0
We're unable to sign you in at this time. Please try again in a few minutes.
Retry
We were able to sign you in, but your subscription(s) could not be found. Please try again in a few minutes.
Retry
There may be a problem with your account. Please contact the AMA Service Center to resolve this issue.
Contact the AMA Service Center:
Telephone: 1 (800) 262-2350 or 1 (312) 670-7827  *   Email: subscriptions@jamanetwork.com
Error Message ......
Original Contribution |

Treatment Failure and Mortality Factors in Patients Receiving Second-Line HIV Therapy in Resource-Limited Countries FREE

Mar Pujades-Rodríguez, MD, PhD; Suna Balkan, MD; Line Arnould, MD; Martin A. W. Brinkhof, PhD; Alexandra Calmy, MD, PhD; for the AIDS Working Group of MSF
[+] Author Affiliations

Author Affiliations: Epicentre (Dr Pujades-Rodríguez), Médecins Sans Frontières (Dr Balkan), Paris, France; Médecins Sans Frontières, Brussels, Belgium (Dr Arnould); Institute of Social and Preventive Medicine, Bern, Switzerland (Dr Brinkhof); and Médecins Sans Frontières Campaign for Access to Essential Medicines, Geneva, Switzerland (Dr Calmy).


JAMA. 2010;304(3):303-312. doi:10.1001/jama.2010.980.
Text Size: A A A
Published online

Context Long-term antiretroviral therapy (ART) use in resource-limited countries leads to increasing numbers of patients with HIV taking second-line therapy. Limited access to further therapeutic options makes essential the evaluation of second-line regimen efficacy in these settings.

Objectives To investigate failure rates in patients receiving second-line therapy and factors associated with failure and death.

Design, Setting, and Participants Multicohort study of 632 patients >14 years old receiving second-line therapy for more than 6 months in 27 ART programs in Africa and Asia between January 2001 and October 2008.

Main Outcome Measures Clinical, immunological, virological, and immunovirological failure (first diagnosed episode of immunological or virological failure) rates, and mortality after 6 months of second-line therapy use. Sensitivity analyses were performed using alternative CD4 cell count thresholds for immunological and immunovirological definitions of failure and for cohort attrition instead of death.

Results The 632 patients provided 740.7 person-years of follow-up; 119 (18.8%) met World Health Organization failure criteria after a median 11.9 months following the start of second-line therapy (interquartile range [IQR], 8.7-17.0 months), and 34 (5.4%) died after a median 15.1 months (IQR, 11.9-25.7 months). Failure rates were lower in those who changed 2 nucleoside reverse transcriptase inhibitors (NRTIs) instead of 1 (179.2 vs 251.6 per 1000 person-years; incidence rate ratio [IRR], 0.64; 95% confidence interval [CI], 0.42-0.96), and higher in those with lowest adherence index (383.5 vs 176.0 per 1000 person-years; IRR, 3.14; 95% CI, 1.67-5.90 for <80% vs ≥95% [percentage adherent, as represented by percentage of appointments attended with no delay]). Failure rates increased with lower CD4 cell counts when second-line therapy was started, from 156.3 vs 96.2 per 1000 person-years; IRR, 1.59 (95% CI, 0.78-3.25) for 100 to 199/μL to 336.8 per 1000 person-years; IRR, 3.32 (95% CI, 1.81-6.08) for less than 50/μL vs 200/μL or higher; and decreased with time using second-line therapy, from 250.0 vs 123.2 per 1000 person-years; IRR, 1.90 (95% CI, 1.19-3.02) for 6 to 11 months to 212.0 per 1000 person-years; 1.71 (95% CI, 1.01-2.88) for 12 to 17 months vs 18 or more months. Mortality for those taking second-line therapy was lower in women (32.4 vs 68.3 per 1000 person-years; hazard ratio [HR], 0.45; 95% CI, 0.23-0.91); and higher in patients with treatment failure of any type (91.9 vs 28.1 per 1000 person-years; HR, 2.83; 95% CI, 1.38-5.80). Sensitivity analyses showed similar results.

Conclusions Among patients in Africa and Asia receiving second-line therapy for HIV, treatment failure was associated with low CD4 cell counts at second-line therapy start, use of suboptimal second-line regimens, and poor adherence. Mortality was associated with diagnosed treatment failure.

Figures in this Article

Universal access to combined antiretroviral therapy (ART) includes access to first- and second-line regimens.1 The use of standardized and affordable first-line fixed-dose combinations of antiretroviral drugs including 2 nucleoside reverse transcriptase inhibitors (NRTIs) and 1 nonnucleoside reverse transcriptase inhibitor (NNRTI) has been critical to allow the scale-up of combined ART in resource-limited countries with high human immunodeficiency virus (HIV) prevalence.2 Most countries currently provide guidelines for the use of second-line regimens, and extensive use of ART has led to increasing numbers of patients receiving second-line therapy for long periods. It is thus likely that increasing numbers of patients in these settings will have diminished response to second-line treatment.

The issue of second-line failure has been addressed for the first time in the recent 2009 revision of the World Health Organization (WHO) treatment guidelines, which recommend third-line regimens be made available in all countries.3 Experience in highly resourced countries has shown that triple-class experienced patients can be successfully treated with effective new antiretroviral regimens.35 However, because third-line regimens are costly and not readily available in resource-limited countries, second-line treatment is often the last therapeutic option available for patients in these settings. Maximizing the duration of second-line regimens, quantifying the need for third-line therapy in patients receiving ART in resource-limited countries, and decreasing the price of new well-tolerated antiretroviral regimens are therefore important.

In this study, we describe rates of failure for second-line therapy and investigate individual- and programmatic-level factors associated with second-line therapy failure and with death in patients treated with second-line therapy in 27 African and Asian ART programs supported by Médecins Sans Frontières (MSF).

Routine clinical and laboratory patient data (FUCHIA software, Epicentre, Paris, France) collected between January 2001 and October 2008 by 27 MSF-supported programs were compiled in February 2009 and analyzed. Some patients were included in a prior analysis of switch rates to second-line therapy, early second-line outcomes, and determinants of switching and early mortality in resource-limited settings.6 For the updated analyses with longer follow-up presented herein, 4 new cohorts are included, 2 from Africa and 2 from Asia. A total of 193 patients receiving second-line therapy for more than 6 months in these new cohorts are included herein. Two patients who initiated ART after December 2006 were treated with the prior cohort but were not included in the prior cohort analysis because the database had been censored in December 2006. Of the 437 patients included in both analyses, 132 were switched to second-line therapy after the time of the prior analysis. For 307 patients derived from the prior cohort, longer follow-up time while taking second-line therapy was available and analyzed herein (database update, October 30, 2008).

Programs have been described previously.6 Briefly, physicians, clinical officers (health practitioners receiving medical training who are more qualified than nurses but less qualified than physicians, with the level of training varying by country), or nurses provide free care and treatment to patients with HIV. WHO recommendations for ART were followed,7 and adherence counseling was offered before and after ART start. Generic antiretroviral drugs were prescribed, and CD4 cell counts were measured using automated or manual methods.

In agreement with health ministries, data were prospectively collected. No patient identifiers were kept in the data sets analyzed, and proposals for analysis were approved by the International Ethics Review Board of MSF. As advised by the National Commission of Informatics and Liberties, given the contexts in which MSF works, information for patients about the data collection system in the facility and its use was judged to be more effective than obtaining written informed consent.

Inclusion Criteria and Definitions

All patients who were aged 15 years or older, had known sex information (all patients who had been switched to second-line therapy had sex information recorded), had initiated NNRTI-based ART, were antiretroviral-naive at the start of treatment, and were receiving a second-line protease inhibitor–based regimen for more than 6 months were eligible for inclusion. WHO treatment guidelines recommend allowing at least 6 months of treatment with a given regimen before diagnosing treatment failure. This is to avoid misclassifying patients who did not receive treatment for a sufficient length of time as having treatment failure (eg, patients need to be exposed to active drugs for a sufficient amount of time to show signs of immunovirological recovery and to reach the level of immunological recovery that protects against the development of opportunistic infections). Patients were considered lost to follow-up if the time between their last visit and the closing date of the corresponding cohort was more than 1 year.

Programs in which 50% or more of patients had 1 or more measurements of viral load 3 to 9 months after second-line therapy start were considered to have had their viral loads routinely monitored. Recorded values of CD4 cell count and viral load closest to the starting date of first-line or second-line ART (3 months before and up to 2 weeks after starting ART) were defined as initial counts. For CD4 cell counts lower than 25/μL and HIV viral load higher than 100 000 copies/mL, the time window for first-line treatment was extended to 1 year before starting therapy because these patients, who were eligible for ART at the time of testing, were likely to have received treatment for another infection or pathological condition before being prescribed ART. These patients might not have been retested for CD4 cell count, viral load, or either at the start of ART. In fact, only 5 patients had CD4 cell count values at ART start that were quantified between 3 and 6 months before the start of ART.

We defined clinical failure for those receiving second-line therapy as the diagnosis of a new WHO clinical stage 3 or 4 event after 6 months of second-line use; immunological failure as a decrease in CD4 cell count to its initial value or lower, decline of at least 50% from highest measurement while receiving treatment, or CD4 cell count lower than 100/μL (or <50/μL for sensitivity analyses) at 1 year of second-line therapy, confirmed within 6 months; and virological failure as a recorded plasma viral load value of more than 10 000 copies/mL confirmed within 6 months.8 Failure of any type was the earlier recorded failure event for each patient (clinical, immunological, or virological).

For the purposes of the analyses herein, we considered failure and mortality or retention as different outcomes in the analysis rather than considering that any death occurring after 6 months of second-line therapy start was due to treatment failure. The precise cause of death of the patients is not recorded in the data sets; thus, we were not able to distinguish between AIDS-related deaths attributable to treatment failure (eg, deaths occurring in the absence of biological confirmation of failure because testing was not performed or results not recorded) and non−AIDS-related deaths.

Information on the clinic appointment date for each patient visit was available for 433 patients (68.5%) analyzed (66.3% of those without second-line treatment failure, and 78.2% of those who had recorded failure). As a surrogate marker for treatment adherence, we created an index by dividing the number of scheduled visits attended with delay (>1 day after the corresponding appointment date) by the number of months receiving second-line therapy and multiplying this by 100. We considered that patients who attended fewer than 5% of their appointments with delay were 95% or more adherent, those with 5% to 19% of delayed appointments were 80% to 94% adherent, and those with 20% or more delayed appointments were less than 80% adherent to therapy. We grouped patients according to the incidence of appointments attended without delay (<80%, 80%-94%, or ≥95%). The cutoff values were selected based on previous studies investigating effects of adherence indicators on treatment outcomes for HIV patients.9,10 Previous research has shown that patients with 95% or better adherence had better rates of virological suppression, less resistance, greater increase in CD4 cell counts, and lower hospital rates than patients with lower adherence.9,11 Furthermore, patients taking NNRTIs or boosted protease inhibitors with adherence of 80% or more had failure rates of less than 10%.12

The adherence variable was created taking into account each consecutive date of appointment and the following recorded date of visit. For patients who did not return, the last recorded appointment date is not considered in the computation of the adherence index because there is no recorded clinical visit afterward for the assessment. These patients would have been considered lost to follow-up in the analysis.

Statistical Analysis

We measured time from 6 months after starting second-line therapy to diagnosis of treatment failure, last clinical visit, or death. We split follow-up periods into 6 to 11, 12 to 17, and 18 or more months. We described characteristics of patients by failure status, calculated times to failure and rates per 1000 person-years with Poisson exact 95% confidence intervals (CIs), and fitted random-effect Poisson models using adaptive Gauss-Hermite quadrature methods13 to identify factors associated with treatment failure.

Factors considered were type of setting (urban or rural); health care level (tertiary or other); use of routine viral load monitoring; use of viral load testing for failure confirmation; sex; age group (15-29, 30-39, 40-49, and ≥50 years); calendar year of second-line therapy start (<2005, 2005, 2006, 2007-2008); split follow-up time on second-line therapy; imputed WHO clinical stage and CD4 cell count at ART start (stages 1 or 2, 3, and 4; <50, 50-99, and ≥100/μL, respectively) and second-line therapy initiation (stages 1 or 2, and 3 or 4; <50, 50-99, 100-199, and ≥200/μL, respectively); initial NRTI-backbone (stavudine and other) and NNRTI drug (efavirenz and nevirapine); duration on first-line therapy (as a continuous variable); type of second-line regimen, ie, type of protease inhibitor (lopinavir boosted with ritonavir, nelfinavir, and other) and number of NRTI changes at switch; and adherence index (≥95%, 80%-94%, and <80%), for the subgroup of patients with available appointment dates (68.5%).

Information on clinical stage and CD4 cell counts was missing in some patients (1% and 14%, respectively, for first-line therapy start; 24% and 32%, respectively, for second-line therapy start). To impute missing data we used multiple imputation by chained equation methods.14,15 Prediction equations included cohort (degree of urbanization for second-line imputed variables), age at ART start, sex, first CD4 cell count value and time to measurement after pretreatment CD4 cell count (for first-line or second-line regimens), patient outcome (death, lost to follow-up, or alive), and follow-up time on second-line therapy censored at 1 year of second-line therapy use. Equations for imputed second-line therapy-related data also included nadir CD4 cell count before second-line therapy start, and indicators for clinical, immunological, and virological failure. Before imputation, continuous variables were normalized using square-root transformation for absolute and nadir CD4 cell counts, and log-transformation for age, follow-up time, and time to first pretreatment CD4 cell measurement. We used interval censoring to ensure imputation of CD4 cell values within the appropriate range (0-800/μL). We generated 10 imputed data sets. Individual and programmatic factors associated with failure in univariable analyses (P < .2) were included in adjusted models. Estimates of coefficients were derived by averaging, and appropriate standard errors calculated applying Rubins rules.1618 Two sensitivity analyses were performed, 1 for any type of failure using the definition of immunological failure as less than 50 cells/μL instead of less than 100 cells/μL,8 and 1 for immunovirological failure (first diagnosed episode of immunological or virological failure using the <100/μL definition). The WHO 2006 guidelines,8 in attempting to standardize treatment failure definition criteria, defined immunological failure as a decrease in CD4 cell count to or lower than baseline, a decrease of 50% from the treatment peak value reached, or a persistent CD4 cell count lower than 100/μL. The rationale for using the CD4 cell count lower than 100 cells was because the other 2 criteria would require either a CD4 cell measurement at the start of therapy or would be difficult to implement in practice. Because many patients initiate first-line therapy with very low values of CD4 cell counts (median of 45/μL; second-line therapy, with CD4 cell counts close to 100/μL) in our MSF sites, we used the 50-cell threshold to increase the specificity of the immunological criteria to correctly identify treatment failure in sensitivity analyses. Results from adjusted models are provided for the whole cohort; estimates further adjusted for adherence data are given only for patients with this information.

We also examined associations between mortality after 6 months of second-line therapy and the aforementioned factors, failure of any type (with sensitivity analyses using the 2 mentioned alternative failure definitions), and occurrence of change in the prescribed second-line regimen for any reason. Because of fewer deaths than failure events, we grouped patients in the following categories: follow-up time on second-line therapy: 6 to 17 months and 18 or more months; clinical stage at first-line therapy: stages 1, 2, or 3 and stage 4; CD4 cell counts at second-line therapy start: less than 50, 50-99, and 100/μL or higher; and adherence index: at least 95% and less than 95%. In our data set, only 5 patients who died had reported adherence of less than 80%. We used Weibull random-effect survival models19,20 and combined estimates from imputed data sets as before. We performed sensitivity analysis of associations using cohort attrition (combined mortality and care discontinuation) instead of death.

The objective of the study was to investigate the effect of a number of programmatic-level and individual-level factors on the 2 main outcomes, therapeutic failure and mortality. Because there was no specific hypothesis for testing, power was not calculated prior to data analysis. However, when the estimates of effect for the failure outcome were examined, no important effects appeared to have been missed. Analyses investigating effects on mortality showed statistically significant associations for diagnosed second-line therapy failure, sex, and type of setting. However, because of the small number of deaths during the time of study follow-up (n = 34) and the small number of sites using viral load monitoring, inability to detect a potential association between mortality and viral load monitoring could have occurred. When the power by simulation technique21,22 was applied to the sample herein using the Weibull model, it was estimated that the study sample size would allow a hazard rate ratio (HR) of 0.21 (estimate from our final model) to be found statistically significant in 7.4% of samples, assuming ±5% precision.

Analyses were performed using Stata version 10.1 (Stata Corp, College Station, Texas), with α = .05. Results are presented as adjusted incidence rate ratios (IRRs) or HRs with 95% CIs and P values from Wald tests for associations calculated across variable categories. Statistical significance was defined as P < .05.

Sites and Study Population

Twenty-seven HIV cohorts from 13 countries were included (Table 1). All sites had access to second-line regimens and routinely performed CD4 cell count testing, 4 routinely monitored viral load, and 19 used viral load to confirm treatment failure. After excluding 335 patients (34.6%) who received second-line therapy for less than 6 months, we analyzed data from a total of 632 adults (>14 years).

Table Graphic Jump LocationTable 1. Characteristics of Human Immunodeficiency Virus Programs Included in Analyses by Geographical Area

A total of 516 patients (81.7%) were treated in urban sites and 77.7% in primary health care facilities. At second-line therapy start, the median age was 35 years (interquartile range [IQR], 30-42 years), median time using ART 24.1 months (IQR, 16.1-31.1 months), and median decrease in CD4 cell count since ART start −49/μL (IQR, −136 to 0/μL]. At second-line therapy start, 71.8% of patients received lopinavir boosted with ritonavir−containing regimens, and 79.3% had changed 2 NRTI drugs (eTable 1).

Rates and Time to Failure Using Second-Line Therapy

Cumulative probabilities of treatment failure from 6 to 30 months of second-line therapy are shown in the Figure. At 1 and 2 years of second-line start, 12% and 28% of patients, respectively, had recorded failure of any type; and 2% and 8% had immunovirological failure, respectively. During follow-up, 119 patients (18.8%) met failure criteria after a median time of 11.9 months of second-line therapy use (IQR, 8.7-17.0 months; Figure). Time to failure was similar when clinical- or laboratory-based criteria were used (Table 2). When the definitions immunological failure of less than 100/μL and less than 50/μL were applied, 30-month failure rates were 214.6 (95% CI, 178.9-257.4) and 205.9 (95% CI, 171.1-247.8) per 1000 person-years, respectively. Rates of immunovirological failure were 52.9 (95% CI, 37.4-74.8) and 46.0 (95% CI, 31.8-66.7) per 1000 person-years, respectively.

Place holder to copy figure label and caption
Figure. Cumulative Probability of Second-Line Regimen Failure, Death, and Attrition
Graphic Jump Location

The shaded area represents 95% confidence intervals for Kaplan-Meier estimates.

Table Graphic Jump LocationTable 2. Numbers of Patients Meeting Criteria of Treatment Failure, Time, and Rates of Failure per 1000 Person-Years at 30 Months of Second-Line Therapya
Factors Associated With Failure Using Second-Line Therapy

Rates of second-line therapy failure of any type were greater in patients who were treated in hospitals than in primary health care facilities (289.5 vs 173.6 per 1000 person-years; adjusted IRR, 1.61; 95% CI, 1.01-2.57; Table 3). Rates increased with lower CD4 cell counts at second-line therapy start, from 156.3 vs 96.2 per 1000 person-years; adjusted IRR, 1.59 (95% CI, 0.78-3.25) for 100-199/μL; to 336.8 per 1000 person-years; adjusted IRR, 3.32 (95% CI, 1.81-6.08) for less than 50/μL than with 200/μL or more. Estimates of effect for women compared with men, before and after adjusting for adherence data, are shown in Table 3. The rate of failure for women without adjustment for adherence information was not statistically significant (adjusted IRR, 1.28; 95% CI, 0.86-1.91, P = .23). The estimate when we adjusted further for adherence information was of borderline significance (adjusted IRR, 1.56; 95% CI, 0.99-2.44, P = .05).

Table Graphic Jump LocationTable 3. Association Between Second-Line Failure and Programmatic and Individual Level Factorsa

Patients who changed 2 of their NRTI drugs at second-line therapy start were less likely to fail treatment than those who changed only 1 (179.2 per 1000 person-years; 95% CI, 144.9-221.7 vs 251.6; 95% CI, 179.8-352.1; adjusted IRR, 0.64; 95% CI, 0.42-0.96). Failure was most frequent during the 6- to 11-month period after switching regimens (250.0 vs 123.2 per 1000 person-years in ≥18-month period; adjusted IRR, 1.90; 95% CI, 1.19-3.02); and in patients with an adherence index of less than 80% (383.5 vs 176.0 per 1000 person-years in ≥95% adherent; adjusted IRR; 3.14, 95% CI 1.67-5.90). Use of second-line regimens containing lopinavir boosted with ritonavir was associated with lower rates of failure (226.9 vs 168.4 per 1000 person-years; adjusted IRR, 1.74; 95% CI, 1.14-2.67 for nelfinavir; 336.5 per 1000 person-years; adjusted IRR, 1.61; 95% CI, 0.73-3.56 for other protease inhibitor vs lopinavir boosted with ritonavir). However, this association was no longer significant after adjustment for adherence (P = .25).

Use of routine viral load monitoring and of viral load testing for confirmation of clinical and immunological failure were unrelated to failure diagnosis (P = .96 and P = .41, respectively). Analysis based on the definition of immunological failure as a CD4 cell count of less than 50/μL instead of less than 100/μL gave similar results (eTable 2). Rates of immunovirological failure were also higher in patients with lower CD4 cell counts at second-line therapy start (79.1 vs 23.5 per 100 person-years; adjusted IRR, 3.05; 95% CI, 1.19-7.83 for 50-199/μL; 83.8 per 1000 person-years; adjusted IRR, 3.41; 95% CI, 1.16-9.99 for <50/μL vs ≥100/μL; eTable 2).

Mortality Rates With Second-Line Therapy and Associated Factors

A total of 34 patients (5.4%) died and 23 (3.6%) were lost from care after the first 6 months of follow-up taking second-line therapy. Fifteen of the deaths occurred in patients with no recorded failure. The death rate at 30 months of second-line therapy use was 44.2 (95% CI, 30.5-64.0), dropout rate, 33.2 per 1000 person-years (95% CI, 21.6-50.9), and attrition rate, 77.4 per 1000 person-years (95% CI, 58.5-102.4). Death occurred in a median of 15.1 months (IQR, 11.9-25.7 months), and care discontinuation, 12.2 months (IQR, 9.7-20.0 months) after second-line therapy start (Figure).

Mortality was lower in rural than in urban sites (33.6 vs 49.0 per 1000 person-years; adjusted HR, 0.33; 95% CI, 0.12-0.91), but was unrelated to use of both routine viral load monitoring and viral load testing for confirmation of clinical or immunological failure (P = .14 and P = .09, respectively; eTable 3). Women had lower rates than men (32.4 vs 68.3 per 1000 person-years; adjusted HR, 0.45; 95% CI, 0.23-0.91). Patients who met criteria for treatment failure of any type (91.9 vs 28.1 per 1000 person-years; adjusted HR, 2.83; 95% CI, 1.38-5.80 for <100/μL; 90.2 vs 29.6 per 1000 person-years; adjusted HR, 2.74; 95% CI, 1.35-5.58 for <50/μL; P = .005) and those diagnosed with immunovirological failure (91.5 vs 27.7 per 1000 person-years; adjusted HR, 3.17; 95% CI, 1.18-8.50; P = .02) were at increased risk of death compared with other patients. Length of time on first-line therapy before second-line therapy start was unrelated to mortality (P = .78).

Results from sensitivity analyses examining associations with cohort attrition instead of death were consistent with those for mortality (eTable 4). In addition, we found evidence of decreased attrition in patients who experienced a change in their second-line regimen during follow-up (70.3 vs 86.8 per 1000 person-years; adjusted HR, 0.42; 95% CI, 0.22-0.80).

In this prospective, multicohort analysis of 27 ART programs in resource-limited settings, 19% of patients receiving second-line therapy for more than 6 months experienced treatment failure, 5% died, and 4% were lost to follow-up. Higher failure rates were observed in patients who initiated second-line therapy at lower CD4 cell counts, in those who changed 1 NRTI drug instead of 2 at second-line therapy start, in those who did not receive lopinavir boosted with ritonavir−based second-line regimens, and during the 6- to 17-month period after second-line start. Furthermore, mortality was higher in men, patients with diagnosed second-line failure, and those treated in urban sites.

In our cohorts, the overall rate of failure using second-line therapy was 46% higher than failure using first-line (195 vs 134 per 1000 person-years).23 Others have reported failure rates on first-line ART of 24,24 and 78 per 1000 person-years.25 Possible explanations for the observed higher failure rates with second-line therapy than with first-line therapy are worse tolerability of regimens, acquired drug resistance in the course of therapy, or difficulty in maintaining adherence to therapy after a long time receiving ART. This is supported by our analysis in which poorly adherent patients were 3 times more likely to fail second-line therapy than those with optimal adherence to therapy. Patients switched to second-line therapy have already experienced a first episode of treatment failure, often resulting from episodic periods of suboptimal adherence to ART. In Western countries, worse patient outcomes are frequently observed in antiretroviral-experienced than antiretroviral-naive patients.26,27 In an urban South African cohort, higher rates of viral suppression for second-line therapy were achieved by patients who had switched to second-line therapy for reasons unrelated to nonadherence.28 Our findings highlight the need for strengthening of treatment adherence by counselors and support groups.

In the absence of specific criteria to diagnose second-line therapy failure, we chose to use the same definitions as failure for first-line therapy. Use of highly sensitive but unspecific criteria to define second-line therapy failure could partly explain the observed high failing rates during second-line therapy because the time needed to achieve immunological recovery might be longer for second than for first line. In resource-rich countries, the same definitions of failure are used regardless of therapy line.29 Furthermore, patients who met failing criteria with or without viral load confirmation were at increased risk of death and attrition, and findings were robust to changes in definitions of failure.

In our analysis, several factors were found to be associated with second-line therapy failure. Regimen potency was critical to prevent treatment failure. Hence, patients who underwent a change of at least 2 NRTIs instead of 1, and those who used lopinavir boosted with ritonavir−based therapy instead of other protease inhibitors, were less likely to fail second-line therapy. Most patients in resource-limited countries are only switched to second-line therapy after the diagnosis of clinical or immunological failure. Delayed treatment modification has been associated with poorer outcomes among patients failing NNRTI-based first-line therapy, as opposed to those receiving protease inhibitor–based regimens.30 Exposure to replicating virus for relatively long periods could lead to accumulation of mutations and compromise the efficacy of second-line regimens.31 In our analysis, we did not find evidence of association between duration of first-line therapy and rate of failure on second-line therapy, even after adjusting for differences in adherence. However, the absence of genotyping data did not allow differentiating patients who had developed mutations conferring resistance from those who did not.

We and others have previously documented the role of CD4 cell count nadir as a predictor of success of first-line therapy.6,25 Before the widespread use of genotyping in Europe, high CD4 count levels at the start of protease inhibitor–based therapy were found to predict subsequent virological suppression.32 Our findings suggest that CD4 count levels at second-line therapy start are more important determinants of failure than those observed at the beginning of therapy. This highlights the need for early detection of suboptimal adherence to first-line therapy.

We also observed relatively high mortality and loss to follow-up rates (attrition of 77 per 1000 person-years) after 6 months of second-line use. Although no similar data are currently available from resource-limited countries, our estimates are within the ranges of those reported for first-line therapy (30-370 per 1000 person-years during the first 6 months of therapy, and decrease to 13-100 per 1000 person-years in later periods33).

The strongest factor associated with death using second-line therapy was diagnosis of second-line therapy failure. Patients meeting failure diagnosis were nearly 3 times more likely to die than other patients.

Women and patients attending rural health facilities experienced lower mortality than other patients. Differences in overall patient adherence achieved by men and women could explain these findings. Studies conducted in Africa have reported poorer adherence to treatment34 and higher mortality and loss to follow-up rates among men than among women.3538

The lower second-line therapy failure rates observed in rural settings might be explained by better access and closer monitoring of patients and better support to achieve second-line adherence than in larger health facilities.

Finally, we found lower attrition but not mortality in patients who had experienced a change in the initial second-line regimen prescribed. In the absence of differences in length of follow-up and in numbers of deaths and lost to follow-up patients between individuals who changed or not their second-line regimen, those who did (eg, because of toxicity) might have been more likely to continue taking their drugs and remain in care than those who were kept with the same regimen.

WHO 2006 guidelines recommended that patients receiving but not responding to second-line therapy continue the same drug regimen unless serious toxicity or drug interactions develop. In the past 3 years, the drug formulary against HIV/AIDS has increased dramatically, and viral load suppression can now be achieved regardless of presence of mutated virus. None of the newly developed drugs (ie, etravirine, darunavir, raltegravir) are currently registered and marketed in resource-limited countries, but new WHO guidelines will now recommend the provision of third-line regimens.39 Sequencing well-tolerated and efficient regimens from first-line to second-line therapy, and from second-line to third-line therapy, is critical, however, to ensure good therapeutic outcomes.

Other limitations of our analysis not yet discussed include residual confounding by incomplete estimate adjustment, limited time of follow-up, small number of sites using viral load monitoring, small number of deaths, and the possibility of bias by differential loss to follow-up. Complete and accurate recording of routine data are difficult to achieve in resource-limited countries; MSF invests considerable resources to improve and maintain the quality of the information collected. Analyses were adjusted for an important number of programmatic and individual factors, but information such as patient adherence or diagnosed clinical nonstaging conditions was unavailable for some or all the study patients. Finally, sensitivity analyses for which associations with attrition instead of death were investigated showed similar results.

The observed rates of second-line therapy failure and associated mortality in this multicentric study in resource-limited countries highlight the need for both strengthening patient adherence to second-line therapy and increased availability of alternative and better tolerated fixed-dose combinations of antiretroviral drugs. New drugs with different modes of action and without cross-resistance to NNRTI and protease inhibitor–based regimens are also needed for patients who live in resource-limited countries.

Corresponding Author: Mar Pujades-Rodríguez, MD, PhD, Epicentre, Médecins Sans Frontières 42-bis, Bd, 8 rue Saint Sabin, 75011 Paris, France (mar.pujades@epicentre.msf.org).

Author Contributions: Dr Pujades-Rodriguez had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Pujades-Rodriguez, Calmy.

Acquisition of data: Pujades-Rodriguez, Balkan, Arnould.

Analysis and interpretation of data: Pujades-Rodriguez, Balkan, Brinkhof, Calmy.

Drafting of the manuscript: Pujades-Rodriguez, Calmy.

Critical revision of the manuscript for important intellectual content: Pujades-Rodriguez, Balkan, Arnould, Brinkhof, Calmy.

Statistical analysis: Pujades-Rodriguez, Brinkhof.

Obtained funding: Pujades-Rodriguez, Calmy.

Administrative, technical, or material support: Pujades-Rodriguez, Balkan, Arnould, Calmy.

Study supervision: Pujades-Rodriguez, Balkan, Calmy.

Financial Disclosures: None reported.

Funding/Support: This study was partly supported by grant 32473B-122116 from the Swiss National Science Foundation, from which external funds for this particular project were provided. Salaries of Drs Pujades-Rodriguez, Balkan, and Arnould are ensured by MSF (Drs Balkan and Arnould are employed by the Medical Department of MSF; Dr Pujades-Rodriguez is employed by Epicentre, a satellite nongovernmental organization closely linked to and funded by MSF).

Role of the Sponsor: The sponsor did not participate in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review, or approval of the manuscript.

Additional Contributions: We thank our partners in the ministries of health of the countries and the MSF field teams for their daily work and efforts to provide care to the patients and for data collection; the members of the MSF AIDS Working Group for their advice and support; and the Epicentre FUCHIA team for their support in data collection and data quality maintenance. We also thank Oliver Yun, MA, ELS, Medical Editor of Doctors Without Borders/Médecins Sans Frontières (MSF), New York, New York, for his editorial work. Mr Yun, who is employed by MSF, did not receive compensation for his editorial work; he provides free editorial support to authors employed by MSF/Epicentre.

HIV/AIDS Department of the World Health Organization.  Towards universal access: scaling up priority HIV/AIDS interventions in the health sector. Progress report 2008. http://www.who.int/hiv/pub/2008progressreport/en/index.html. Accessed April 20, 2010
Calmy A, Klement E, Teck R, Berman D, Pécoul B, Ferradini L. Simplifying and adapting antiretroviral treatment in resource-poor settings: a necessary step to scaling-up.  AIDS. 2004;18(18):2353-2360
PubMed
 Rapid Advice: Antiretroviral Therapy for HIV Infection in Adults and Adolescents. Geneva, Switzerland: World Health Organization; November 2009
Katlama C, Haubrich R, Lalezari J,  et al; DUET-1, DUET-2 study groups.  Efficacy and safety of etravirine in treatment-experienced, HIV-1 patients: pooled 48 week analysis of two randomized, controlled trials.  AIDS. 2009;23(17):2289-2300
PubMed   |  Link to Article
Yazdanpanah Y, Fagard C, Descamps D,  et al.  High rate of virologic success with raltegravir plus etravirine and darunavir/ritonavir in treatment-experienced patients with multidrug-resistant virus: results of the ANRG 139 TRIO trial. Paper presented at: XVII International AIDS Conference; August 3-8, 2008; Mexico City, Mexico. Abstract THAB0406
Pujades-Rodríguez M, O’Brien D, Humblet P, Calmy A. Second-line antiretroviral therapy in resource-limited settings: the experience of Médecins Sans Frontières.  AIDS. 2008;22(11):1305-1312
PubMed   |  Link to Article
The World Health Organization.  Scaling up Antiretroviral Therapy in Resource-Limited Settings: Treatment Guidelines for a Public Health Approach. Geneva, Switzerland: World Health Organization; 2003
The World Health Organization.  Antiretroviral Therapy for HIV Infection in Adults and Adolescents in Resource-Limited Settings: Towards Universal Access: Recommendations for a Public Health Approach. Geneva, Switzerland: World Health Organization; 2006
Lima VD, Gill VS, Yip B, Hogg RS, Montaner JS, Harrigan PR. Increased resilience to the development of drug resistance with modern boosted protease inhibitor-based highly active antiretroviral therapy.  J Infect Dis. 2008;198(1):51-58
PubMed   |  Link to Article
Weidle PJ, Wamai N, Solberg P,  et al.  Adherence to antiretroviral therapy in a home-based AIDS care programme in rural Uganda.  Lancet. 2006;368(9547):1587-1594
PubMed   |  Link to Article
Paterson DL, Swindells S, Mohr J,  et al.  Adherence to protease inhibitor therapy and outcomes in patients with HIV infection.  Ann Intern Med. 2000;133(1):21-30
PubMed   |  Link to Article
Martin M, Del Cacho E, Codina C,  et al.  Relationship between adherence level, type of the antiretroviral regimen, and plasma HIV type 1 RNA viral load: a prospective cohort study.  AIDS Res Hum Retroviruses. 2008;24(10):1263-1268
PubMed   |  Link to Article
Rabe-Hesketh S. Reliable estimation of generalized linear mixed models using adaptive quadrature.  Stata J. 2002;2(1):1-21
Royston P. Multiple imputation of missing values.  Stata J. 2004;4(3):227-241
Royston P. Multiple imputation of missing values: update of ice.  Stata J. 2005;5(4):527-536
Rubin DB. Multiple Imputation for Nonresponse in Surveys. New York, NY: J Wiley & Sons; 1987
Carlin JB, Galati JC, Royston P. A framework for managing and analyzing multiple imputed data in Stata.  Stata J. 2008;8(1):49-67
Royston P, Carlin JB, White IR. Multiple imputation of missing values: new features for mim.  Stata J. 2009;9(2):252-264
Keiding N, Andersen PK, Klein JP. The role of frailty models and accelerated failure time models in describing heterogeneity due to omitted covariates.  Stat Med. 1997;16(1-3):215-224
PubMed   |  Link to Article
Gutierrez RG. Parametric frailty and shared frailty survival models.  Stata J. 2002;2(1):22-44
Feiveson AH. Power by simulation.  Stata J. 2002;2(2):107-124
Feiveson A. Calculating power by simulation. http://www.stata.com/support/faqs/stat/power.html. Updated July 2009. Accessed June 15, 2010
Pujades-Rodriguez M, Ferreyra C, Calmy A, Balkan S. Failure on First Line Therapy and Inequalities in Switching to Second Line in Adults Treated in Urban and Rural ART Programs: Multicentric Analysis in 28 MSF-Supported African and Asian Sites. Poster presented at: XVIII International AIDS Conference; July 18-23, 2010; Vienna, Austria. Abstract PE0126
Keiser O, Tweya H, Boulle A,  et al; ART-LINC of IeDEA Study Group.  Switching to second-line antiretroviral therapy in resource-limited settings: comparison of programmes with and without viral load monitoring.  AIDS. 2009;23(14):1867-1874
PubMed   |  Link to Article
Zhou J, Li PC, Kumarasamy N,  et al; TREAT Asia HIV Observational Database.  Deferred modification of antiretroviral regimen following documented treatment failure in Asia: results from the TREAT Asia HIV Observational Database (TAHOD).  HIV Med. 2010;11(1):31-39
PubMed
Ananworanich J. Reaching undetectable viral loads after intial HIV treatment.  Future HIV Therapy. 2007;1(1):81-89
Link to Article
Willig JH, Abroms S, Westfall AO,  et al.  Increased regimen durability in the era of once-daily fixed-dose combination antiretroviral therapy.  AIDS. 2008;22(15):1951-1960
PubMed   |  Link to Article
Fox MP, Ive P, Long L, Maskew M, Sanne I. High rates of survival, immune reconstitution, and virologic suppression on second-line antiretroviral therapy in South Africa.  J Acquir Immune Defic Syndr. 2010;53(4):500-506
PubMed   |  Link to Article
Hammer SM, Eron JJ Jr, Reiss P,  et al; International AIDS Society-USA.  Antiretroviral treatment of adult HIV infection: 2008 recommendations of the International AIDS Society-USA panel.  JAMA. 2008;300(5):555-570
PubMed   |  Link to Article
Petersen ML, van der Laan MJ, Napravnik S, Eron JJ, Moore RD, Deeks SG. Long-term consequences of the delay between virologic failure of highly active antiretroviral therapy and regimen modification.  AIDS. 2008;22(16):2097-2106
PubMed   |  Link to Article
Hosseinipour MC, van Oosterhout JJ, Weigel R,  et al.  The public health approach to identify antiretroviral therapy failure: high-level nucleoside reverse transcriptase inhibitor resistance among Malawians failing first-line antiretroviral therapy.  AIDS. 2009;23(9):1127-1134
PubMed   |  Link to Article
Mocroft A, Phillips AN, Miller V,  et al; EuroSIDA study group.  The use of and response to second-line protease inhibitor regimens: results from the EuroSIDA study.  AIDS. 2001;15(2):201-209
PubMed   |  Link to Article
Lewden C, Balestre E, Dabis F.Survival and loss-to-follow-up of HIV infected adults who have started antiretroviral therapy in low and middle income countries; 2009. Bordeaux, France: Institut de Santé Publique, d’Epidémiologie et de Développement (ISPED) Université Victor SegalenINSERM U897; INSERM U897
Rougemont M, Stoll BE, Elia N, Ngang P. Antiretroviral treatment adherence and its determinants in Sub-Saharan Africa: a prospective study at Yaounde Central Hospital, Cameroon.  AIDS Res Ther. 2009;6:21
PubMed   |  Link to Article
Weidle PJ, Malamba S, Mwebaze R,  et al.  Assessment of a pilot antiretroviral drug therapy programme in Uganda: patients' response, survival, and drug resistance.  Lancet. 2002;360(9326):34-40
PubMed   |  Link to Article
Calmy A, Pinoges L, Szumilin E, Zachariah R, Ford N, Ferradini L.Médecins Sans Frontieres.  Generic fixed-dose combination antiretroviral treatment in resource-poor settings: multicentric observational cohort.  AIDS. 2006;20(8):1163-1169
PubMed   |  Link to Article
Chen SC, Yu JK, Harries AD,  et al.  Increased mortality of male adults with AIDS related to poor compliance to antiretroviral therapy in Malawi.  Trop Med Int Health. 2008;13(4):513-519
PubMed   |  Link to Article
Brinkhof MW, Boulle A, Weigel R,  et al; International Epidemiological Databases to Evaluate AIDS (IeDEA).  Mortality of HIV-infected patients starting antiretroviral therapy in sub-Saharan Africa: comparison with HIV-unrelated mortality.  PLoS Med. 2009;6(4):e1000066
PubMed   |  Link to Article
Crowley S, Rollins N, Shaffer N, Guerma T, Vitoria M, Lo YR. New WHO HIV treatment and prevention guidelines.  Lancet. 2010;375(9718):874-875
PubMed   |  Link to Article

Figures

Place holder to copy figure label and caption
Figure. Cumulative Probability of Second-Line Regimen Failure, Death, and Attrition
Graphic Jump Location

The shaded area represents 95% confidence intervals for Kaplan-Meier estimates.

Tables

Table Graphic Jump LocationTable 1. Characteristics of Human Immunodeficiency Virus Programs Included in Analyses by Geographical Area
Table Graphic Jump LocationTable 2. Numbers of Patients Meeting Criteria of Treatment Failure, Time, and Rates of Failure per 1000 Person-Years at 30 Months of Second-Line Therapya
Table Graphic Jump LocationTable 3. Association Between Second-Line Failure and Programmatic and Individual Level Factorsa

References

HIV/AIDS Department of the World Health Organization.  Towards universal access: scaling up priority HIV/AIDS interventions in the health sector. Progress report 2008. http://www.who.int/hiv/pub/2008progressreport/en/index.html. Accessed April 20, 2010
Calmy A, Klement E, Teck R, Berman D, Pécoul B, Ferradini L. Simplifying and adapting antiretroviral treatment in resource-poor settings: a necessary step to scaling-up.  AIDS. 2004;18(18):2353-2360
PubMed
 Rapid Advice: Antiretroviral Therapy for HIV Infection in Adults and Adolescents. Geneva, Switzerland: World Health Organization; November 2009
Katlama C, Haubrich R, Lalezari J,  et al; DUET-1, DUET-2 study groups.  Efficacy and safety of etravirine in treatment-experienced, HIV-1 patients: pooled 48 week analysis of two randomized, controlled trials.  AIDS. 2009;23(17):2289-2300
PubMed   |  Link to Article
Yazdanpanah Y, Fagard C, Descamps D,  et al.  High rate of virologic success with raltegravir plus etravirine and darunavir/ritonavir in treatment-experienced patients with multidrug-resistant virus: results of the ANRG 139 TRIO trial. Paper presented at: XVII International AIDS Conference; August 3-8, 2008; Mexico City, Mexico. Abstract THAB0406
Pujades-Rodríguez M, O’Brien D, Humblet P, Calmy A. Second-line antiretroviral therapy in resource-limited settings: the experience of Médecins Sans Frontières.  AIDS. 2008;22(11):1305-1312
PubMed   |  Link to Article
The World Health Organization.  Scaling up Antiretroviral Therapy in Resource-Limited Settings: Treatment Guidelines for a Public Health Approach. Geneva, Switzerland: World Health Organization; 2003
The World Health Organization.  Antiretroviral Therapy for HIV Infection in Adults and Adolescents in Resource-Limited Settings: Towards Universal Access: Recommendations for a Public Health Approach. Geneva, Switzerland: World Health Organization; 2006
Lima VD, Gill VS, Yip B, Hogg RS, Montaner JS, Harrigan PR. Increased resilience to the development of drug resistance with modern boosted protease inhibitor-based highly active antiretroviral therapy.  J Infect Dis. 2008;198(1):51-58
PubMed   |  Link to Article
Weidle PJ, Wamai N, Solberg P,  et al.  Adherence to antiretroviral therapy in a home-based AIDS care programme in rural Uganda.  Lancet. 2006;368(9547):1587-1594
PubMed   |  Link to Article
Paterson DL, Swindells S, Mohr J,  et al.  Adherence to protease inhibitor therapy and outcomes in patients with HIV infection.  Ann Intern Med. 2000;133(1):21-30
PubMed   |  Link to Article
Martin M, Del Cacho E, Codina C,  et al.  Relationship between adherence level, type of the antiretroviral regimen, and plasma HIV type 1 RNA viral load: a prospective cohort study.  AIDS Res Hum Retroviruses. 2008;24(10):1263-1268
PubMed   |  Link to Article
Rabe-Hesketh S. Reliable estimation of generalized linear mixed models using adaptive quadrature.  Stata J. 2002;2(1):1-21
Royston P. Multiple imputation of missing values.  Stata J. 2004;4(3):227-241
Royston P. Multiple imputation of missing values: update of ice.  Stata J. 2005;5(4):527-536
Rubin DB. Multiple Imputation for Nonresponse in Surveys. New York, NY: J Wiley & Sons; 1987
Carlin JB, Galati JC, Royston P. A framework for managing and analyzing multiple imputed data in Stata.  Stata J. 2008;8(1):49-67
Royston P, Carlin JB, White IR. Multiple imputation of missing values: new features for mim.  Stata J. 2009;9(2):252-264
Keiding N, Andersen PK, Klein JP. The role of frailty models and accelerated failure time models in describing heterogeneity due to omitted covariates.  Stat Med. 1997;16(1-3):215-224
PubMed   |  Link to Article
Gutierrez RG. Parametric frailty and shared frailty survival models.  Stata J. 2002;2(1):22-44
Feiveson AH. Power by simulation.  Stata J. 2002;2(2):107-124
Feiveson A. Calculating power by simulation. http://www.stata.com/support/faqs/stat/power.html. Updated July 2009. Accessed June 15, 2010
Pujades-Rodriguez M, Ferreyra C, Calmy A, Balkan S. Failure on First Line Therapy and Inequalities in Switching to Second Line in Adults Treated in Urban and Rural ART Programs: Multicentric Analysis in 28 MSF-Supported African and Asian Sites. Poster presented at: XVIII International AIDS Conference; July 18-23, 2010; Vienna, Austria. Abstract PE0126
Keiser O, Tweya H, Boulle A,  et al; ART-LINC of IeDEA Study Group.  Switching to second-line antiretroviral therapy in resource-limited settings: comparison of programmes with and without viral load monitoring.  AIDS. 2009;23(14):1867-1874
PubMed   |  Link to Article
Zhou J, Li PC, Kumarasamy N,  et al; TREAT Asia HIV Observational Database.  Deferred modification of antiretroviral regimen following documented treatment failure in Asia: results from the TREAT Asia HIV Observational Database (TAHOD).  HIV Med. 2010;11(1):31-39
PubMed
Ananworanich J. Reaching undetectable viral loads after intial HIV treatment.  Future HIV Therapy. 2007;1(1):81-89
Link to Article
Willig JH, Abroms S, Westfall AO,  et al.  Increased regimen durability in the era of once-daily fixed-dose combination antiretroviral therapy.  AIDS. 2008;22(15):1951-1960
PubMed   |  Link to Article
Fox MP, Ive P, Long L, Maskew M, Sanne I. High rates of survival, immune reconstitution, and virologic suppression on second-line antiretroviral therapy in South Africa.  J Acquir Immune Defic Syndr. 2010;53(4):500-506
PubMed   |  Link to Article
Hammer SM, Eron JJ Jr, Reiss P,  et al; International AIDS Society-USA.  Antiretroviral treatment of adult HIV infection: 2008 recommendations of the International AIDS Society-USA panel.  JAMA. 2008;300(5):555-570
PubMed   |  Link to Article
Petersen ML, van der Laan MJ, Napravnik S, Eron JJ, Moore RD, Deeks SG. Long-term consequences of the delay between virologic failure of highly active antiretroviral therapy and regimen modification.  AIDS. 2008;22(16):2097-2106
PubMed   |  Link to Article
Hosseinipour MC, van Oosterhout JJ, Weigel R,  et al.  The public health approach to identify antiretroviral therapy failure: high-level nucleoside reverse transcriptase inhibitor resistance among Malawians failing first-line antiretroviral therapy.  AIDS. 2009;23(9):1127-1134
PubMed   |  Link to Article
Mocroft A, Phillips AN, Miller V,  et al; EuroSIDA study group.  The use of and response to second-line protease inhibitor regimens: results from the EuroSIDA study.  AIDS. 2001;15(2):201-209
PubMed   |  Link to Article
Lewden C, Balestre E, Dabis F.Survival and loss-to-follow-up of HIV infected adults who have started antiretroviral therapy in low and middle income countries; 2009. Bordeaux, France: Institut de Santé Publique, d’Epidémiologie et de Développement (ISPED) Université Victor SegalenINSERM U897; INSERM U897
Rougemont M, Stoll BE, Elia N, Ngang P. Antiretroviral treatment adherence and its determinants in Sub-Saharan Africa: a prospective study at Yaounde Central Hospital, Cameroon.  AIDS Res Ther. 2009;6:21
PubMed   |  Link to Article
Weidle PJ, Malamba S, Mwebaze R,  et al.  Assessment of a pilot antiretroviral drug therapy programme in Uganda: patients' response, survival, and drug resistance.  Lancet. 2002;360(9326):34-40
PubMed   |  Link to Article
Calmy A, Pinoges L, Szumilin E, Zachariah R, Ford N, Ferradini L.Médecins Sans Frontieres.  Generic fixed-dose combination antiretroviral treatment in resource-poor settings: multicentric observational cohort.  AIDS. 2006;20(8):1163-1169
PubMed   |  Link to Article
Chen SC, Yu JK, Harries AD,  et al.  Increased mortality of male adults with AIDS related to poor compliance to antiretroviral therapy in Malawi.  Trop Med Int Health. 2008;13(4):513-519
PubMed   |  Link to Article
Brinkhof MW, Boulle A, Weigel R,  et al; International Epidemiological Databases to Evaluate AIDS (IeDEA).  Mortality of HIV-infected patients starting antiretroviral therapy in sub-Saharan Africa: comparison with HIV-unrelated mortality.  PLoS Med. 2009;6(4):e1000066
PubMed   |  Link to Article
Crowley S, Rollins N, Shaffer N, Guerma T, Vitoria M, Lo YR. New WHO HIV treatment and prevention guidelines.  Lancet. 2010;375(9718):874-875
PubMed   |  Link to Article
CME
Meets CME requirements for:
Browse CME for all U.S. States
Accreditation Information
The American Medical Association is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The AMA designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 CreditTM per course. Physicians should claim only the credit commensurate with the extent of their participation in the activity. Physicians who complete the CME course and score at least 80% correct on the quiz are eligible for AMA PRA Category 1 CreditTM.
Note: You must get at least of the answers correct to pass this quiz.
You have not filled in all the answers to complete this quiz
The following questions were not answered:
Sorry, you have unsuccessfully completed this CME quiz with a score of
The following questions were not answered correctly:
Commitment to Change (optional):
Indicate what change(s) you will implement in your practice, if any, based on this CME course.
Your quiz results:
The filled radio buttons indicate your responses. The preferred responses are highlighted
For CME Course: A Proposed Model for Initial Assessment and Management of Acute Heart Failure Syndromes
Indicate what changes(s) you will implement in your practice, if any, based on this CME course.
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Multimedia

Some tools below are only available to our subscribers or users with an online account.

Web of Science® Times Cited: 29

Related Content

Customize your page view by dragging & repositioning the boxes below.

Articles Related By Topic
Related Topics
PubMed Articles
JAMAevidence.com