Journal Article > ResearchFull Text
BMC Infect Dis. 29 April 2016; Volume 16 (Issue 1); DOI:10.1186/s12879-016-1520-4
Blaizot S, Maman D, Riche B, Mukui I, Kirubi B, et al.
BMC Infect Dis. 29 April 2016; Volume 16 (Issue 1); DOI:10.1186/s12879-016-1520-4
Multiple prevention interventions, including early antiretroviral therapy initiation, may reduce HIV incidence in hyperendemic settings. Our aim was to predict the short-term impact of various single and combined interventions on HIV spreading in the adult population of Ndhiwa subcounty (Nyanza Province, Kenya).
Journal Article > ResearchFull Text
PLOS One. 23 February 2012; Volume 7 (Issue 2); e31706.; DOI:10.1371/journal.pone.0031706
Henriques J, Pujades M, McGuire M, Szumilin E, Iwaz J, et al.
PLOS One. 23 February 2012; Volume 7 (Issue 2); e31706.; DOI:10.1371/journal.pone.0031706
OBJECTIVE
The evaluation of HIV treatment programs is generally based on an estimation of survival among patients receiving antiretroviral treatment (ART). In large HIV programs, loss to follow-up (LFU) rates remain high despite active patient tracing, which is likely to bias survival estimates and survival regression analyses.
METHODS
We compared uncorrected survival estimates derived from routine program data with estimates obtained by applying six correction methods that use updated outcome data by a field survey targeting LFU patients in a rural HIV program in Malawi. These methods were based on double-sampling and differed according to the weights given to survival estimates in LFU and non-LFU subpopulations. We then proposed a correction of the survival regression analysis.
RESULTS
Among 6,727 HIV-infected adults receiving ART, 9% were LFU after one year. The uncorrected survival estimates from routine data were 91% in women and 84% in men. According to increasing sophistication of the correction methods, the corrected survival estimates ranged from 89% to 85% in women and 82% to 77% in men. The estimates derived from uncorrected regression analyses were highly biased for initial tuberculosis mortality ratios (RR; 95% CI: 1.07; 0.76-1.50 vs. 2.06 to 2.28 with different correction weights), Kaposi sarcoma diagnosis (2.11; 1.61-2.76 vs. 2.64 to 3.9), and year of ART initiation (1.40; 1.17-1.66 vs. 1.29 to 1.34).
CONCLUSIONS
In HIV programs with high LFU rates, the use of correction methods based on non-exhaustive double-sampling data are necessary to minimise the bias in survival estimates and survival regressions.
The evaluation of HIV treatment programs is generally based on an estimation of survival among patients receiving antiretroviral treatment (ART). In large HIV programs, loss to follow-up (LFU) rates remain high despite active patient tracing, which is likely to bias survival estimates and survival regression analyses.
METHODS
We compared uncorrected survival estimates derived from routine program data with estimates obtained by applying six correction methods that use updated outcome data by a field survey targeting LFU patients in a rural HIV program in Malawi. These methods were based on double-sampling and differed according to the weights given to survival estimates in LFU and non-LFU subpopulations. We then proposed a correction of the survival regression analysis.
RESULTS
Among 6,727 HIV-infected adults receiving ART, 9% were LFU after one year. The uncorrected survival estimates from routine data were 91% in women and 84% in men. According to increasing sophistication of the correction methods, the corrected survival estimates ranged from 89% to 85% in women and 82% to 77% in men. The estimates derived from uncorrected regression analyses were highly biased for initial tuberculosis mortality ratios (RR; 95% CI: 1.07; 0.76-1.50 vs. 2.06 to 2.28 with different correction weights), Kaposi sarcoma diagnosis (2.11; 1.61-2.76 vs. 2.64 to 3.9), and year of ART initiation (1.40; 1.17-1.66 vs. 1.29 to 1.34).
CONCLUSIONS
In HIV programs with high LFU rates, the use of correction methods based on non-exhaustive double-sampling data are necessary to minimise the bias in survival estimates and survival regressions.
Journal Article > ResearchFull Text
BMC Infect Dis. 22 January 2013; Volume 13; 27.; DOI:10.1186/1471-2334-13-27
Bastard M, Soulinphumy K, Phimmasone P, Saadani A, Ciaffi L, et al.
BMC Infect Dis. 22 January 2013; Volume 13; 27.; DOI:10.1186/1471-2334-13-27
BACKGROUND
In April 2003, Médecins Sans Frontières launched an HIV/AIDS programme to provide free HAART to HIV-infected patients in Laos. Although HIV prevalence is estimated as low in this country, it has been increasing in the last years. This work reports the first results of an observational cohort study and it aims to identify the principal determinants of the CD4 cells evolution and to assess mortality among patients on HAART.
METHODS
We performed a retrospective database analysis on patients initiated on HAART between 2003 and 2009 (CD4<200cells/μL or WHO stage 4). We excluded from the analysis patients who were less than 16 years old and pregnant women. To explore the determinants of the CD4 reconstitution, a linear mixed model was adjusted. To identify typical trajectories of the CD4 cells, a latent trajectory analysis was carried out. Finally, a Cox proportional-hazards model was used to reveal predictors of mortality on HAART including appointment delay greater than 1 day.
RESULTS
A total of 1365 patients entered the programme and 913 (66.9%) received an HAART with a median CD4 of 49 cells/μL [IQR 15–148]. High baseline CD4 cell count and female gender were associated with a higher CD4 level over time. In addition, this gender difference increased over time. Two typical latent CD4 trajectories were revealed showing that 31% of women against 22% of men followed a high CD4 trajectory. In the long-term, women were more likely to attend appointments without delay. Mortality reached 6.2% (95% CI 4.8-8.0%) at 4 months and 9.1% (95% CI 7.3-11.3%) at 1 year. Female gender (HR=0.17, 95% CI 0.07-0.44) and high CD4 trajectory (HR=0.19, 95% CI 0.08-0.47) were independently associated with a lower death rate.
CONCLUSIONS
Patients who initiated HAART were severely immunocompromised yielding to a high early mortality. In the long-term on HAART, women achieved a better CD4 cells reconstitution than men and were less likely to die. This study highlights important differences between men and women regarding response to HAART and medical care, and questions men’s compliance to treatment.
In April 2003, Médecins Sans Frontières launched an HIV/AIDS programme to provide free HAART to HIV-infected patients in Laos. Although HIV prevalence is estimated as low in this country, it has been increasing in the last years. This work reports the first results of an observational cohort study and it aims to identify the principal determinants of the CD4 cells evolution and to assess mortality among patients on HAART.
METHODS
We performed a retrospective database analysis on patients initiated on HAART between 2003 and 2009 (CD4<200cells/μL or WHO stage 4). We excluded from the analysis patients who were less than 16 years old and pregnant women. To explore the determinants of the CD4 reconstitution, a linear mixed model was adjusted. To identify typical trajectories of the CD4 cells, a latent trajectory analysis was carried out. Finally, a Cox proportional-hazards model was used to reveal predictors of mortality on HAART including appointment delay greater than 1 day.
RESULTS
A total of 1365 patients entered the programme and 913 (66.9%) received an HAART with a median CD4 of 49 cells/μL [IQR 15–148]. High baseline CD4 cell count and female gender were associated with a higher CD4 level over time. In addition, this gender difference increased over time. Two typical latent CD4 trajectories were revealed showing that 31% of women against 22% of men followed a high CD4 trajectory. In the long-term, women were more likely to attend appointments without delay. Mortality reached 6.2% (95% CI 4.8-8.0%) at 4 months and 9.1% (95% CI 7.3-11.3%) at 1 year. Female gender (HR=0.17, 95% CI 0.07-0.44) and high CD4 trajectory (HR=0.19, 95% CI 0.08-0.47) were independently associated with a lower death rate.
CONCLUSIONS
Patients who initiated HAART were severely immunocompromised yielding to a high early mortality. In the long-term on HAART, women achieved a better CD4 cells reconstitution than men and were less likely to die. This study highlights important differences between men and women regarding response to HAART and medical care, and questions men’s compliance to treatment.
Journal Article > ResearchAbstract Only
AIDS Res Hum Retroviruses. 8 November 2016; Volume 33 (Issue 5); DOI:10.1089/AID.2016.0123
Blaizot S, Kim AA, Zeh C, Riche B, Maman D, et al.
AIDS Res Hum Retroviruses. 8 November 2016; Volume 33 (Issue 5); DOI:10.1089/AID.2016.0123
OBJECTIVES
Estimating HIV incidence is critical for identifying groups at risk for HIV infection, planning and targeting interventions, and evaluating these interventions over time. The use of reliable estimation methods for HIV incidence is thus of high importance. The aim of this study was to compare methods for estimating HIV incidence in a population-based cross-sectional survey.
DESIGN/METHODS
The incidence estimation methods evaluated included assay-derived methods, a testing history-derived method, and a probability-based method applied to data from the Ndhiwa HIV Impact in Population Survey (NHIPS). Incidence rates by sex and age and cumulative incidence as a function of age were presented.
RESULTS
HIV incidence ranged from 1.38 [95% confidence interval (CI) 0.67-2.09] to 3.30 [95% CI 2.78-3.82] per 100 person-years overall; 0.59 [95% CI 0.00-1.34] to 2.89 [95% CI 0.86-6.45] in men; and 1.62 [95% CI 0.16-6.04] to 4.03 [95% CI 3.30-4.77] per 100 person-years in women. Women had higher incidence rates than men for all methods. Incidence rates were highest among women aged 15-24 and 25-34 years and highest among men aged 25-34 years.
CONCLUSION
Comparison of different methods showed variations in incidence estimates, but they were in agreement to identify most-at-risk groups. The use and comparison of several distinct approaches for estimating incidence are important to provide the best-supported estimate of HIV incidence in the population.
Estimating HIV incidence is critical for identifying groups at risk for HIV infection, planning and targeting interventions, and evaluating these interventions over time. The use of reliable estimation methods for HIV incidence is thus of high importance. The aim of this study was to compare methods for estimating HIV incidence in a population-based cross-sectional survey.
DESIGN/METHODS
The incidence estimation methods evaluated included assay-derived methods, a testing history-derived method, and a probability-based method applied to data from the Ndhiwa HIV Impact in Population Survey (NHIPS). Incidence rates by sex and age and cumulative incidence as a function of age were presented.
RESULTS
HIV incidence ranged from 1.38 [95% confidence interval (CI) 0.67-2.09] to 3.30 [95% CI 2.78-3.82] per 100 person-years overall; 0.59 [95% CI 0.00-1.34] to 2.89 [95% CI 0.86-6.45] in men; and 1.62 [95% CI 0.16-6.04] to 4.03 [95% CI 3.30-4.77] per 100 person-years in women. Women had higher incidence rates than men for all methods. Incidence rates were highest among women aged 15-24 and 25-34 years and highest among men aged 25-34 years.
CONCLUSION
Comparison of different methods showed variations in incidence estimates, but they were in agreement to identify most-at-risk groups. The use and comparison of several distinct approaches for estimating incidence are important to provide the best-supported estimate of HIV incidence in the population.
Journal Article > ResearchFull Text
PLOS One. 21 November 2011; Volume 6 (Issue 11); e28112.; DOI:10.1371/journal.pone.0028112
Pujades-Rodriguez M, Dantony E, Pinoges LLP, Ecochard R, Etard JF, et al.
PLOS One. 21 November 2011; Volume 6 (Issue 11); e28112.; DOI:10.1371/journal.pone.0028112
BACKGROUND
To compare the incidence and timing of toxicity associated with the use of a reduced dose of stavudine from 40 to 30 mg in first-line antiretroviral therapy (ART) for HIV treatment and to investigate associated risk factors.
METHODS
Multicohort study including 23 HIV programs in resource-limited countries. Adults enrolled between January 2005 and December 2009. Four-year rates of all-cause and stavudine-specific toxicity were estimated. Multilevel mixed-effect Poisson and accelerated failure models were used to investigate factors associated with toxicity and timing of diagnosis.
FINDINGS
A total of 48,785 patients contributed 62,505 person-years of follow-up. Rate of all-cause toxicity was 7.80 (95%CI 7.59-8.03) per 100 person-years, but varied greatly across sites (range 0.41-21.76). Patients treated with stavudine 40 mg had higher rates of toxicity (adjusted rate ratio [aRR] 1.18, 95%CI 1.06-1.30 during the first year of ART; and 1.51, 95%CI 1.32-1.71 during the second year). Women, older age, initial advanced clinical stage, and low CD4 count were associated with increased toxicity rate ratios. Timing of lipodystrophy and peripheral neuropathy diagnosis were 12% and 13% shorter, respectively, in patients treated with stavudine 40 mg than in those receiving 30 mg stavudine dose (P = 0.03 and 0.07, respectively). INSTERPRETATION: Higher rates of drug-related toxicity were reported in patients receiving stavudine 40 mg compared with 30 mg, and the time to toxicity diagnosis was shorter in patients treated with the higher dose. Higher rates of toxicity were observed during the first two years of ART.
To compare the incidence and timing of toxicity associated with the use of a reduced dose of stavudine from 40 to 30 mg in first-line antiretroviral therapy (ART) for HIV treatment and to investigate associated risk factors.
METHODS
Multicohort study including 23 HIV programs in resource-limited countries. Adults enrolled between January 2005 and December 2009. Four-year rates of all-cause and stavudine-specific toxicity were estimated. Multilevel mixed-effect Poisson and accelerated failure models were used to investigate factors associated with toxicity and timing of diagnosis.
FINDINGS
A total of 48,785 patients contributed 62,505 person-years of follow-up. Rate of all-cause toxicity was 7.80 (95%CI 7.59-8.03) per 100 person-years, but varied greatly across sites (range 0.41-21.76). Patients treated with stavudine 40 mg had higher rates of toxicity (adjusted rate ratio [aRR] 1.18, 95%CI 1.06-1.30 during the first year of ART; and 1.51, 95%CI 1.32-1.71 during the second year). Women, older age, initial advanced clinical stage, and low CD4 count were associated with increased toxicity rate ratios. Timing of lipodystrophy and peripheral neuropathy diagnosis were 12% and 13% shorter, respectively, in patients treated with stavudine 40 mg than in those receiving 30 mg stavudine dose (P = 0.03 and 0.07, respectively). INSTERPRETATION: Higher rates of drug-related toxicity were reported in patients receiving stavudine 40 mg compared with 30 mg, and the time to toxicity diagnosis was shorter in patients treated with the higher dose. Higher rates of toxicity were observed during the first two years of ART.
Journal Article > ResearchAbstract
Stat Methods Med Res. 1 September 2014; Volume 26 (Issue 1); DOI:10.1177/0962280214549040
Subtil F, Boussari O, Bastard M, Etard JF, Ecochard R, et al.
Stat Methods Med Res. 1 September 2014; Volume 26 (Issue 1); DOI:10.1177/0962280214549040
Classifying patients according to longitudinal measures, or trajectory classification, has become frequent in clinical research. The k-means algorithm is increasingly used for this task in case of continuous variables with standard deviations that do not depend on the mean. One feature of count and binary data modeled by Poisson or logistic regression is that the variance depends on the mean; hence, the within-group variability changes from one group to another depending on the mean trajectory level. Mixture modeling could be used here for classification though its main purpose is to model the data. The results obtained may change according to the main objective. This article presents an extension of the k-means algorithm that takes into account the features of count and binary data by using the deviance as distance metric. This approach is justified by its analogy with the classification likelihood. Two applications are presented with binary and count data to show the differences between the classifications obtained with the usual Euclidean distance versus the deviance distance.
Journal Article > ResearchFull Text
PLOS One. 19 June 2015; Volume 10 (Issue 6); e0130387.; DOI:10.1371/journal.pone.0130387
Blaizot S, Riche B, Maman D, Mukui I, Kirubi B, et al.
PLOS One. 19 June 2015; Volume 10 (Issue 6); e0130387.; DOI:10.1371/journal.pone.0130387
BACKGROUND
Mathematical models have played important roles in the understanding of epidemics and in the study of the impacts of various behavioral or medical measures. However, modeling accurately the future spread of an epidemic requires context-specific parameters that are difficult to estimate because of lack of data. Our objective is to propose a methodology to estimate context-specific parameters using Demographic and Health Survey (DHS)-like data that can be used in mathematical modeling of short-term HIV spreading.
METHODS AND FINDINGS
The model splits the population according to sex, age, HIV status, and antiretroviral treatment status. To estimate context-specific parameters, we used individuals' histories included in DHS-like data and a statistical analysis that used decomposition of the Poisson likelihood. To predict the course of the HIV epidemic, sex- and age-specific differential equations were used. This approach was applied to recent data from Kenya. The approach allowed the estimation of several key epidemiological parameters. Women had a higher infection rate than men and the highest infection rate in the youngest age groups (15-24 and 25-34 years) whereas men had the highest infection rate in age group 25-34 years. The immunosuppression rates were similar between age groups. The treatment rate was the highest in age group 35-59 years in both sexes. The results showed that, within the 15-24 year age group, increasing male circumcision coverage and antiretroviral therapy coverage at CD4 ≤ 350/mm3 over the current 70% could have short-term impacts.
CONCLUSIONS
The study succeeded in estimating the model parameters using DHS-like data rather than literature data. The analysis provides a framework for using the same data for estimation and prediction, which can improve the validity of context-specific predictions and help designing HIV prevention campaigns.
Mathematical models have played important roles in the understanding of epidemics and in the study of the impacts of various behavioral or medical measures. However, modeling accurately the future spread of an epidemic requires context-specific parameters that are difficult to estimate because of lack of data. Our objective is to propose a methodology to estimate context-specific parameters using Demographic and Health Survey (DHS)-like data that can be used in mathematical modeling of short-term HIV spreading.
METHODS AND FINDINGS
The model splits the population according to sex, age, HIV status, and antiretroviral treatment status. To estimate context-specific parameters, we used individuals' histories included in DHS-like data and a statistical analysis that used decomposition of the Poisson likelihood. To predict the course of the HIV epidemic, sex- and age-specific differential equations were used. This approach was applied to recent data from Kenya. The approach allowed the estimation of several key epidemiological parameters. Women had a higher infection rate than men and the highest infection rate in the youngest age groups (15-24 and 25-34 years) whereas men had the highest infection rate in age group 25-34 years. The immunosuppression rates were similar between age groups. The treatment rate was the highest in age group 35-59 years in both sexes. The results showed that, within the 15-24 year age group, increasing male circumcision coverage and antiretroviral therapy coverage at CD4 ≤ 350/mm3 over the current 70% could have short-term impacts.
CONCLUSIONS
The study succeeded in estimating the model parameters using DHS-like data rather than literature data. The analysis provides a framework for using the same data for estimation and prediction, which can improve the validity of context-specific predictions and help designing HIV prevention campaigns.
Journal Article > ResearchFull Text
BMC Infect Dis. 26 July 2017; Volume 17 (Issue 1); DOI:10.1186/s12879-017-2612-5
Blaizot S, Huerga H, Riche B, Ellman T, Shroufi A, et al.
BMC Infect Dis. 26 July 2017; Volume 17 (Issue 1); DOI:10.1186/s12879-017-2612-5
Combined prevention interventions, including early antiretroviral therapy initiation, may substantially reduce HIV incidence in hyperendemic settings. Our aim was to assess the potential short-term impact of combined interventions on HIV spreading in the adult population of Mbongolwane and Eshowe (KwaZulu-Natal, South Africa) using sex- and age-specific scenarios, and age-targeted interventions.
Journal Article > ResearchFull Text
AIDS. 17 July 2012; Volume 26 (Issue 11); 1393-8.; DOI:10.1097/QAD.0b013e328352d054
Maman D, Pujades-Rodriguez M, Nicholas S, McGuire M, Szumilin E, et al.
AIDS. 17 July 2012; Volume 26 (Issue 11); 1393-8.; DOI:10.1097/QAD.0b013e328352d054
OBJECTIVE
We investigated the association between immune response and mortality in four HIV African programs supported by Médecins Sans Frontières.
DESIGN
Multicentric retrospective cohort study.
METHODS
All antiretroviral therapy (ART) naive adults (>15 years) who initiated therapy between March 2001 and November 2010 and receiving therapy for 9 months or more were included. We described the evolution of mortality over time. Mixed Poisson models were used to assess the effect of updated CD4 cell counts and other potential risk factors on mortality.
FINDINGS
A total of 24 037 patients, of which 68% were women, contributed 69 516.2 person-years of follow-up. At ART initiation, 5718 patients (23.7%) were classified as WHO clinical stage 4, 1587 (6.6%) had a BMI below 16 kg/m and 2568 (10.7%) had CD4 cell count below 50 cells/μl. A total of 568 (2.4%) deaths were recorded during the study period. In the CD4 response categories 500 cells/μl or more, 350-499, 200-349, 50-199 cells/μl and less than 50 cells/μl, unadjusted mortality rates were 0.36; 0.58; 0.88; 1.91 and 7.43 per 100 person-years, respectively. In multivariate analysis, higher mortality was observed in patients with CD4 response levels 350-499 cells/μl [adjusted hazard ratio (aHR) 1.70, 95% confidence interval (CI) 1.26-2.30] and for those between 200-349 (aHR 2.56; 95% CI 1.93-3.38), compared to those with 500 cells/μl or more.
INTERPRETATION
The observed higher survival of patients with a CD4 response to ART higher than 500 cells/μl supports the need of further research to evaluate the individual benefit of initiating ART at higher CD4 levels in sub-Saharan Africa.
We investigated the association between immune response and mortality in four HIV African programs supported by Médecins Sans Frontières.
DESIGN
Multicentric retrospective cohort study.
METHODS
All antiretroviral therapy (ART) naive adults (>15 years) who initiated therapy between March 2001 and November 2010 and receiving therapy for 9 months or more were included. We described the evolution of mortality over time. Mixed Poisson models were used to assess the effect of updated CD4 cell counts and other potential risk factors on mortality.
FINDINGS
A total of 24 037 patients, of which 68% were women, contributed 69 516.2 person-years of follow-up. At ART initiation, 5718 patients (23.7%) were classified as WHO clinical stage 4, 1587 (6.6%) had a BMI below 16 kg/m and 2568 (10.7%) had CD4 cell count below 50 cells/μl. A total of 568 (2.4%) deaths were recorded during the study period. In the CD4 response categories 500 cells/μl or more, 350-499, 200-349, 50-199 cells/μl and less than 50 cells/μl, unadjusted mortality rates were 0.36; 0.58; 0.88; 1.91 and 7.43 per 100 person-years, respectively. In multivariate analysis, higher mortality was observed in patients with CD4 response levels 350-499 cells/μl [adjusted hazard ratio (aHR) 1.70, 95% confidence interval (CI) 1.26-2.30] and for those between 200-349 (aHR 2.56; 95% CI 1.93-3.38), compared to those with 500 cells/μl or more.
INTERPRETATION
The observed higher survival of patients with a CD4 response to ART higher than 500 cells/μl supports the need of further research to evaluate the individual benefit of initiating ART at higher CD4 levels in sub-Saharan Africa.