Abstract
Background:
Antiretroviral therapy (ART) for HIV-infected sex workers is an important HIV prevention strategy. However, sex workers may have additional challenges for retention in ART care. The objectives of this study were to assess retention of sex workers on ART in a routine setting in Ivory Coast and identify risk factors for loss to follow-up (LTFU).
Methods:
The design was a retrospective cohort study. An analysis of clinic files was conducted in 2 sites providing ART services to sex workers in Ivory Coast. Demographic, behavior, and clinical data of female and male sex workers on ART were abstracted onto a standardized anonymous data collection form. Data collection took place between May 11 and 28, 2010.
Results:
A total of 376 female and 38 male sex workers were included in the analysis. The retention probability was 75% at 6 months, 68% at 12 months, 55% at 24 months, and 47% at 36 months. Attrition was mainly because of LTFU. Factors significantly associated with LTFU in bivariate analysis were lower schooling level, later calendar year of starting ART, and not receiving initial adherence counseling. Later year of starting ART and not receiving adherence counseling at ART initiation remained significantly associated with LTFU in a multivariate Cox regression model.
Conclusions:
To improve the retention of sex workers on ART, there is a need for more in-depth investigation of the role of pre-ART counseling and the increasing rates of LTFU with each calendar year.
Keywords: ART, sex workers, retention on ART, loss to follow-up
INTRODUCTION
Advances in HIV treatment have resulted in substantial increases in longevity and quality of life among people living with HIV/AIDS in high-income countries.1 In recent years, access to antiretroviral therapy (ART) in low- and middle-income countries has greatly improved, and impressive progress has been made with 7.5 million people receiving ART in the African region by the end of 2012.2 ART is not only improving quality of life and decreasing morbidity in those receiving treatment, but the public health impact of scaling up ART may contribute significantly to prevention efforts.3,4 A recent study has shown that early provision of ART reduced transmission of HIV in discordant couples by 96%.5 The results of that study and other recent studies show the enormous potential of treatment for prevention.6,7
The ability to successfully engage key hard-to-reach populations, such as sex workers, in HIV care and treatment programs is important to maximize effectiveness of treatment as prevention strategies.8 Providing ART to sex workers may have an important impact on the HIV epidemic in many countries. However, there are many hurdles to effective treatment, one of the major ones being retention on ART. Sex workers have additional challenges, including their mobility, irregular working hours, and difficult or no family relationships, that increase risk for loss to follow-up (LTFU).9
Other studies have shown effectiveness of ART and high retention rates of sex workers in ART care in research settings.8 These results may not be representative for sex workers on ART in a nonresearch “routine” care setting for high-risk populations.
Ivory Coast in West Africa has a generalized HIV epidemic with an HIV prevalence of 3.7% among the general population in 2005.10 In collaboration with multiple partners, the Ministry of Health rapidly scaled up ART services between 2004 and 2007. More than 10,000 patients started ART during that time and >30,000 HIV-infected patients were receiving ART by 2007.11 Collaboration with partners, decentralization, and integration with primary care systems have been important strategies used by the Ministry of Health to rapidly improve access to HIV care and treatment nationally. On a routine basis, site-specific aggregate program data and limited demographic and clinical information on patients receiving ART are reported.
Sex workers bear a disproportionate burden of the HIV epidemic in Ivory Coast, and HIV prevalence as high as 50% among male and 31% among female sex workers has been found.12,13 A large-scale prevention and care program for sex workers named “Projet d’Assistance aux Populations Hautement Vulnérables (PAPO-HV)” was set up in 2004, with the financial support of the President’s Emergency Plan for AIDS Relief (PEPFAR). Services for sex workers include condom promotion, screening, and management of sexually transmitted infections, HIV testing and counseling, HIV care and support services, and access to ART. Specific counseling and referral services are available for sex workers younger than 18 years. Services were gradually scaled up, and by 2008, sex worker clinics were operational in 10 of the largest cities in the country. At the same time, 7 of the sex worker clinics were providing ART in accordance with the national ART guidelines. Although the exact magnitude of the problem is not known, LTFU is reported frequently by the clinic staff and should be addressed to increase effectiveness of the ART program among sex workers.
The objectives of this study were to assess retention of sex workers on ART in Ivory Coast and to identify risk factors for LTFU.
METHODS
Study Design and Procedures
A retrospective cohort analysis of clinic files was conducted in 2010. All ART sites, supported by Project PAPO-HV that had supported 50 or more sex workers on ART, as of December 31, 2007, were eligible for inclusion in the study. Of a total of 7 PAPO-HV sites, which had ever provided ART services to adults, only 2 had provided ART services in 2007. The two sites were Clinique de Confiance in Abidjan and Centre Espérance in San Pedro. Five other sites started to provide ART to sex workers in 2008 but accounted only for 38% of the total number sex workers on ART by December 2009 (with 7–31 sex workers per clinic).
At these 2 clinics, all medical records belonging to female and male sex workers aged 15 years and older at the time of ART initiation, who had initiated ART at least 6 months before the date of chart abstraction, were eligible for inclusion. Before beginning data collection, all data abstractors underwent 3–5 day training on the protocol, forms, and study procedures. For each eligible patient, data were abstracted from 3 different sources onto a standardized data collection form without personal identifiers. Clinical data were abstracted from standardized Ministry of Health-recommended medical records. If CD4 counts were not recorded in the charts, laboratory records were reviewed. Finally, data related to sex work were abstracted from a registration form that is routinely completed for all sex workers at enrollment in the clinics. In addition to key outcome data, such as retention, reasons for attrition (death, LTFU, or stopping ART), patient demographic, and behavior characteristics were abstracted. Data collection took place between May 11 and 28, 2010.
To ensure quality of the data, all data abstraction forms were reviewed and 10% of the randomly selected charts were re-abstracted by the team facilitators.
Analyses
For statistical analyses, we used STATA (version 11.1; StataCorp LP, College Station, TX). A patient was considered retained on ART if he/she had attended the clinic for any reason, including a doctor’s appointment or pharmacy refill, within the 90 days preceding data abstraction, and was not documented to have died or stopped ART. A patient was considered LTFU if he/she had not attended the clinic within 90 days preceding data abstraction and was not documented to have died, stopped ART, or being transferred out for treatment (definition of LTFU used by the Ministry of Health, Ivory Coast). A Kaplan-Meier curve was used to define the cumulative retention probabilities at 6, 12, 24, and 36 months of follow-up. The follow-up started from the date the sex worker initiated ART treatment. Patients referred to another health facility were censored at the time of transfer.
A Cox proportional hazards regression model, setting LTFU as the primary outcome, was used to estimate unadjusted and adjusted hazard ratios, 95% confidence intervals, and P values for patient-level characteristics at ART initiation, which are potential risk factors for LTFU.1,11,14,15 To best manage missing data for a priori patient-level characteristics of interest, which were assumed to be missing at random, multiple imputations with chained equations were used to impute the missing data.16 The ice17–19 procedure in STATA was used to create 20 imputed data sets for the key outcome: LTFU. The imputation model included the event indicator, all study variables, and the Nelson-Aalen estimate of cumulative hazard.20 Final estimates were combined across the imputed data sets according to Rubin’s rules14 using the mutiple imputation of missing values procedure in STATA.21 The proportional hazards assumption was assessed using visual methods and the Grambsch and Therneau22 test.
Ethics Considerations
The protocol received approval from the ethics committees of the Ministry of Health, Ivory Coast; the Institutional Review Board of the Institute of Tropical Medicine and the ethics committee of the University of Antwerp, Belgium; the Institutional Review Board of the US Centers for Disease Control and Prevention; and Family Health International’s Protection of Human Subjects Committee, the United States. A waiver of informed consent for medical record abstraction was obtained. Although no names or other personal identifiers were abstracted, the abstractors saw names on the charts that were reviewed. The abstractors signed a form stating that they would not disclose any information from the charts.
RESULTS
Baseline Characteristics
Between August 2005 and December 2008, 384 female and 40 male sex workers started treatment in 1 of the 2 clinics. Nineteen patients (17 females and 2 males) were transferred in from other treatment centers and were excluded from the analysis. Baseline characteristics of the study participants are presented in Table 1. The median age was 32 years for both female [interquartile range (IQR), 26–37] and male (IQR, 27–34.5) sex workers. Twenty percent of the female and 66% of the male sex workers attended secondary school or higher.
TABLE 1.
Characteristics of Female and Male Sex Workers at ART Initiation in 2 Clinics in Ivory Coast
| Characteristic | FSW (N = 367) |
MSW (N = 38) |
|---|---|---|
| No. Observations (%) | No. Observations (%) | |
| Sociodemographics | ||
| Clinic site | ||
| Abidjan | 177 (48.2) | 38 (100) |
| San Pedro | 190 (51.8) | 0 (0.0) |
| Age group, yrs | ||
| <25 | 37 (10.1) | 8 (21.1) |
| 25–29 | 94 (25.6) | 8 (21.1) |
| 30–39 | 145 (39.5) | 16 (42.1) |
| ≥40 | 62 (16.9) | 4 (10.5) |
| Missing | 29 (7.9) | 2 (5.3) |
| Schooling level | ||
| Married/cohabitating | 30 (8.2) | 1 (2.6) |
| Single | 227 (61.9) | 36 (94.7) |
| Divorced | 7 (1.9) | 0 (0.0) |
| Widowed | 19 (5.2) | 0 (0.0) |
| Missing | 84 (22.9) | 1 (2.6) |
| Educational status | ||
| None | 189 (51.5) | 1 (2.6) |
| Primary school | 101 (27.5) | 11 (28.9) |
| Secondary school | 68 (18.5) | 22 (57.9) |
| University | 4(1.1) | 3 (7.9) |
| Missing | 5 (1.4) | 1 (2.6) |
| Number of children | ||
| 0 | 67 (18.3) | 17 (44.7) |
| 1 | 82 (22.3) | 4 (10.5) |
| 2 or more | 145 (39.5) | 2 (5.3) |
| Missing | 73 (19.9) | 15 (39.5) |
| Characteristics related to sex work | ||
| Country of origin | ||
| Ivory Coast | 268 (73.0) | 36 (94.7) |
| Other | 66 (18.0) | 1 (2.6) |
| Missing | 33 (9.0) | 1 (2.6) |
| Having a nonpaying partner | ||
| Yes | 132 (36.0) | 22 (57.9) |
| No | 186 (50.7) | 16 (42.1) |
| Missing | 49 (13.4) | 0 (0.0) |
| Duration in sex work, yrs | ||
| <2 | 78 (21.3) | 3 (7.9) |
| 2 | 70 (19.1) | 3 (7.9) |
| 3 | 86 (23.4) | 6 (15.8) |
| ≥4 | 97 (26.4) | 22 (57.9) |
| Missing | 36 (9.8) | 4 (10.5) |
| Number of clients during last working day | ||
| ≤1 | 103 (28.1) | 17 (44.7) |
| 2 | 109 (29.7) | 10 (26.3) |
| ≥3 | 91 (24.8) | 7 (18.4) |
| Missing | 64 (17.4) | 4 (10.5) |
| Condom use with clients | ||
| Always | 102 (27.8) | 22 (57.9) |
| Many times | 57 (15.5) | 11 (28.9) |
| Sometimes | 89 (24.3) | 1 (2.6) |
| Never | 78 (21.3) | 0 (0.0) |
| Missing | 41 (11.2) | 4 (10.5) |
| Money received from last client, in CFA* | ||
| ≤1000 CFA | 171 (46.6) | 3 (7.9) |
| 1001–3000 CFA | 80 (21.8) | 6 (15.8) |
| >3000 CFA | 54 (14.7) | 24 (63.2) |
| Missing | 62 (16.9) | 5 (13.2) |
| Usual amount of money received from clients, in CFA | ||
| ≤1000 CFA | 218 (59.4) | 7 (18.4) |
| 1001–3000 CFA | 58 (15.8) | 9 (23.7) |
| >3000 CFA | 30 (8.2) | 15 (39.5) |
| Missing | 61 (16.6) | 7 (18.4) |
| Current drug use | ||
| Yes | 0 (0.0) | 5 (13.2) |
| No | 306 (83.4) | 30 (78.9) |
| Missing | 61 (16.6) | 3 (7.9) |
| Biomedical characteristics | ||
| Weight, kg | ||
| <50 | 129 (35.1) | 6 (15.8) |
| 50–54.9 | 78 (21.3) | 5 (13.2) |
| 55–55.9 | 62 (16.9) | 10 (26.3) |
| ≥60 | 74 (20.2) | 10 (26.3) |
| Missing | 24 (6.5) | 7 (18.4) |
| Hemoglobin, g/dL | ||
| <8.5 | 79 (21.5) | 4 (10.5) |
| 8.5–10 | 93 (25.3) | 6 (15.8) |
| 10–11.5 | 113 (30.8) | 2 (5.3) |
| ≥11.5 | 52 (14.2) | 22 (57.9) |
| Missing | 30 (8.2) | 4 (10.5) |
| CD4 cells/mm3 | ||
| <200 | 180 (49.0) | 16 (42.1) |
| 200–49 | 152 (41.4) | 16 (42.1) |
| ≥350 | 29 (7.9) | 5 (13.2) |
| Missing | 6 (1.6) | 1 (2.6) |
| CDC† clinical stage | ||
| A | 27 (7.4) | 2 (5.3) |
| B | 249 (67.8) | 23 (60.5) |
| C | 25 (6.8) | 8 (21.1) |
| Missing | 66 (18.0) | 5 (13.2) |
1 US $= 500 F CFA (at the time of the study).
CDC, centers for disease control and prevention; CFA, Communauté Financière Africaine; FSW, female sex worker; MSW, male sex worker.
The median duration in sex work was 3 years (IQR, 2–4) for the female and 5 years (IQR, 3–10) for the male sex workers. Female sex workers reported a median of 2 clients a day, and male sex workers reported a median of 1.5 clients a day. The usual amount of money received from clients was US $2 [median (IQR): 2–3] for female and US $6 [median (IQR): 4–20] for male sex workers. At ART initiation, the median weight of the female sex workers was 53 kg [median (IQR): 47–60] and 58 kg for the male sex workers [median (IQR): 53–63]. No viral loads were performed, but CD4 counts were recorded for most of the patients. The median CD4 count was 202 cells per cubic millimeter (IQR, 117–279) for female and 218 cells per cubic millimeter (IQR, 125–295) for male sex workers.
Time Sequences
All patients were ART naive at the time of ART initiation. A majority of sex workers had their HIV test performed on the same day they registered for the first time at the clinic (79%). The time between HIV diagnosis and ART at the sex worker clinic was less than 1 month for 92% of the sex workers. The median time between HIV diagnosis and ART treatment was 18 days (IQR, 10–85) for female and 120 days (IQR, 9–406) for male sex workers.
Survival Probabilities
The 367 female sex workers accumulated a total of 677.0 person-years of follow-up, whereas the 38 male sex workers accumulated 68.6 person-years of follow-up.
The retention probability was 75% at 6 months, 68% at 12 months, 55% at 24 months, and 47% at 36 months. Table 2 shows the cumulative incidence of death, LTFU, stopping ART, and attrition as composite end point among sex workers starting ART during 2005–2008. Figure 1 shows the cumulative incidence curves for each competing outcome stratified by calendar year of ART initiation.
TABLE 2.
Cumulative Incidence of Death, LTFU, and Stopping ART Among Sex Workers Starting ART During 2005–2008
| Time After ART Initiation, yrs | 2005, % | 2006, % | 2007, % | 2008, % | |
|---|---|---|---|---|---|
| Dead | |||||
| 0.5 | 3 | 2 | 3 | 1 | |
| 1 | 3 | 2 | 3 | 2 | |
| 2 | 6 | 5 | 7 | 4 | |
| 3 | 6 | 8 | 9 | — | |
| 4 | 9 | 10 | — | — | |
| Stopped | |||||
| 0.5 | 0 | 0 | 0 | 0 | |
| 1 | 0 | 0 | 0 | 0 | |
| 2 | 0 | 0 | 0 | 0 | |
| 3 | 0 | 0 | 2 | — | |
| 4 | 0 | 1 | — | — | |
| LTFU | |||||
| 0.5 | 3 | 18 | 26 | 24 | |
| 1 | 12 | 23 | 29 | 33 | |
| 2 | 21 | 28 | 35 | 49 | |
| 3 | 24 | 31 | 45 | — | |
| 4 | 30 | 38 | — | — | |
| *Attrition | |||||
| 0.5 | 6 | 20 | 28 | 25 | |
| 1 | 15 | 25 | 32 | 35 | |
| 2 | 27 | 33 | 42 | 53 | |
| 3 | 30 | 39 | 55 | — | |
| 4 | 39 | 49 | — | — |
Attrition is the combined outcome of death, LTFU, and stopping ART.
FIGURE 1.
Cumulative incidence of death, LTFU, and stopping ART among sex workers starting ART in Côte d’Ivoire during 2005–2008.
Risk Factors for LTFU
Factors significantly associated with LTFU in bivariate analysis were lower schooling level, later year of starting ART, and not receiving initial adherence counseling at ART initiation (Table 3). Taking all variables into consideration in the multivariate model, only later year of starting ART, and not receiving adherence counseling at ART initiation remained significantly associated with LTFU.
TABLE 3.
Baseline Characteristics as Predictors for LTFU in Bivariate and Multivariate Analyses (Cox Proportional Hazards Model)
| Crude |
Adjusted* |
|||||||
|---|---|---|---|---|---|---|---|---|
| N | Rate/100 PY | HR | 95% CI | P | AHR | 95% CI | P | |
| Sociodemographics | ||||||||
| Sex | ||||||||
| Male | 40 | 12.7 | 1 | 1 | ||||
| Female | 384 | 21.7 | 1.68 | 0.89 to 3.19 | 0.111 | 1.13 | 0.53 to 2.40 | 0.748 |
| HIV type | ||||||||
| HIV-1 | 388 | 20.3 | 1 | 1 | ||||
| HIV-2 or dual HIV-1 and 2 | 33 | 26.2 | 1.32 | 0.80 to 2.17 | 0.280 | 1.37 | 0.78 to 2.40 | 0.266 |
| Age group, yrs | ||||||||
| <30 | 151 | 26.2 | 1 | 1 | ||||
| ≥30 | 273 | 18.0 | 0.74 | 0.54 to 1.02 | 0.065 | 0.77 | 0.51 to 1.15 | 0.198 |
| Marital status | ||||||||
| Married/living with partner | 33 | 16.5 | 1 | 1 | ||||
| Single/divorced/widowed | 306 | 21.5 | 1.23 | 0.69 to 2.18 | 0.484 | 1.34 | 0.71 to 2.51 | 0.361 |
| Schooling level | ||||||||
| Primary or less | 314 | 23.2 | 1 | 1 | ||||
| Secondary or higher | 104 | 14.4 | 0.65 | 0.44 to 0.95 | 0.027 | 0.72 | 0.47 to 1.11 | 0.140 |
| Number of children | ||||||||
| 0–1 | 176 | 22.0 | 1 | 1 | ||||
| 2 or more | 154 | 19.6 | 0.97 | 0.70 to 1.34 | 0.837 | 0.99 | 0.66 to 1.48 | 0.964 |
| Operations characteristics | ||||||||
| Study site | ||||||||
| Abidjan | 232 | 20.0 | 1 | 1 | ||||
| San Pedro | 192 | 21.6 | 1.16 | 0.86 to 1.57 | 0.329 | 1.00 | 0.64 to 1.58 | 0.995 |
| Adherence counseling pre-ART | ||||||||
| Yes | 299 | 15.3 | 1 | 0.43 to 0.90 | 0.013 | 1 | ||
| No | 125 | 23.1 | 1.59 | 1.11 to 2.30 | 0.012 | 2.35 | 1.49 to 3.70 | <0.001 |
| Year of starting ART at the sex worker clinic | ||||||||
| 2005–2006 | 166 | 13.1 | 1 | 1 | ||||
| 2007 | 149 | 24.3 | 1.63 | 1.12 to 2.37 | 0.010 | 1.85 | 1.23 to 2.77 | 0.003 |
| 2008 | 109 | 39.6 | 2.21 | 1.48 to 3.29 | 0.000 | 3.06 | 1.93 to 4.87 | 0.000 |
| Prescription of cotrimoxazole at initiation care | ||||||||
| No | 282 | 24.5 | 1 | 1 | ||||
| Yes | 142 | 15.6 | 0.73 | 0.52 to 1.01 | 0.058 | 0.76 | 0.53 to 1.09 | 0.140 |
| Type of initial ARV regimen | ||||||||
| D4T combination | 359 | 20.0 | 1 | 1 | ||||
| AZT combination | 63 | 26.0 | 1.19 | 0.78 to 1.80 | 0.416 | 1.17 | 0.74 to 1.87 | 0.491 |
| Sex work related | ||||||||
| Time since starting sex work | ||||||||
| Up to 3 yrs | 162 | 22.6 | 1 | 1 | ||||
| 3 yrs or more | 216 | 19.7 | 0.93 | 0.67 to 1.28 | 0.643 | 1.19 | 0.84 to 1.69 | 0.319 |
| Number of clients last working day | ||||||||
| 0–2 | 248 | 19.2 | 1 | 1 | ||||
| ≥3 | 102 | 25.0 | 1.24 | 0.86 to 1.79 | 0.249 | 1.33 | 0.88 to 2.02 | 0.171 |
| Usual client price† | ||||||||
| ≤1000 CFA | 234 | 22.7 | 1 | 1 | ||||
| >1000 CFA | 116 | 17.2 | 0.75 | 0.53 to 1.06 | 0.103 | 0.81 | 0.54 to 1.22 | 0.315 |
| Reported condom use with clients | ||||||||
| Not always | 133 | 20.4 | 1 | 1 | ||||
| Always | 240 | 20.4 | 1.02 | 0.74 to 1.42 | 0.888 | 1.10 | 0.73 to 1.65 | 0.654 |
| Biomedical | ||||||||
| CD4 cells/mm3 | ||||||||
| <200 | 190 | 22.1 | 1 | 1 | ||||
| 200–349 | 164 | 20.2 | 0.90 | 0.65 to 1.24 | 0.506 | 0.90 | 0.64 to 1.26 | 0.528 |
| ≥350 | 25 | 14.9 | 0.65 | 0.31 to 1.39 | 0.269 | 0.61 | 0.28 to 1.34 | 0.217 |
| Hemoglobin, g/dL | ||||||||
| <10 | 171 | 24.7 | 1 | 1 | ||||
| ≥10 | 180 | 17.7 | 0.77 | 0.56 to 1.07 | 0.115 | 0.80 | 0.57 to 1.14 | 0.216 |
| CDC disease stage | ||||||||
| A | 32 | 18.3 | 1 | 1 | ||||
| B | 283 | 21.7 | 1.21 | 0.66 to 2.23 | 0.532 | 1.19 | 0.63 to 2.24 | 0.596 |
| C | 35 | 15.6 | 0.88 | 0.38 to 2.02 | 0.763 | 0.83 | 0.35 to 2.01 | 0.687 |
Adjusted: all variables were included in the final model.
1 US $= 500 F CFA (at the time of the study).
ART, antiretroviral therapy; AZT, zidovudine; CI, confidence intervals; D4T, stavudine; CFA, Communauté Financière Africaine; HR, hazard ratio.
DISCUSSION
This is one of the first studies assessing retention of sex workers on ART in a routine setting in a low-income country. The retention of sex workers on ART in Ivory Coast was 55% after 2 years.
Two other studies, 1 from Burkina Faso and 1 from Benin, showed the effectiveness of ART provided for female sex workers in a research setting. The study in Burkina Faso measured the long-term virologic, immunological, and clinical efficacy of highly active ART in a cohort of female sex workers in Bobo-Dioulasso.23 Only 47 female sex workers were followed up, counting for a total of 111 person-years of follow-up. During the observation time, 2 women dropped out and 4 women died. However, as Huet et al pointed out that their results were obtained within the framework of a research study, which used frequent visits combined with intensive counseling and case management services to maintain patients in care. Therefore, these results may not be representative for sex workers on ART in other programs.8 A similar study in Benin showed the response to ART among 53 female sex workers to be lower than the response in the general population because of poor adherence.24 None of the 2 studies focused on retention rates and predictors for LTFU.
Compared with general population cohorts in Ivory Coast, our retention probabilities are slightly lower. In a recent study, using the same methodology among a representative sample of all ART facilities in Ivory Coast, retention was 79% after 6, 74% after 12, 65% after 24, and 56% after 36 months.25 Reasons for the lower retention of among sex workers, as compared with the general population, may include their higher mobility but should be further investigated.24
Compared with patients starting ART in high-income countries, patients in low- and middle-income countries tend to start ART with lower CD4 counts and have higher early mortality.26 The sex workers in our studies were late presenters, with more than half of them on ART less than 1 month after their HIV diagnosis (median time between diagnosis and ART 21 days, IQR 10–112). Because their CD4 at initiation of ART was very low, they were at high risk of early mortality. It is likely that our study underestimates true mortality in this cohort because deaths were not always recorded and a proportion of LTFU may be because of unrecorded deaths. Other studies have shown that mortality is not the major reason for patient attrition in large ART programs in developing countries.14,15 The greater threat to the success of the ART program may be the high levels of LTFU, insofar as this outcome reflects patients who have truly left care.27
Increased global access and use of ART has been postulated to undermine HIV prevention efforts by changing individual risk-taking behavior.28 If behavior of sex workers was similar, the effect of a lower viral load through ART may be compromised by an increase in risky behavior. However, studies in Kenya and Benin found that ART initiation was not associated with an increase in unprotected sexual contacts by female sex workers.24,29 We do not have longitudinal data on condom use after initiating ART in our cohort.
As the ART program expanded during 2004 through 2007, retention of newly enrolled ART patients decreased, a temporal trend that has been observed in other programs in resource-constrained settings.27,30 This declining trend of the retention rates with later calendar year of ART initiation is worrisome, and reasons for this trend should be explored. The lower retention rate may be related to the increasing workload in the clinics after scaling up of ART treatment. Scaling up of ART treatment may also be related to an increase in observed LTFU because of undocumented transfer of patients to other health facilities. Finally, between 2007 and 2009, there was a perceived pressure from international donors to increase the numbers of patients on ART. This may have resulted in a less thorough screening for ART readiness before starting ART and a selection of patients who are less motivated to remain on treatment. A study in a primary health care setting in South Africa has demonstrated the importance of careful selection and preparation of patients for their retention in ART care.31 Further qualitative research is needed to explain this declining retention with later calendar years.
In addition to earlier calendar year of ART initiation, the counseling session at the ART initiation seems to be very important for retention of sex workers in ART care. This is in line with findings of an ecologic study of 349 clinics in 10 countries in sub-Saharan Africa. Availability of counseling services was significantly associated with higher retention rates in this study.32
This study also has some limitations. Because this was a retrospective chart review, other important structural predictors of retention, such as mode of transport, income, and social support systems could not be assessed. In addition, the retrospective design resulted in many challenges of incomplete data. They highlight the importance of strengthening data collection systems to better respond and assess retention to care and treatment.
Another limitation is the not randomized selection of study sites, which could have introduced some selection bias. However, at the time of the study, the 2 sites were the only sites providing ART care to sex workers since more than 12 months in Ivory Coast. It was estimated that these 2 sites adequately reflect the spectrum of ART care that has been provided to sex workers in Cote d’Ivoire.
CONCLUSIONS
The question/debate is not any more whether to put sex workers on ART but how to retain them in ART care to reduce HIV transmission. To improve the retention of sex workers on ART in Ivory Coast, there is a need for more in-depth investigation of the role of pre-ART counseling and the increasing rates of LTFU with each calendar year.
Acknowledgments
Supported by the President’s Emergency Plan for AIDS Relief (PEPFAR) through the US Centers for Disease Control and Prevention (CDC) and the Belgian Directorate-General for Development Cooperation (DGD) through a Framework Agreement with the Institute of Tropical Medicine.
The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the US Centers for Disease Control and Prevention. Use of trade names is for identification only and does not imply endorsement by the US Centers for Disease Control and Prevention or the US Department of Health and Human Services.
Footnotes
The authors have no conflicts of interest to disclose.
REFERENCES
- 1.Egger M, May M, Chene G, et al. Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet. 2002;360:119–129. [DOI] [PubMed] [Google Scholar]
- 2.WHO. Global update on HIV treatment 2013: results, impact and Opportunities. Version 2. 2013. Available at: http://apps.who.int/iris/bitstream/10665/85326/1/9789241505734_eng.pdf Accessed January23, 2014.
- 3.Mahy M, Stover J, Stanecki K, et al. Estimating the impact of antiretroviral therapy: regional and global estimates of life-years gained among adults. Sex Transm Infect. 2010;86(suppl 2):ii67–ii71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Delva W, Eaton JW, Meng F, et al. HIV treatment as prevention: optimising the impact of expanded HIV treatment programmes. PLoS Med. 2012;9:e1001258. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Cohen MS, Chen YQ, McCauley M, et al. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med. 2011;365:493–505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Eaton JW, Johnson LF, Salomon JA, et al. HIV treatment as prevention: systematic comparison of mathematical models of the potential impact of antiretroviral therapy on HIV incidence in South Africa PLoS Med. 2012;9:e1001245. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Tanser F, Bärnighausen T, Grapsa E, et al. High coverage of ART associated with decline in risk of HIV acquisition in rural KwaZulu-Natal, South Africa. Science. 2013;339:866–971. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Montague BT, Vuylsteke B, Buve A. Sustainability of programs to reach high risk and marginalized populations living with HIV in resource limited settings: implications for HIV treatment and prevention. BMC Public Health. 2011;11:701. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Richter M, Chersich MF, Vearey J, et al. Migration status, work conditions and health utilization of female sex workers in three South African cities. JImmigr Minor Health. 2012;16:7–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Institut National de la Statistique et ICF International. Enquête Démographique et de Santé et à Indicateurs Multiples de Côte d’Ivoire 2011–2012. Calverton, MD: INS et ICF International; 2012. [Google Scholar]
- 11.Toure S, Kouadio B, Seyler C, et al. Rapid scaling-up of antiretroviral therapy in 10,000 adults in Cote d’Ivoire: 2-year outcomes and determinants. AIDS. 2008;22:873–882. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Vuylsteke B, Semde G, Sika L, et al. High prevalence of HIV and sexually transmitted infections among male sex workers in Abidjan, Cote d’Ivoire: need for services tailored to their needs. Sex Transm Infect. 2012;88:288–293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Vuylsteke B, Semde G, Sika L, et al. HIV and STI prevalence among female sex workers in Cote d’Ivoire: why targeted prevention programs should be continued and strengthened. PLoS One. 2012;7: e32627. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Wandeler G, Keiser O, Pfeiffer K, et al. Outcomes of antiretroviral treatment programs in rural Southern Africa. J Acquir Immune Defic Syndr. 2012;59:e9–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Brinkhof MW, Dabis F, Myer L, et al. Early loss of HIV-infected patients on potent antiretroviral therapy programmes in lower-income countries. Bull World Health Organ. 2008;86:559–567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Rubin DB. Multiple Imputation for Nonresponse in Surveys. New York, NY: J. Wiley & Sons; 1987. [Google Scholar]
- 17.Royston P Multiple imputation of missing values. StataJ. 2004;4:227–241. [Google Scholar]
- 18.Royston P Multiple imputation of missing values: update. Stata J. 2005; 5:188–201. [Google Scholar]
- 19.Royston P Multiple imputation of missing values: update of ice. Stata J. 2005;5:527–536. [Google Scholar]
- 20.White IR, Royston P. Imputing missing covariate values for the Cox model. Stat Med. 2009;28:1982–1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Royston P, Carlin JB, White IR. Multiple imputation of missing values: new features for mim. Stata J. 2009;9:252–264. [Google Scholar]
- 22.Grambsch PM, Therneau TM. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika. 1994;81:515–526. [Google Scholar]
- 23.Huet C, Ouedraogo A, Konate I, et al. Long-term virological, immunological and mortality outcomes in a cohort of HIV-infected female sex workers treated with highly active antiretroviral therapy in Africa. BMC Public Health. 2011;11:700. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Diabate S, Chamberland A, Zannou DM, et al. Sexual behaviour after antiretroviral therapy initiation in female sex workers and HIV-positive patients from the general population, Cotonou, Benin. AIDS Care. 2013; 25:1426–1432. [DOI] [PubMed] [Google Scholar]
- 25.Auld AF, Ekra KA, Shiraishi RW, et al. Temporal trends in treatment outcomes for HIV-1 and HIV-2-infected adults enrolled in côte d’Ivoire’s national antiretroviral therapy program. PLoS One. 2014;9:e98183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Braitstein P, Brinkhof MW, Dabis F, et al. Mortality of HIV-1-infected patients in the first year of antiretroviral therapy: comparison between low-income and high-income countries. Lancet. 2006;367:817–824. [DOI] [PubMed] [Google Scholar]
- 27.Cornell M, Grimsrud A, Fairall L, et al. Temporal changes in programme outcomes among adult patients initiating antiretroviral therapy across South Africa, 2002–2007. AIDS. 2010;24:2263–2270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Doyle JS, Degenhardt L, Pedrana AE, et al. Effects of HIV antiretroviral therapy on sexual and Injecting risk-taking behavior: a systematic review and meta-analysis. Clin Infect Dis. 2014;59:1483–1484. [DOI] [PubMed] [Google Scholar]
- 29.McClelland RS, Graham SM, Richardson BA, et al. Treatment with antiretroviral therapy is not associated with increased sexual risk behavior in Kenyan female sex workers. AIDS. 2010;24:891–897. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Auld AF, Mbofana F, Shiraishi RW, et al. Four-year treatment outcomes of adult patients enrolled in Mozambique’s rapidly expanding antiretroviral therapy program. PLoS One. 2011;6:e18453. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Coetzee D, Boulle A, Hildebrand K, et al. Promoting adherence to antiretroviral therapy: the experience from a primary care setting in Khayelitsha, South Africa. AIDS. 2004;18(suppl 3):S27–S31. [DOI] [PubMed] [Google Scholar]
- 32.Lamb MR, El-Sadr WM, Geng E, et al. Association of adherence support and outreach services with total attrition, loss to follow-up, and death among ART patients in sub-Saharan Africa. PLoS One 2012;7:e38443. [DOI] [PMC free article] [PubMed] [Google Scholar]