002 with newly diagnosed HIV infection were included. Time intervals were chosen to reflect the possible changing demographics of our cohort over time. Data collection had previously taken place at our department for 2002, 2007 and 2012. Following on from this, data were then collected for the most recent year that had passed (2014). Patients who had already received care for their HIV infection at another centre were excluded from the study.
Variable of interest
The primary variable of interest was CD4 cell count (cells/mm3) at presentation. Patients were identified by a departmental patient database; these were new attenders for HIV care over defined time-points (2014, 2012, 2007 and 2002). By electronic chart review, various demographic data including age, gender, risk group, geographic origin, previous HIV testing status, previous investigation of unexplained symptoms prior to HIV diagnosis, clinical indicator condition at time of diagnosis and CD4 cell count at presentation of new HIV diagnoses over 2014 were collected on Excel Database. HIV transmission risk group was categorised as MSM, IDU, heterosexual and other/unknown. Using data available, further comparison was then made between variables of this group and those of late presenters from three previous time-points over the past decade. (2002, 2007 and 2012). These variables included age, gender, mode of acquisition, country of origin and CD4 cell count at presentation. Data were coded anonymously and statistical analysis was performed using Graphpad Instat. Wilcoxon, ANOVA and χ2 tests were used to compare variables. Late presentation (LPS) was defined as those who presented with a CD4 cell count of <350 cells/mm3. Moderate immunodeficiency (MI) was defined with an initial CD4 cell count between 200 and 350 cells/mm3 and advanced immunodeficiency (AI) was defined as an initial CD4 cell count of <200 cells/mm3.
Results
In 2014, a total number of 231 patients were referred for management of newly diagnosed HIV infection. Seventy-five (32.6%) patients presented with a CD4 cell count of <350 cells/mm3 (LPS). Of these, 55 (73.3%) were male and 20 (26.7%) were female. Mean (SD) age at presentation was 39.7 (11) years. The majority of patients (n = 33, 44%) were Irish. Mean (SD) CD4 cell count was 166 (109) cells/mm3. See Figure 1 for information on probable route of transmission and geographic origin and see Table 1 for detailed demographic information for each presentation stage.
LPS by mode of acquisition and country of origin. Characteristics of First HIV Presenters to an Ambulatory HIV service from 2002–2014.
Moderate immunodeficiency in 2014 (n = 32)
The majority of patients (n = 23, 72%) were tested in healthcare settings outside our hospital. Seventeen (53.1%) patients had previously been tested for HIV infection. Eight (50%) of these patients had a HIV test within 2 years prior to HIV diagnosis and 7 (88%) of these were MSM.
Advanced immunodeficiency in 2014 (n = 43)
Late diagnosis 2014 demographics.
Comparison of groups over time
Proportions of male LPS compared with female LPS have increased over time (p ≤ 0.001). With this, proportions of MSM compared with other risk groups diagnosed over time have increased (p ≤ 0.001) (Figure 1). A decreased proportion of those with heterosexual risk in the LPS group was seen over time, however in 2014, the risk group most likely to present as LPS was heterosexual: HS (58%) vs. MSM (22.8%). Proportions of LPS from Sub-Saharan Africa in comparison with other geographic origins have decreased from 2002 to 2014 (p ≤ 0.0001). In contrast to this, proportions of LPS attending from South America have increased significantly from 2002 to 2014 (p ≤ 0.001). As rates of those from Sub-Saharan Africa have fallen, another significant proportion (44%) of LPS with heterosexual mode of acquisition in 2014 was from Ireland. No overall significant age difference was seen amongst LPS over time (p = 0.593). However, a direct comparison between 2002 and 2014 shows a significant difference in age, where LPS are presenting at an older age in 2014 than in 2002 (p ≤ 0.001) (Table 1).
Risk of acquisition of HIV.
The overall proportion of LPS newly diagnosed with HIV attending our centre for care has decreased over time (p ≤ 0.001). However, the proportion of those who presented with advanced immunodeficiency (AI) over time has not significantly changed (p = 0.69) (Figure 3).
Stages of first presenters to HIV service.
Rate of change over time (% of each category who presented with CD4 cell count <350).
Discussion
Over time, proportions of those with newly diagnosed HIV-1 infection presenting at a late stage in their illness are decreasing. Proportions of those with very late presentation presenting over the past 12 years, with a CD4 cell count of <200 cells/mm3, remain unchanged. The demographics of those with late HIV presentation are changing over time. Those with heterosexual mode of acquisition, older patients and those from South America are more likely to present with late HIV infection in recent years.
Despite current HIV prevention strategies in place, including earlier introduction of ART, the widespread availability of post-exposure prophylaxis and ED screening interventions, our findings show that a significant portion of those presenting with HIV continue to present at a late point in their illness, with a subsequent negative impact on overall morbidity and mortality. These data have shown that in 2014, a number of patients presented to a range of healthcare facilities for investigation of unexplained symptoms that later were diagnosed as HIV clinical indicator illness, without having a HIV test done, including medical and surgical outpatients, general practitioners and Emergency Departments. High rates of missed opportunities for HIV diagnosis have previously been reported; one US study reports 35% of individuals newly diagnosed with HIV infection had previously attended a healthcare provider over the preceding 1 year with HIV associated signs and symptoms. 7 The missed opportunities identified in this study further highlights the ongoing need for appropriate physician education on HIV clinical indicant illness, the need for HIV testing and the removal of barriers to performing a HIV test for patients.
Comparison between those presenting with moderate and advanced immunodeficiency in 2014 shows a significantly smaller number of those with advanced immunodeficiency had previously undergone HIV testing. At least 50% of patients in both groups had previously tested for HIV in the 2-year period prior to testing positive; the majority (69%) of these were MSM, reflecting high rates of HIV testing in this risk group. Early HIV infection with low CD4 cell count needs to be considered as a possible explanation for higher rates of LPS MSM diagnoses with higher testing rates, and suggests the definition for late HIV presentation needs to be revised. It also highlights the role for newer laboratory techniques to be developed and evaluated to distinguish recent and established HIV infection.
When comparison was made between patterns of demographics of those presenting in 2014 and over three other time points since 2002, it is clear that currently the rate of LPS in all risk groups is decreasing, however similar proportions of those with advanced immunodeficiency continue to present. Furthermore, people are likely to be older when presenting as LPS. These findings reflect those that have been found at a national level. 8
Rates of LPS in our cohort are lower than that found internationally. The Cohere Observational Cross-European study of LPS from 34 countries showed a LPS rate of 48.7% in 2013, with overall highest rates of LPS from 2010 to 2013 in heterosexual males and females. 9 While risk factors for late HIV presentation seen in this study are similar to our cohort, rates of LPS are higher than seen in our cohort. A possible explanation for this is that AIDS defining illness was not included as a definition for LPS in our cohort. Furthermore, a 2015 meta-analysis of gender differences at risk of late HIV presentation showed male gender to be at higher odds of late HIV presentation, similar to our cohort. 10 A review of HIV presentations in the United States over a 10-year time period, with LPS defined as CD4 cell count of <350 only, showed rates of LPS decreased over this time period, with LPS rates of 53% seen most recently in 2007. The effect of the use of earlier ART in the US cohort in more recent times remains to be seen. 11
Various intervention strategies have been employed in recent years to reduce the rates of late HIV diagnosis, missed opportunities for HIV testing and the detrimental sequelae of such an outcome. In the United States, to accelerate progress toward reducing undiagnosed HIV infection, the CDC and its partners have pursued an approach that includes expanding HIV testing in communities with high HIV infection rates. 12 It has previously been shown that barriers amongst healthcare providers exist, including lack of time to conduct testing and staff feeling ill-prepared to answer patient queries.13,14 It is possible that our patients did not have a HIV test previously for these reasons.
Limitations
This study has a number of limitations that are inherent with a retrospective cohort study. We were unable to ascertain the true timing of HIV acquisition in our patients. As late presenters were categorised on the basis of CD4 cell count only, it is likely that a proportion of these patients included in the data analysis had recently acquired infection. Because of limitations with retrospective chart review, the authors did not use the LP consensus definition, 15 where AIDS defining illness was not included in our definition of LPS. As a result of this, it is possible that rates of LPS are lower in our cohort than seen internationally.
Conclusions
Potential targets for future HIV testing include testing initiatives at other healthcare settings and GP practices, where a large proportion of those LPS were diagnosed. In line with higher rates of HIV diagnosis in MSM, the proportion of MSM LPS as compared with other risk groups is increasing. A significantly smaller number of MSM were diagnosed with advanced immunodeficiency, representing the higher frequency of HIV tests performed for this risk group. With this in mind, our knowledge that the heterosexual risk group now most likely to present as LPS in 2014 and high rates of recent previous testing in the MSM group, it is now evident that other risk groups need to be targeted for HIV testing. Given the increasing age of patients who are presenting with late HIV, all age and risk groups should be considered for a HIV test. To employ these interventions effectively, widespread HIV testing in all healthcare facilities including GP practices needs to be implemented. Furthermore, the need to re-structure the education of healthcare providers, to re-emphasise the need for HIV clinical suspicion in all risk and age groups when presenting for care and to increase the awareness of the role for widespread HIV testing in areas of different prevalence and demography has been highlighted. While blood borne virus screening is now standard of care at the emergency medicine department at our centre, further strategies needs to be employed to diagnose and treat HIV early, to prevent onward HIV transmission and avoid adverse healthcare consequence.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.