Starting HAART in a very advanced stage of disease is assumed to be the most prevalent form of initiation in HIV-infected subjects in developing countries. Data from Latin America and the Caribbean is still lacking. Our main objective was to determine the frequency, risk factors and trends in time for being late HAART initiator (LHI) in this region.
Cross-sectional analysis from 9817 HIV-infected treatment-naïve patients initiating HAART at 6 sites (Argentina, Chile, Haiti, Honduras, Peru and Mexico) from October 1999 to July 2010. LHI had CD4+ count ≤200cells/mm3 prior to HAART. Late testers (LT) were those LHI who initiated HAART within 6 months of HIV diagnosis. Late presenters (LP) initiated after 6 months of diagnosis. Prevalence, risk factors and trends over time were analyzed.
Among subjects starting HAART (n = 9817) who had baseline CD4+ available (n = 8515), 76% were LHI: Argentina (56%[95%CI:52–59]), Chile (80%[95%CI:77–82]), Haiti (76%[95%CI:74–77]), Honduras (91%[95%CI:87–94]), Mexico (79%[95%CI:75–83]), Peru (86%[95%CI:84–88]). The proportion of LHI statistically changed over time (except in Honduras) (p≤0.02; Honduras p = 0.7), with a tendency towards lower rates in recent years. Males had increased risk of LHI in Chile, Haiti, Peru, and in the combined site analyses (CSA). Older patients were more likely LHI in Argentina and Peru (OR 1.21 per +10-year of age, 95%CI:1.02–1.45; OR 1.20, 95%CI:1.02–1.43; respectively), but not in CSA (OR 1.07, 95%CI:0.94–1.21). Higher education was associated with decreased risk for LHI in Chile (OR 0.92 per +1-year of education, 95%CI:0.87–0.98) (similar trends in Mexico, Peru, and CSA). LHI with date of HIV-diagnosis available, 55% were LT and 45% LP.
LHI was highly prevalent in CCASAnet sites, mostly due to LT; the main risk factors associated were being male and older age. Earlier HIV-diagnosis and earlier treatment initiation are needed to maximize benefits from HAART in the region.
Citation: Crabtree-Ramírez B, Caro-Vega Y, Shepherd BE, Wehbe F, Cesar C, Cortés C, et al. (2011) Cross-Sectional Analysis of Late HAART Initiation in Latin America and the Caribbean: Late Testers and Late Presenters. PLoS ONE 6(5): e20272. doi:10.1371/journal.pone.0020272
Editor: Beatriz G. J. Grinsztejn, Instituto de Pesquisa Clínica Evandro Chagas/Fundação Oswaldo Cruz, Brazil
Received: March 2, 2011; Accepted: April 17, 2011; Published: May 26, 2011
Copyright: © 2011 Crabtree-Ramírez et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was supported by the National Institutes of Health Cooperative Agreement (1-U01-AI069923). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: BCR: speaker for BMS and MSD and grant for clinical trial of MSD; C. Cortés: speaker for MSD and GSK; EG: received grants from Pfizer and MSD; PC: board membership of MSD, GSK, Pfizer, Tibotec, Shering Plough, and speaker for Abbot, Tibotec, BMS, GSK, MSD, Pfizer and Boeringer Ingelheim; JSM: speaker for ViiV and Stendahl and has received grant for research from Pfizer. Authors YCV, BES, FW, C. Cesar, DP, SK, CM, and DM have no conflict of interest.
Since the introduction of highly active antiretroviral therapy (HAART) for the treatment of HIV infection, the benefits in survival have been clearly established . The stage of HIV infection at the time of HAART initiation plays an important role in patients' prognosis after treatment. Patients with advanced disease when starting HAART are less likely to achieve virological suppression, more likely to modify their therapy due to adverse events, have a higher mortality rate and represent a financial strain in public health services, as compared with those who initiate earlier –. Therefore, in recent years there has been a push towards earlier initiation of HAART to improve survival and decrease AIDS complications –. Furthermore, earlier HAART initiation may result in a lower risk of HIV transmission, as suppressing HIV-1 RNA levels has been shown to diminish transmission rates , . As a result, early diagnosis and treatment of HIV constitutes a major public health issue.
Despite efforts to improve the care of people living with HIV, a considerable proportion of subjects do not obtain medical attention until they are in very advanced stages of the disease , –. Several studies conducted in high income countries in the past decade found that 20% to 40% of patients diagnosed with HIV were in advanced disease stages (in most studies defined as presenting with CD4 cell counts less than 200 cells/mm3 and/or a history of an AIDS-defining illness). Consequently, these patients were late HAART initiators . Late HAART initiation can be attributed to late HIV diagnosis due to late testing (LT), or to late presentation (LP) to care after diagnosis has been established. Although late testing and late presentation may have similar consequences for the prognosis of individual patients when starting HAART –, differences between LT and LP may reflect failures at different steps of health care access . In Latin America and the Caribbean, studies show high frequencies of late stage of disease at HAART initiation –, –. However, risk factors associated with LHI have not been well described. The aims of this study were to evaluate the proportion of patients who were LHI between the years of 2000 and 2010, to assess trends across the study period, to look for risk factors associated with LHI, and to determine if LHI was due to late diagnosis or late presentation using data from the Caribbean, Central and South American network for HIV Research (CCASAnet) .
The CCASAnet cohort (http://ccasanet.vanderbilt.edu/) has been described elsewhere –, . The collaboration was established in 2006 as Region 2 of the International Epidemiologic Databases to Evaluate AIDS (IeDEA; www.iedea-hiv.org) with the purpose of collecting HIV data from Central and South America and the Caribbean to be able to describe the unique characteristics of the HIV epidemic in this region. Six CCASAnet sites contributed data to this study: Fundación Huésped in Buenos Aires, Argentina (FH-Argentina); Fundación Arriarán in Santiago, Chile (FA-Chile); Le Groupe Haïtien d'Etude du Sarcome de Kaposi et des Infections Opportunistes in Port-au-Prince, Haiti (GHESKIO-Haiti); Instituto Hondureño de Seguridad Social and Hospital de Especialidades in Tegucigalpa, Honduras (IHSS/HE-Honduras); El Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán in Mexico City, Mexico (INNSZ-Mexico); and El Instituto de Medicina Tropical Alexander von Humboldt in Lima, Perú (IMTAvH-Peru). Institutional review board approval was obtained locally for each participating site and for the CCASAnet data coordinating center (DCC) at Vanderbilt University, Nashville, TN, USA.
Study design and population
The source population for this cross-sectional analysis was HIV-infected adults receiving medical care at the 6 participating CCASAnet sites. Treatment-naïve patients who started HAART between October 1999 and July 2010 were included. HAART was defined in all sites as all regimens with at least 3 antiretroviral drugs (mainly composed with 2 nucleoside/nucleotide analogs of reverse transcriptase inhibitors plus a no-nucleoside/nucleotide analogs of reverse transcriptase inhibitors or a protease inhibitor). All participant sites are part of local programs for universal free access to HAART. Late HAART initiators (LHI) were defined as those subjects with a non-missing baseline CD4 cell count (6 months before or 1 week after starting HAART) who either had baseline CD4≤200 cell/mm3 or an AIDS-defining illness (ADI) prior to or at HAART initiation. LHI were further classified as either late testers (LT) or late presenters (LP), corresponding to those who started HAART within 6 months of HIV diagnosis (LT) or more than 6 months after HIV diagnosis (LP). The remaining patients were considered non-late HAART initiators (NLHI). All study definitions were chosen a priori.
Each participant site contributed data from their database using a standard data transfer protocol. Demographic, clinical, and laboratory data for all sites were prospectively collected and periodically recorded. The databases are updated routinely and sent to the DCC. Data were checked for errors and inconsistencies.
We collected data on sex, age at HAART initiation, probable route of infection (self-reported as sexual – either homosexual or heterosexual intercourse – or non-sexual), marital status (only available in GHESKIO-Haiti, INNSZ-Mexico and IMTAvH-Peru), employment status (employed/unemployed), and years of education at first clinic visit. In sites with available data (FH-Argentina and INNSZ-Mexico), sexual probable route of infection was further classified as homosexual or heterosexual. Baseline CD4 cell count was defined as the measurement closest to HAART initiation but not more than 6 months prior to, or 7 days after, the date of HAART start. Values not available within this time frame were considered missing.
Trends in the proportion of LHI across the study period were assessed non-parametrically using Friedman's ‘super smoother’ . The association between LHI and risk factors was evaluated using univariate and multivariate logistic regression models. The primary multivariate analysis included age, sex, years of education, and date of HAART initiation – variables which were measured in all sites. Age and years of education were included as continuous variables. Secondary, site-specific multivariate models also included probable route of infection, marital status, and employment status, when available. Date of HAART initiation was included in the models using restricted cubic splines to account for non-linear time-trends . Combined-site estimates were computed using random effects methods , . Risk factors for LT and LP were assessed using multinomial logistic regression. In primary analyses, patients missing baseline CD4 cell count (LHI analyses) and baseline CD4 cell count or date of HIV diagnosis (LT/LP analyses) were excluded. Analyses were repeated incorporating data from all patients using multiple imputation techniques . Statistical analyses were performed using STATA version 10 and R version 2.11.1.
A total of 9817 subjects started HAART during the study period. Patient characteristics stratified by site and combined are shown in Table 1 and Table S1 (Supporting Information section). The median age at HAART initiation was 36 years (interquartile range 30–43); 40% of patients were female, ranging from 55% in GHESKIO-Haiti to 11% in FA-Chile. The probable route of HIV infection was sexual for the majority of patients; only FH-Argentina had a substantial portion of patients whose probable route of infection was non-sexual. The mean years of education at first clinic visit was 8.6 (Standard Deviation 4.9) ranging from 6.1 years in GHESKIO-Haiti to 13.1 in FA-Chile. Twenty percent of patients were married, although marital status was only available in GHESKIO-Haiti, IHSS/HE-Honduras, INNSZ-Mexico and IMTAvH-Peru. Employment status at entrance to care was available in all sites except GHESKIO-Haiti; 73% of the cohort was employed. The median CD4 cell count at HAART initiation was 129 cells/mm3 (IQR 49–219); baseline CD4 cell count was missing for 1302 subjects (13%). The median time between HIV diagnosis and HAART initiation was 153 days (5.1 months; IQR 38–712 days); date of diagnosis was missing for 1836 (18.7%) subjects. The time period of the study differed between sites, although all sites contributed data from 2003–2008.
Rates of LHI
Of 8515 patients with a recorded baseline CD4 cell count, 6431 (76% [95%CI: 74–77]) were LHI: 56% [95%CI: 52–59] in FH-Argentina, 80% [95%CI: 77–82] in FA-Chile, 76% [95%CI: 74–77] in GHESKIO-Haiti, 91%[95%CI: 87–94] in IHSS/HE-Honduras, 79% [95%CI: 75–83]) in INNSZ-Mexico and 86% [95%CI: 84–88] in IMTAvH-Peru. At HAART initiation, 70.4% of patients had CD4<200 (with or without ADI), 30.5% had an ADI (regardless of CD4+ count), and 25.6% had both characteristics. The proportion of LHI over time for each site is shown in Figure 1. For all of the sites except IHSS/HE-Honduras, the proportion of LHI statistically changed over time (p<0.001 for FH-Argentina, FA-Chile, GHESKIO-Haiti, and IMTAvH-Peru; p = 0.02 for INNSZ-Mexico; and p = 0.86 for IHSS/HE-Honduras). Rates of LHI tended to be lower in more recent years for most of the sites. The rates of LHI in INNSZ-Mexico and GHESKIO-Haiti increased in earlier years followed by a decrease in later years. In contrast, in FH-Argentina the rate of LHI decreased in earlier years and stayed roughly constant from 2004 to 2009. Note that 95.2% of patients started HAART with CD4<350 or an ADI.
NOTE. FH-Argentina, Fundación Huésped in Buenos Aires, Argentina; FA-Chile, Fundación Arriarán in Santiago, Chile; GHESKIO-Haiti, Le Groupe Haïtien d'Etude du Sarcome de Kaposi et des Infections Opportunistes in Port-au-Prince, Haiti; IHSS/HE-Honduras Instituto Hondureño de Seguridad Social and Hospital de Especialidades in Tegucigalpa, Honduras; INNSZ-Mexico, El Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán in Mexico City, Mexico; and IMTAvH-Peru, El Instituto de Medicina Tropical Alexander von Humboldt in Lima, Perú.
Risk factors for LHI
In multivariate logistic analyses including age, sex, years of education, and date of HAART initiation, male sex was associated with an increased risk of LHI in FA-Chile, GHESKIO-Haiti, IMTAvH-Peru, and in the combined site analyses (OR = 2.93, 95% CI 1.59–5.40; OR = 1.27, CI 1.06–1.51; OR = 2.28, CI 1.64–3.18; OR = 1.69, CI 1.24–2.30; respectively; Table 2). Older patients were more likely to be LHI in FH-Argentina and IMTAvH-Peru (OR 1.21 per 10 year increase in age, 95% CI 1.02–1.45; OR 1.20, CI 1.02–1.43; respectively), although this trend was not seen across all cohorts (OR 1.07, 95% CI 0.94–1.21). Higher education was statistically associated with a decreased risk of LHI in FA-Chile (OR 0.92 per additional year of education, CI 0.87–0.98), and similar trends, although not statistically significant, were seen in INNSZ-Mexico, IMTAvH-Peru, and in the combined site analysis. The estimated probability of LHI over time after adjusting for age, sex, and education was similar to the unadjusted trends seen above, except perhaps in FH-Argentina where there appeared to be a stronger recent trend to lower rates of LHI (see Figure S1 of Supporting Information section).
In secondary multivariate analyses that included other predictors which were recorded at specific sites, employment was associated with lower rates of LHI in FH-Argentina (OR 0.50, 95% CI 0.27–0.92), but not in FA-Chile, IHSS/HE-Honduras, INNSZ-Mexico, or IMTAvH-Peru (p>0.1 for all) (see Table S2 of Supporting Information section). In FH-Argentina, men who have sex with men (MSM) and those whose probable route of infection was heterosexual tended to have lower rates of LHI than those whose probable route of infection was non-sexual (OR = 0.49, 95% CI 0.24–1.00; OR = 0.66, 95% CI 0.32–1.35; respectively); in INNSZ-Mexico, MSM also tended to have lower rates of LHI than heterosexuals, although this was not statistically significant (OR = 0.81, 95% CI 0.41–1.58). Being married had little association with LHI in any of the sites that recorded marital status (OR 0.92, 95% CI 0.73–1.16, p = 0.48; OR 0.58, 95% CI 0.29–1.17, p = 0.13; OR 0.76, 95% CI 0.53–1.07, p = 0.11 for GHESKIO-Haiti, INNSZ-Mexico, and IMTAvH-Peru, respectively).
Rates of LT and LP
Of 6047 LHI with a recorded date of HIV diagnosis, 3331 (55%) were late testers and 2716 (45%) were late presenters, corresponding to 42% and 34% of the 7901 patients in the overall cohort with complete data. The proportion of late HAART initiators who were LT was 50% for FH-Argentina, 23% for FA-Chile, 63% for GHESKIO-Haiti, 56% for IHSS/HE-Honduras, 48% for INNSZ-Mexico, and 59% for IMTAvH-Peru. Temporal trends were seen across sites (Figure 2). Notably, the proportions of late testers and late presenters both dropped substantially in GHESKIO-Haiti in recent years, although in 2007–2008 over 60% of GHESKIO-Haiti's LHI were late testers.
NOTE. The left and center plots are the corresponding proportions based on the entire cohort population (denominator = all HAART initiators). The right plot shows trends in the proportion of late HAART initiators who were late testers (denominator = all late HAART initiators). Rates of LT (left plot) statistically changed over time for most sites (p<0.001 for all sites except FH-Argentina where p = 0.42). Rates of LP (center plot) also changed over time for most sites (p = 0.11, 0.007, <0.001, 0.01, <0.001, and 0.01 for FH-Argentina, FA-Chile, GHESKIO-Haiti, IHSS/HE-Honduras, INNSZ-Mexico, and IMTAvH-Peru, respectively). FH-Argentina, Fundación Huésped in Buenos Aires, Argentina; FA-Chile, Fundación Arriarán in Santiago, Chile; GHESKIO-Haiti, Le Groupe Haïtien d'Etude du Sarcome de Kaposi et des Infections Opportunistes in Port-au-Prince, Haiti; IHSS/HE-Honduras Instituto Hondureño de Seguridad Social and Hospital de Especialidades in Tegucigalpa, Honduras; INNSZ-Mexico, El Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán in Mexico City, Mexico; and IMTAvH-Peru, El Instituto de Medicina Tropical Alexander von Humboldt in Lima, Perú. LHI, late HAART initators.
Risk Factors for LT and LP
Male sex was associated with late presentation and especially late testing for most sites and when all sites were considered together (Table 3; combined OR = 1.51, 95% CI 1.12–2.03, p-value 0.001 for LP; combined OR = 1.76, 95% CI 1.36–2.29, p-value<0.001 for LT). Older age was a risk factor for late testing (OR = 1.17 per 10 year increase in age, 95% CI 1.03–1.34, p-value 0.02), whereas higher education tended to be associated with a decreased risk of late presentation (OR 0.97, 95% CI 0.95–1.00, p-value 0.005).
Analyses were repeated using multiple imputation techniques to include patients with incomplete data. In general, results were very similar to those reported above (Supporting Information section). The estimated rates of LHI were 56% in FH-Argentina, 80% in FA-Chile, 76% in GHESKIO-Haiti, 89% in IHSS/HE-Honduras, 78% in INNSZ-Mexico, and 86% in IMTAvH-Peru. The only notable changes in multivariate analyses were 1) education was associated with LHI in FH-Argentina (OR = 0.97 per year, 95% CI 0.95–0.99) and 2) the association between age and LHI in IMTAvH-Peru slightly weakened (OR = 1.16 per 10 years, 95% CI 0.98–1.37).
The present study conducted among cohorts in different sites from 6 Latin-American countries shows that a significant proportion of patients starting HAART in this region do so after they have experienced an AIDS-defining illness or when the CD4 cell count is lower than 200, far below the cutoff of 350 cells/mm3 that is the recommended by most of the current international guidelines –. The prevalence of late HAART initiation as defined in this study ranges from 56% in FH-Argentina to 91% in IHSS/HE-Honduras with a combined mean of 75%. This finding presents a harsh reality to the current debate on when to start antiretroviral treatment , . While recent World Health Organization guidelines  have raised the target CD4 cell count at which treatment should be initiated, we found that most patients in our region still initiate treatment at a level which is far from optimal. In contrast to previous studies , our study analyzed not only the frequency with which patients present late to care, but as importantly, the frequency at which patients initiate therapy late. It has been assumed from previous studies that the main reason for late HAART initiation is that patients are not diagnosed early enough, often reflecting a failure of the local health systems to implement widespread testing strategies , , . In our study, however, a surprising finding was that a significant proportion of LHI (nearly 50% in the combined cohort; the proportion of patients who were LHI and were diagnosed more than 6 months before HAART initiation ranged from as high as 80% in FA-Chile, to approximately 40% in GHESKIO-Haiti and IMTAvH-Peru) were in fact diagnosed more than 6 months before starting HAART, and therefore, initiating late was not necessarily due to patients testing for HIV at late stages of disease but to other unexplained reasons. These reasons could include a poor linkage between the testing programs and the access to care facilities , , , need to be further studied. The high rates of late presentation are particularly noteworthy because access to HAART should not be a problem in any of the participating sites, as compared to other reports in Latin America , given that each is part of local programs for universal coverage of HAART paid by their respective governments.
Recently, in a large cohort of 14 sites in the USA and Canada (NA-ACCORD), late presentation to care seem to be have been decreasing between 1997 and 2007, based on the change in the median of CD4+ count at presentation to care . In our region, from 2004 to 2009 there appears to be a trend towards fewer patients initiating HAART in advanced disease stages in more recent years. This was particularly apparent in Haiti, where rates of LHI have substantially dropped since 2007. This drop is most likely attributable to a clinic-wide effort to initiate HAART in patients at earlier disease stages and also to the CIPRA study conducted at GHESKIO, which included patients starting HAART with CD4 cell counts that were higher than 200 cells/mm3 . Furthermore, GHESKIO has an integrated testing and treatment center, which facilitates linkages in care , , . The high rates of LHI in the earlier study years could be attributed in part to local guidelines containing late thresholds for initiating HAART –. Trends towards lower rates of LHI in recent years probably reflect in part a change in guidelines. Nevertheless, despite these trends, the prevalence of LHI in the most recent years still remains unacceptably high.
The outstanding risk factor for LHI in the multivariate analysis was male sex. Older age and less education were also associated with higher rates of LHI in many of the sites. Even though the AIDS epidemic in most of the participating countries is driven by MSMs, in the sites where these data were available (INNSZ-Mexico and FH-Argentina) MSMs tended to have a somewhat lower risk of being LHI. In the combined analysis of all cohorts, the same factors (male sex and older age) were significantly associated with late testing (recent diagnosis in LHI) and with late presentation (diagnosis more than 6 months before HAART initiation). These risk factors are similar to those reported in other studies for patients diagnosed in advanced stages .
One limitation of our study is the number of missing CD4 cell counts at HAART initiation from some of the sites. The analysis presented excluded those patients without data, which may bias the results. However, we obtained similar results using multiple imputation techniques which allow the inclusion of all patients initiating HAART. Another potential limitation is that we may have had insufficient power to detect some risk factors for LHI due to there being very few NLHI in some sites. The heterogeneity between rates of LHI at the different sites is also noteworthy, and suggests that there are important differences between clinic settings, perhaps in traditions, culture, HIV stigma, testing services, education, or something else. Unfortunately, with only six sites, we do not have sufficient data to investigate site-specific factors associated with LHI.
It is important to emphasize that our study did not measure the rate of late testing, but rather the proportion of patients who started HAART who were considered late testers. Of course, these patients represent only those who made it to care and received treatment. Our study does not consider persons in late stages of disease who do not know their HIV status, or diagnosed patients in late stages of disease who have not started treatment.
In conclusion our study shows a high rate of HAART initiation at CD4 cell counts <200 cell/mm3 or with an AIDS-defining illness in Latin-American and the Caribbean, which is substantially below the current cutoff recommended in international guidelines –. This observation is mostly explained by late HIV diagnosis however, a significant proportion of patients across all sites that meet criteria for treatment initiation experience considerable delays between the time of HIV diagnosis and the initiation of HAART. Avoiding this delay by improving the linkage between testing and treatment, as well as promoting early diagnosis of HIV, should result in better immunologic status at HAART initiation and better outcomes.
Probability of being a LHI over time adjusting for age, sex, and years of education. NOTE. 95% confidence intervals are reported. The adjusted rates of LHI statistically changed over time for most sites (p<0.001 for all sites except INNSZ-Mexico, p = 0.05, and IHSS/HE-Honduras, p = 0.95). FH-Argentina, Fundación Huésped in Buenos Aires, Argentina; FA-Chile, Fundación Arriarán in Santiago, Chile; GHESKIO-Haiti, Le Groupe Haïtien d'Etude du Sarcome de Kaposi et des Infections Opportunistes in Port-au-Prince, Haiti; IHSS/HE-Honduras Instituto Hondureño de Seguridad Social and Hospital de Especialidades in Tegucigalpa, Honduras; INNSZ-Mexico, El Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán in Mexico City, Mexico; and IMTAvH-Peru, El Instituto de Medicina Tropical Alexander von Humboldt in Lima, Perú. LHI, late HAART initators.
Number of patients by Calendar year of time of HAART initiation by site. NOTE. HAART, highly active antiretroviral therapy; FH-Argentina, Fundación Huésped in Buenos Aires; FA-Chile, Fundación Arriarán in Santiago, GHESKIO-Haiti, Le Groupe Haïtien d'Etude du Sarcome de Kaposi et des Infections Opportunistes in Port-au-Prince; IHSS/HE-Honduras Instituto Hondureño de Seguridad Social and Hospital de Especialidades in Tegucigalpa; INNSZ-Mexico, Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán in Mexico City; and IMTAvH-Peru, Instituto de Medicina Tropical Alexander von Humboldt in Lima; LHI, late HAART initiators.
Odds ratios (95% confidence intervals) for late HAART initiation using multiple imputationa. NOTE. FH-Argentina, Fundación Huésped in Buenos Aires, Argentina; FA-Chile, Fundación Arriarán in Santiago, Chile; GHESKIO-Haiti, Le Groupe Haïtien d'Etude du Sarcome de Kaposi et des Infections Opportunistes in Port-au-Prince, Haiti; IHSS/HE-Honduras Instituto Hondureño de Seguridad Social and Hospital de Especialidades in Tegucigalpa, Honduras; INNSZ-Mexico, El Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán in Mexico City, Mexico; and IMTAvH-Peru, El Instituto de Medicina Tropical Alexander von Humboldt in Lima, Perú. aData shown are odds ratios (95% confidence intervals). Estimates are adjusted for all variables in the table as well as date of HAART initiation.
The CCASAnet collaboration gratefully acknowledges the many patients and collaborating site staff who make this project possible. http://ccasanet.vanderbilt.edu/
Wrote the paper: BCR BES JSM. Conceived and designed this study: BCR JSM. Collected the data and were responsible for the data cohort: FW C. Cesar C. Cortés DP SK EG BCR. Had access to the full data: BCR YCV FW. Responsible for the statistical analysis and presentation of the results: YCV BES. Collected the data and references for discussing the results: FW C. Cesar C. Cortés DP SK EG BCR. Reviewed and approved the manuscript: BCR YCV BES FW C. Cesar C. Cortés DP SK EG PC CM DM JSM.
- 1. Pallela F, Moorman AC, Loveless MO, Fuhrer J, Satten GA, et al. (1998) Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. N Engl J Med 338(13): 853–60.
- 2. Castilla J, Sobrino P, De La Fuente L, Noguer I, Guerra L, et al. (2002) Late diagnosis of HIV infection in the era of highly active antiretroviral therapy: consequences for AIDS incidence. AIDS 16(14): 1945–51.
- 3. Chadborn TR, Bastera K, Delpecha VC, Sabin CA, Sinka K, et al. (2005) No time to wait: how many HIV-infected homosexual men are diagnosed late and consequently die? (England and Wales, 1993–2002) AIDS 19: 513–520.
- 4. Chadborn TR, Delpecha VC, Sabin CA, Sinka C, Evans BG (2006) The late diagnosis and consequent short-term mortality of HIV-infected heterosexuals (England and Wales, 2000–2004). AIDS 20: 2371–2379.
- 5. Egger M, May M, Chêne G, Phillips AN, Ledergerber B, et al. (2002) Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. The ART Cohort Collaboration. The Lancet 360(9327): 119–129.
- 6. Hocking JS, Rodger AJ, Rhodes DG, Crofts N (2000) Late presentation of HIV infection associated with prolonged survival following AIDS diagnosis–characteristics of individuals. Int J STD AIDS 11(8): 503–8.
- 7. Krentz HB, Auld MC, Gill MJ (2004) The high cost of medical care for patients who present late (CD4, 200 cells/mL) with HIV infection. HIV Med 5: 93–98.
- 8. Sabin CA, Smith CJ, Gumley H, Murphy G, Lampe FC, et al. (2004) Late presenters in the era of highly active antiretroviral therapy: uptake of and responses to antiretroviral therapy. AIDS 18(16): 2145–51.
- 9. Cesar C, Shepherd BE, Krolewiecki AJ, Fink VI, Schechter M, et al. (2010) Caribbean, Central and South America Network for HIV Research (CCASAnet) Collaboration of the International Epidemiologic Databases to Evaluate AIDS (IeDEA) Program. Rates and reasons for early change of first HAART in HIV-1-infected patients in 7 sites throughout the Caribbean and Latin America. PLoS One 5(6): e10490.
- 10. Tuboi SH, Schechter M, McGowan CC, Cesar C, Krolewiecki A, et al. (2009) Mortality During the First Year of Potent Antiretroviral Therapy in HIV-1-Infected Patients in 7 Sites Throughout Latin America and the Caribbean. J Acquir Immune Defic Synd 51: 615–623.
- 11. Mills E, Bakanda C, Birungi J, Mwesigwa R, Chan K, et al. (2011) Mortality by baseline CD4 cell count among HIV patients initiating antiretroviral therapy: evidence from a large cohort in Uganda. AIDS 851–55.
- 12. Sterne JA, May M, Costagliola D, de Wolf F, et al. When To Start Consortium (2009) Timing of initiation of antiretroviral therapy in AIDS-free HIV-1-infected patients: a collaborative analysis of 18 HIV cohort studies. Lancet 373(9672): 1352–63.
- 13. Kitahata MM, Gange SJ, Abraham AG, Merriman B, Saag MS, et al. (2009) NA-ACCORD Investigators. Effect of early versus deferred antiretroviral therapy for HIV on survival. N Engl J Med 360(18): 1815–26.
- 14. Mocroft A, Ledergerber B, Katlama C, Kirk O, Reiss P, et al. (2003) EuroSIDA study group. Decline in the AIDS and death rates in the EuroSIDA study: an observational study. Lancet 362(9377): 22–9.
- 15. Severe P, Juste MA, Ambroise A, Eliacin L, Marchand C, et al. (2010) Early versus standard antiretroviral therapy for HIV-infected adults in Haiti. N Engl J Med 363(3): 257–65.
- 16. Castilla J, Del Romero J, Hernando V, Marincovich B, García S, et al. (2005) Effectiveness of highly active antiretroviral therapy in reducing heterosexual transmission of HIV. J Acquir Immune Defic Syndr 40(1): 96–101.
- 17. Lima VD, Johnston K, Hogg RS, Levy AR, Harrigan PR, et al. (2008) Expanded access to highly active antiretroviral therapy: a potentially powerful strategy to curb the growth of the HIV epidemic. J Infect Dis 198(1): 59–67.
- 18. Girardi E, Aloisi MS, Arici C, Pezzotti P, Serraino D, et al. for the ICoNA Behavioural Epidemiology Study Group (2004) Delayed presentation and late testing for HIV: demographic and behavioral risk factors in a multicenter study in Italy. J Acquir Immune Defic Syndr 36(4): 951–9.
- 19. The UK Collaborative HIV Cohort (UK CHIC) Steering Committee (2010) Late diagnosis in the HAART era: proposed common definitions and associations with mortality. AIDS 24(5): 723–7.
- 20. Fisher M (2008) Late diagnosis of HIV infection: major consequences and missed opportunities. Curr Opin Infect Dis 21(1): 1–3.
- 21. Centers for Disease Control and Prevention (CDC)Missed opportunities for earlier diagnosis of HIV infection–South Carolina, 1997–2005. MMWR Morb Mortal Wkly Rep 2006 55(47): 1269–72.
- 22. Nash D, Katyal M, Brinkhof MW, Keiser O, May M, et al. (2008) ART-LINC Collaboration of IeDEA. Long-term immunologic response to antiretroviral therapy in low-income countries: a collaborative analysis of prospective studies. AIDS 22(17): 2291–302.
- 23. Bonjour MA, Montagne M, Zambrano M, Molina G, Lippuner C, et al. (2008) Determinants of late disease-stage presentation at diagnosis of HIV infection in Venezuela: a case-case comparison. AIDS Res Ther 5: 6. DOI:10.1186/1742-6405-5-6.
- 24. McGowan CC, Cahn P, Gotuzzo E, Padgett D, Paper JW, et al. (2007) Cohort Profile: Caribbean, Central and South America Network for HIV research (CCASAnet) collaboration within the International Epidemiologic Databases to Evaluate AIDS (IeDEA) programme. Int J Epidemiol 36(5): 969–76.
- 25. Friedman JH (1984) SMART User's Guide. Laboratory for Computational Statistics, Stanford University Technical Report No. 1.
- 26. Harrell FE (2001) Regression Modeling Strategies with Applications to Linear Models, Logistic Regression, and Survival Analysis. New York, NY: Springer.
- 27. DerSimonian R, Laird N (1986) Meta-analysis in clinical trials. Control Clin Trials 7(3): 177–88.
- 28. Smith-Warner SA, Spiegelman D, Ritz J, Albanes D, Beeson W, et al. (2006) Methods for pooling results of epidemiologic studies: the Pooling Project of Prospective Studies of Diet and Cancer. Am J Epidemiol 163(11): 1053–64. Available: http://aje.oxfordjournals.org/content/163/11/1053.full.pdfhtml. Accessed 22 March 2011.
- 29. Shafer J (1997) Analysis of Incomplete Multivariate Data. London, United Kingdom: Chapman & Hall.
- 30. WHO (2010) Antiretroviral therapy for HIV infection in adults and adolescents: Recommendation for a public health approach. Geneva, Switzerland.
- 31. EACS (2009) Clinical Management and Treatment of HIV infected Adults in Europe.
- 32. OARAC (2009) Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents. Developed by the DHHS Panel – A Working Group of the Office of AIDS Research Advisory Council.
- 33. Gilks CF, Crowley S, Epkini R, Gove S, Perriens J, et al. (2006) The WHO public-health approach to antiretroviral treatment against HIV in resource-limited settings. Lancet 368(9534): 505–10.
- 34. Walensky RP, Wolf LL, Wood R, Fofana MO, Freedberg KA, et al. (2009) When to start antiretroviral therapy in resource-limited settings. Ann Intern Med 151: 157–66.
- 35. Schwarcz S, Hsu L, Dilley JW, Loeb L, Nelson K, et al. (2006) Late diagnosis of HIV infection: trends, prevalence, and characteristics of persons whose HIV diagnosis occurred within 12 months of developing AIDS. J Acquir Immune Defic Syndr 43(4): 491–4.
- 36. Assefa Y, Van Damme W, Mariam DH, Kloos H (2010) Toward universal access to HIV counseling and testing and antiretroviral treatment in Ethiopia: looking beyond HIV testing and ART initiation. AIDS Patient Care STDS 24(8): 521–5.
- 37. Losina E, Basset IV, Giddy J, Chetty S, Regan S, et al. (2010) The “ART” of linkage: pre-treatment loss to care after HIV diagnosis at two PEPFAR sites in Durban, South Africa. PLoS One 5(3): e9538.
- 38. Zala C, Rustad CA, Chan K, Khan NI, Beltran M, et al. (2008) Determinants of treatment access in a population-based cohort of HIV-positive men and women living in Argentina. Medscape J Med 10(4): 78.
- 39. Althoff KN, Gange SJ, Klein MB, Brooks JT, Hogg RS, et al. (2010) Late presentation for human immunodeficiency virus care in the United States and Canada. Clin Infect Dis 50(11): 1512–20.
- 40. WHO (2004) Scaling up antiretroviral therapy in resource-limited settings. Geneva, Switzerland.
- 41. WHO (2006) Antiretroviral therapy for HIV infection in adults and adolescents: recommendations for a public health approach: 2006 revision. Geneva, Switzerland.
- 42. SADI (2007) Recomendaciones para el seguimiento y tratamiento de la infección por HIV. Available: http://www.sadi.org.ar/images/HIV2007.pdf. Accessed 22 March 2011.
- 43. Ministerio de Salud (2007) Recomendaciones para el tratamiento antirretroviral 2007. Available: http://www.msal.gov.ar/sida/pdf/info-equipos-salud/recomendaciones-tratamiento-antirretroviral.pdf. Accessed 22 March 2011.
- 44. CENSIDA (2008) Guía de manejo antirretroviral de las personas con VIH. Cuarta edición. México. Available: Centro Nacional para la Prevención y el Control del VIH/SIDA. http://www.censida.salud.gob.mx/interior/atencion.html. Accessed 8 November 2010.
- 45. Ministerio de Salud (2005) Guía Clínica Síndrome de la Inmunodeficiencia Adquirida VIH/SIDA. 1st Ed. Santiago.