Implementing family-focused HIV care and treatment: the first 2 years’ experience of the mother-to-child transmission -plus program in Abidjan, Côte d’Ivoire

Authors


Corresponding Author Besigin Tonwe-Gold, Programme ACONDA-VS, 23 BP 2485 Abidjan 23, Côte d’Ivoire. E-mail: tonwe@aviso.ci

Summary

Objectives  To describe a family-focused approach to HIV care and treatment and report on the first 2 years experience of implementing the mother-to-child transmission (MTCT)-plus program in Abidjan, Côte d’Ivoire.

Program  The MTCT-plus initiative aims to enrol HIV-infected pregnant and postpartum women in comprehensive HIV care and treatment for themselves and their families.

Main outcomes  Between August 2003 and August 2005, 605 HIV-infected pregnant or postpartum women and 582 HIV-exposed infants enrolled. Of their 568 male partners reported alive, 52% were aware of their wife’s HIV status and 30% were tested for HIV; 53% of these tested partners were found to be HIV-infected and 78% enrolled into the program. Overall only 10% of the women enrolled together with their infected partner. On the other hand, the program involved half of the seronegative men who came for voluntary counselling and testing (VCT) in the care of their families. Of 1624 children <15 years reported alive by their mothers (excluding the last newborn infants of the most recent pregnancy systematically screened for HIV), only 10.8% were brought in for HIV testing, of whom 12.3% were found to be HIV-infected.

Lessons learned and challenges  The family-focused model of HIV care pays attention to the needs of families and household members. The program was successful in enrolling HIV women, their partners and infants in continuous follow-up. However engaging partners and family members of newly enrolled women into care involves numerous challenges such as disclosure of HIV status by women to their partners and family members. Further efforts are required to understand barriers for families accessing HIV services as strategies to improve partner involvement and provide access to care for other children in the households are needed in this West African urban setting.

Abstract

Objectifs:  Décrire une approche de soins et traitement axée sur la famille et rapporter l’expérience des deux premières années de l’implémentation du programme sur la transmission mère-enfant (TME-Plus) à Abidjan en Côte d’Ivoire.

Programme:  L’initiative TME-Plus vise à recruter des femmes enceintes et en post-partum infectées par le VIH dans des soins et traitement complets pour elles-mêmes et leurs familles.

Principaux résultats:  Entre août 2003 et août 2005, 605 femmes enceintes ou en post-partum infectées par le VIH et 582 nourrissons exposés au VIH ont été recrutés. 568 partenaires masculins signalés comme étant en vie, 52% d’entre eux étant au courant de leur du statut VIH de leur épouse et 30% testés pour le VIH. 53% de ces partenaires testés se sont révélés infectés par le VIH et 78% ont été recrutés dans le programme. Au total, seules 10% des femmes ont été inscrites avec leur partenaire infecté. D’autre part, le programme a aussi recruter la moitié des hommes séronégatifs venus pour le CDV dans le cadre des soins pour leur famille. Sur 1624 enfants de moins de 15 ans rapportés vivants par leur mère (à l’exclusion du dernier nouveau-né de la dernière grossesse systématiquement dépistée pour le VIH), seuls 10,8% ont été amenés pour le dépistage du VIH, parmi lesquels 12,3% ont été trouvés infectés par le VIH.

Leçons tirées et défis:  Le modèle de soins VIH axé sur la familial prête attention aux besoins des membres de la famille et du ménage. Le programme a réussi à recruter les femmes VIH, leurs partenaires et les nourrissons en suivi continu. Toutefois engager les partenaires et les membres de la famille des femmes nouvellement recrutées pour les soins implique de nombreux défis, tels que la divulgation du statut VIH par les femmes à leurs partenaires et aux membres de la famille. Des efforts supplémentaires sont nécessaires pour comprendre les obstacles pour les familles à l’accès aux services VIH telles que des stratégies visant à améliorer l’implication des partenaires et procurer l’accès aux soins à d’autres enfants dans les ménages qui sont nécessaires dans cette région urbaine d’Afrique de l’Ouest.

Abstract

Objetivos:  Describir un enfoque basado en la familia de cuidados y tratamiento para VIH, y reportar sobre la experiencia de los primeros dos años de implementación del programa MTCT-Plus en Abidjan, Costa de Marfil.

Programa:  La iniciativa del MTCT-Plus buscaba incluir mujeres VIH positivas, tanto embarazadas como en el postparto, dentro de un programa integral de cuidados y tratamiento del VIH, tanto para ellas como para sus familias.

Principales resultados:  Entre Agosto 2003 y Agosto 2005, 605 mujeres infectadas con VIH, embarazadas o en postparto, y 582 recién nacidos expuestos a VIH fueron incluidos dentro del programa. De las 568 parejas masculinas que habían sido reportadas como vivas, un 52% eran conscientes del seroestatus de su esposa y un 30% se hicieron la prueba del VIH. Un 53% de estas parejas fueron diagnosticadas como VIH-positivas y un 78% incluidos en el programa. En general, solo un 10% de las mujeres participaron junto con sus parejas infectadas. Por otro lado, el programa incluyó a la mitad de los hombres seronegativos que vinieron a recibir aconsejamiento y prueba. De los 1,624 niños <15 años reportados como vivos por sus madres (excluyendo los recién nacidos del embarazo más reciente quienes eran sistemáticamente testados para VIH), solo un 10.8% fueron traídos para ser testados para VIH, y de estos se encontró que un 12.3% estaban infectados con VIH.

Lecciones aprendidas y retos:  El modelo de cuidados para VIH enfocados en la familia se centra en las necesidades de las familias y los miembros del hogar. El programa tuvo éxito incluyendo dentro de un seguimiento continuo a mujeres infectadas con VIH, a sus parejas y recién nacidos. Sin embargo, el involucrar a las parejas y miembros de la familia de mujeres recién incluidas en el programa trae consigo muchos retos, entre ellos el que las mujeres revelen su seropositividad a sus parejas y otros miembros de la familia. Se requieren más esfuerzos para entender las barreras existentes en familias que acceden a los servicios de cuidado del VIH, con el fin de mejorar la participación tanto de las parejas como de los niños miembros de hogares de esta localidad urbana de África del oeste.

Introduction

Effective models of care are being sought to provide successful large-scale strategies for safe, efficient and appropriate HIV care and treatment in poor countries. Owing to a new global dynamic with substantial increases in international funding and lower antiretroviral (ARV) drug prices, antiretroviral treatment (ART) has recently expanded throughout sub-Saharan Africa. Most knowledge of ART delivery comes from industrialized countries (Egger et al. 2002; May et al. 2006) while data from treatment programmes in sub-Saharan Africa are slowly being gathered (Braitstein et al. 2006; Calmy et al. 2004; Ferradini et al. 2006; Stringer et al. 2006). In particular, there are few data examining the application and benefits of a family-focused element to HIV treatment in developing countries.

Various models of care have been adopted by programmes throughout sub-Saharan Africa to scale up the delivery of ART. In most settings, HIV care programmes are centred on the specific needs of the individual adult and child engaged in treatment (Mermin 2005; Arrivéet al. 2008). The family-focused approach is a distinct model of HIV care which was pioneered by the mother-to-child transmission (MTCT)-plus initiative at the International Centre for AIDS Care and Treatment Programs, Columbia University (2007a). The MTCT-plus model of care was established to address the long-term care and treatment needs of HIV-infected women for the prevention of mother-to-child transmission (PMTCT). In this model, the pregnant or postpartum HIV woman is the pivotal person who serves as the guide, steering her family and household members in accessing HIV care and treatment services. The approach is distinguished by: attending to the needs of adults and children; and providing comprehensive prevention and care services for all family members. Taking the position that the environment of an individual (and therefore the family) has a direct impact on the individual’s ability to promote his or her own health, family-focused care offers the opportunity to bring women, their partners and children into HIV care together while encouraging providers to consider the needs of all family members (Abram et al. 2007).

This report describes the challenges and successes of implementing a family-focused approach to HIV care and treatment at two MTCT-Plus sites in Abidjan, Côte d’Ivoire.

The MTCT-plus approach of family care

Setting

In 2002, the MTCT-plus initiative was one of the first internationally-led programmes to support HIV care and treatment in resource-limited settings and the adoption of the family model for delivery of such services. The initiative funded 14 demonstration programmes in multiple urban, peri-urban and rural sites in eight sub-Saharan African countries and in Thailand (Columbia University 2007a). In the crowded and poor districts of Yopougon and Abobo (Abidjan, Cote d’Ivoire), two MTCT-plus sites were established in community-based antenatal clinics.

Patients enrolled in this MTCT-plus programme had access to regular clinical and laboratory assessments, nutritional support, HIV diagnostic testing, family planning services, prophylaxis and management of opportunistic infections and ART initiated according to WHO and local guidelines. The programmes also supported community outreach and education. Standardized approaches facilitated procurement of drugs and supplies, training, and data collection (Columbia University 2007b).

Staffing and training

A multidisciplinary team of health workers composed of physicians, nurses, midwives, counsellors and outreach workers, pharmacy staff, data entry and monitoring personnel, laboratory technicians as well as peers from persons living with HIV/AIDS (PLWHA) was formed at each clinic. Each team aimed to care for 750 adults and children living with HIV (Columbia University 2007b). Training was performed using an MTCT-plus designed curriculum focusing on the team as a whole. This was instrumental in cementing the relationship between the team members and establishing a culture devoid of rigid hierarchy. The training scheme used a competency-based approach focused on building individual and team skills to implement the programme rather than providing only information updates and general HIV training (Columbia University 2007b).

Enrolment into the programme

Women were recruited from two sources in Abidjan: pregnant HIV-infected women attending routine antenatal visits at six urban clinics and postpartum women within 18 months of delivery who were previously enrolled in the PMTCT research protocol ANRS 1201/1202 Ditrame Plus (Dabis et al. 2005; Ekouevi et al. 2004). Women were invited to participate in the programme if they lived in one of the two urban districts where the MTCT-plus centres were established.

Each woman was systematically asked to list her family and household members and was encouraged to bring her partner, children and other household members for HIV testing. Identifying family members allowed providers to evaluate family ties, strengths and needs, concerns and in particular to identify key decision makers within the family. Counselling to encourage partner notification was given to all women who had not yet informed their partners of their HIV status. When needed, sessions were scheduled with a psychologist to develop and support individual disclosure plans. Male partners who tested negative for HIV were not enrolled into the programme but were encouraged to provide support to their families by visiting the clinic with their families, participating in adherence support sessions for their wives and children, and attending a peer support group.

All infants born to pregnant women, and those most recently born to postpartum women, were enrolled in the programme and followed closely until their final HIV status was confirmed.

Programme components

Patients enrolled in the programme received a full array of clinical, supportive and laboratory services (Rouet et al. 2005). As shown in Figure 1 this included referrals to psychosocial and nutritional support, as well as family planning and tuberculosis services. The follow-up schedule depended on the stage of HIV disease as determined by clinical assessment, CD4+ cell count and whether the patient received ART. This consisted of: cotrimoxazole prophylaxis in patients with CD4+ cell count <500/mm3, weekly follow-up for 8 weeks for those starting ART according to WHO criteria (Tonwe Gold et al. 2007) and then monthly clinic visits; total blood cell count and CD4+ cell count every 6 months; management of clinical events at the clinic following standardized algorithms of investigation and treatment; and systematic follow-up of participants who did not keep scheduled appointments. All care, as well as ART, was provided free-of-charge (including transport and consultations). Patients contributed to other costs such as replacement feeding (if this was their informed choice), hospitalizations, specific medical investigations, drugs other than those for opportunistic infections (OIs) and vitamins. RNA–PCR testing allowed the routine diagnosis of paediatric HIV infection at 4–6 weeks of age (Rouet et al. 2005), with subsequent confirmation according to the breastfeeding pattern after the child reached his or her first birthday. Home visits were provided by peer workers systematically at enrolment for those initiating ART and follow-up visits if needed in case of non-adherence or social difficulties. A form was filled in and transmitted to the clinic’s team and the case was discussed during weekly multidisciplinary meetings. Home visits included therefore adherence counselling, psychosocial support and occasionally nutritional support.

Figure 1.

 Care and interventions offered in the mother-to-child transmission-plus program (Abidjan, Côte d’Ivoire) since 2003.

Antiretroviral first-line regimens for adults followed WHO (2003): a non-nucleoside reverse transcriptase inhibitor (NNRTI)-based regimen with, generally, nevirapine in combination with two nucleoside reverse transcriptase inhibitors, zidovudine (ZDV) and lamivudine (3TC). For children, the first-line regimen was the protease inhibitor (nelfinavir) in association with ZDV and 3TC, except for infants <12 months who received a Ritonavir-boosted Lopinavir regimen. Both adults and children had access to other ARV drugs in case of toxicity and to second-line ART in case of clinical or immunological failure. ART was prescribed only in the context of ongoing adherence support.

Monitoring

A comprehensive medical record system was developed to record individual patient information and to support patient care and follow-up. Programme implementation and outcomes were routinely assessed by monitoring: (i) the number of individuals enrolled; (ii) percent of eligible individuals receiving ART or OI prophylaxis; (iii) programme discontinuation (death, loss-to-follow-up and patient withdrawal); (iv) determination of infant HIV status; and (v) CD4 cell count. The ability of each enrolled woman to bring her family and household members into care was assessed. Any family or household members found to be HIV-infected were enrolled into the programme.

An electronic visit scheduling system was utilized to track and promptly identify individuals who missed visits. Contact information with neighbourhood maps of patients’ residences was updated periodically and used to conduct home visits.

Monitoring of database and statistical analysis

Data were collected every day by physicians, nurses and counsellors and keyed in daily by a trained data clerk at the clinics. All information was entered into Access software provided by Columbia University. The database was sent monthly to data management centres at Columbia University. Each month we received data queries for cleansing. Three forms were used for adults, three for infants and two forms for both adults and infants. For statistical analysis, Group comparisons used Student’s t-test or non-parametric Mann–Whitney U-test for continuous variables, and Chi-square test or Fisher’s exact test for categorical variables. All analyses were performed with stataTM 9.0 (Stata Corporation, College Station, TX, USA).

The Abidjan MTCT-plus program was approved by the Ministry of Health of Côte d’Ivoire. As a service demonstration programme, the MTCT-plus initiative was exempted from formal review by the Columbia University Institutional Review Board.

Results

Enrolment of index women

During the first 2 years of the programme, 605 index women were enrolled. Their median age was 28 years (range 16–46) and the median number of household members was five (range 1–19). A total of 159 women (26%) lived alone or were widowed. The median parity per woman at enrolment (excluding the current pregnancy birth) was two [inter-quartile range (IQR) 1–3]. A third of the women were illiterate (33.6%). At enrolment, 22% of women had a CD4+ count <200 cells/μl, their median CD4+ count was 351/μl (IQR: 219–527), and 24% were WHO stage 3 or 4. Overall, 42.8% of women fulfiled eligibility criteria for ART. As Table 1 shows, 94.9% initiated treatment.

Table 1.   Characteristics of HIV-infected women enrolled in the mother-to-child transmission-plus program in Abidjan, August 2003 to August 2005 (n = 605)
 NumberPercentage
  1. ART, antiretroviral treatment; IQR, interquartile range; WHO, World Heath Organization.

  2. *Between August 2003 and December 2004: stage 4, stage 3 or stage 2 and CD4 count <350/mm3 and CD4 <200/mm3 and since January 2006 the patients with Stage 2 and CD4 <350/mm3 were not eligible for receiving ART.

Median age (range)28 (15–46) 
 <18 years4 0.7
 18–24 years13321.9
 25–29 years23438.7
 ≥30 years23438.7
Time of enrolment
 Prepartum29148.1
 Postpartum31451.9
Median number of children (IQR) per index woman (excluding current pregnancy)2 (1–3) 
 06510.7
 114624.1
 216226.8
 ≥323238.3
Education
 Illiterate20333.6
 Primary school education21435.4
 Secondary school education16427.1
 University education244.0
CD4+ lymphocytes
 Median CD4/mm3 (IQR)351 (219–527) 
 <200/mm313321.9
 200–349/mm316427.1
 350–499/mm312320.4
 ≥500/mm318530.6
WHO staging
 Stages 1 and 246076.0
 Stages 3 and 414524.0
Eligibility for ART
 Eligible (2003 WHO Criteria)*25942.8
 Starting ART at enrolment246/25994.9
Partner identification
 No partner registered213.5
 Died before enrolment 162.6
 Living partner registered at enrolment56893.9

At the time of enrolment 291 index women (48.1%) were pregnant. Women enrolled prepartum were somewhat younger (28.0 vs. 29.0 years; P = 0.018), had lower parity (1.9 vs. 2.6; P < 0.001), a higher CD4+ count (median 429 vs. 377 cells/mm3 P = 0.012) and lower WHO staging: 15.8% women enrolled antepartum classified WHO stage 3 or 4, whereas 31.8% women enrolled postpartum (P ≤ 0.001) did. Women enrolled postpartum were enrolled at a median of 17.1 months (IQR 6.1–25.1) after delivery.

Partner engagement in the programme

Of 605 HIV-infected women, 568 (94%) identified living partners. Three hundred (52.8%) of these women disclosed their HIV status to their partners during a median follow-up of 14.3 months (IQR 9.2–18.7). One hundred and sixty-nine of the 568 identified male partners were tested for HIV, of these 88 (52%) were HIV-infected.

For women who decided not to disclose their HIV status, 63% cited fear of their partner’s reaction (rejection) as the primary reason. Sixty-nine of the 88 (78%) HIV-infected partners were enrolled in the MTCT-plus program. Of the 19 HIV-infected partners not enrolled, four died prior to enrolment, three never returned after learning their HIV status, 10 were followed at other centres and two had appointments for future enrolment.

Figure 2 shows the cascade of effort deployed to engage partners in the process of care. The male partners at enrolment had more advanced HIV disease (median CD4 count of 234/μl) than women (median CD4 count of 351/μl (P < 0.001)). Twenty-one adult household members of the index women were HIV-tested. Twelve were diagnosed with HIV, including three co-spouses, all of whom were enrolled into the programme.

Figure 2.

 Involvement of male partners of the index HIV-infected women in the mother-to-child transmission-plus program, from HIV voluntary counselling and testing to the initiation of antiretroviral treatment. August 2003 to August 2005, Abidjan, Côte d’Ivoire.
Declared: partner identified by index woman, informed: women disclosed to male partner identified, Tested: HIV tested among male partners of index women aware of woman’s status, HIV positive: of male partners, those found HIV positive.

Enrolment of children

Of the 291 women enrolled during pregnancy, 270 delivered live born infants who were followed in care; 267 were tested at a median age of 32 days (IQR: 30–37). At 12 months 222 infants had a final HIV diagnosis; 21 infants were lost to follow-up and 14 died. Nine of 267 children (3.4% 95% CI; 1.5–6.3%) were diagnosed as HIV-infected at week 4 postpartum and another four were diagnosed as infected at 1 year of age (Tonwe Gold et al. 2007).

Of the 314 women who were enrolled postpartum, 312 delivered live born infants who were followed in care: 30 children (9.6%) were diagnosed as HIV-infected.

For all 605 women, 1042 children were reported living (excluding babies from the recent pregnancy described above). Of the 1042 children, 146 were brought to the programme for HIV testing and 18 (12.3%) were HIV-infected. The median age at screening was 6.3 years, IQR (4.8–8.5 years). The 15% of the children living in households of women enrolled postpartum and 11% of those enrolled antepartum were brought forward for testing.

Thus during the first 2-year period of the programme, 57 HIV-infected children were followed in the programme. Four children were enrolled at other care centres. The majority of children (n = 254) enrolled in this MTCT-plus program were originally HIV-exposed infants of indeterminate infection status, 71% of whom were breastfed from birth, thus requiring long-term follow-up to ascertain their final HIV status.

Family approach of HIV care

Table 2 summarizes the different family networks followed in our programme. The family structure varied primarily in accordance to whether the male partner was aware of the HIV status of their female partners and whether they themselves were aware of their own HIV status. Overall, 59% of women had a partner unaware of their own HIV status (not tested in our programme) and 10% were followed with their partner in the programme. Furthermore, 121 (20%) families had two infected household members identified and 11 families (1.8%) had three or more household members with HIV infection.

Table 2.   Family structure of HIV-infected index women enrolled in the mother-to-child transmission-plus program. August 2003 to August 2005 in Abidjan, Côte d’Ivoire*
Category of familyn (%)
  1. *Eleven HIV-infected infants were also enrolled whose mothers had died before enrolment into the programme (index infant).

1: Family with one HIV-infected member identified
 With no partner37 (6.1)
 With partner and children of unknown HIV status355 (58.7)
 With HIV-negative partner and no HIV infected child diagnosed81 (13.4)
2: Family with two HIV infected members identified
 Woman and HIV-infected partner followed up in the programme, no infected child diagnosed59 (9.8)
 Woman and HIV infected partner followed-up in another programme, no child infected diagnosed19 (3.1)
 Woman and partner of unknown status and at least one child HIV-infected36 (6.0)
 Woman and partner with unknown status and at least one other HIV-infected adult patient, no infected child diagnosed7 (1.2)
3: Family with three or more HIV-infected members identified
 Woman and HIV-infected partner and at least one HIV-infected child followed-up in the programme9 (1.5)
 Woman and HIV-infected partner and at least one other HIV-infected adult followed up in the programme and no infected child diagnosed1 (0.2)
 Woman and at least one HIV-infected child and at least one other HIV-infected adult followed up in the programme (outside of the partner)1 (0.2)

Antiretroviral treatment

Fewer than 2% of women and 9% of partners were receiving ART before enrolment. Table 3 demonstrates that during the following 2-year period, 251 (41.5%) women and 41 (65%) partners initiated ART, accounting for 3182 and 482 person-months of follow-up, respectively. An NNRTI-based regimen was initiated in 98% of the index women and 90% of the partners; 43 of the infected children (n = 57 started ART, 39 with a protease inhibitors-based regimen at a median age of 23 months (IQR 11.7–31.5).

Table 3.   ART regimens and status on treatment for adults and children in the mother-to-child transmission plus program. August 2003 to August 2005, Abidjan, Côte d’Ivoire
 Adult index womenAdult menChildren
  1. ART, antiretroviral treatment; NRTI, nucleoside reverse transcriptase inhibitors; NNRTI, non-nucleoside reverse transcriptase inhibitors; PI, protease inhibitors; IQR, interquartile range.

  2. * This includes 246 index women and five other adult women.

Initial ART regimen251*4143
 2 NRTI + 1 NNRTI245374
 2NRTI + 1 PI5439
 3 NRTI100
Follow-up on ART
 Cumulative person – months3182498434
 Median per patient in months (IQR)13 (9–17)13 (8–18)12 (5–15)
 Status as of August 31st 2005
  Alive2393641
  Lost to follow-up6 (2.5)2 (5.5)0 (0.0)
  Deceased6 (2.5)3 (8.3)2 (4.9)

Retention of individuals on ART was very high during the first 2 years of the programme with 2.5% of index women, 5.5% of partners and none of the infected children lost-to-follow-up. Retention was also high for those not eligible for ART with 98% of index women and exposed children and 100% of male partners still in follow-up at the time of the analysis. Table 3 shows that during the 2-year follow-up period 10 of the 605 index women, five of the 69 male partners and 17 of the 582 children (two of whom were HIV-infected) died.

Discussion

The MTCT-plus initiative supported a unique model of care catered to engaging not only individuals, but also family members, in HIV care and treatment. It emphasized the importance of HIV-infected women in assisting their family/household members to access HIV care and treatment. While the programme has had substantial success, it also faced many challenges. Women are often in vulnerable social situations and therefore face multiple barriers in trying to include their families in care; 26% of women enrolled in the programme were not married and not living with their current partner.

In our cohort of 605 women, 53% of the 568 who had a living partner indicated that they had disclosed their HIV status to their male partner. The reasons for non-disclosure have been cited in several studies. Fear of accusations of infidelity, abandonment, discrimination and violence are the primary reasons for non-disclosure. Medley et al. (2004) in her meta-analysis reported that rates and barriers to disclosure among women varied from 16.7% to 86% and that women attending free-standing VCT clinics were more likely to disclose their HIV status to their sexual partners than women who were tested in the context of their antenatal care. Between 3.5% and 14.6% of women reported experiencing a violent reaction from a partner following disclosure. Surprisingly, the disclosure rate in our programme averaged what has been reported elsewhere, despite unrestricted access to care including ART, an unprecedented situation in Abidjan in 2003–2005. Research surrounding the issue of disclosure in the new context of unrestricted access to ART still needs to be done.

Involvement of partners in the programme (i.e. partners agreeing to VCT) was 30% with only 12% of the male partners registered, i.e. actually enrolled due to HIV diagnosis. Bringing men into reproductive health services is known to be very taxing, and finding ways to reach them is complicated (Mbizvo & Bassett 1996). It has been shown that decision-making regarding child bearing, even though there is a dual commitment by both partners, is still seen as the woman’s responsibility (Mbizvo & Bassett 1996). The fact that the MTCT-plus program at our sites was implemented primarily at maternal and child heath facilities (generally believed to be health structures designed for women), may have prevented a larger number of men from choosing to access the services provided. Taking into account the high number of serodiscordant couples, we would still have expected approximately 300 HIV-infected partners (48% of partners of the index women) to be enrolled. But as Table 2 shows, for those male partners who actually did come for testing (even though the majority were seronegative) the programme was successful in continuing to involve them in the care of their families. Having attended clinic visits with their families and participated in adherence sessions when needed, these partners were an invaluable source of support for their wives and children. This suggests that the involvement and support of men is important in delivering comprehensive care to their families.

Regarding the involvement of children in the programme, more than 728 infants and children gained access to VCT services and were screened for HIV infection during this period. This can be interpreted in two ways. We succeeded in providing to 582 neonates systematic enrolment in the programme, close follow-up during the first year of life and availability of early infant diagnostic testing as paediatric diagnosis. But only 146 other children were tested in the programme and out of the children reported alive by enrolled women (more than 1200 have yet to be screened). The challenges in accessing these children are still very important even though intensive counselling was provided by the staff at our sites. One major difficulty believed to hinder access was the possibility that many of the children lived away from the mother’s household with other relatives in distant communities. Data regarding this hypothesis are now being gathered for future assessment, including more formative qualitative data on the barriers mothers perceive for testing their children. Universal screening of infants and children at immunization and sick baby clinics could overcome this barrier through ‘opt-out’ and provider initiated testing on a routine basis and this could lead to improved access and care for HIV-exposed and infected children (Rollins et al. 2006; WHO, UNAIDS, UNICEF 2007).

One of the keys of the success of a family-focused programme relies on the availability of a comprehensive set of services that reflect the needs of every individual within the family and the family unit as a whole. Care in this model requires committed staff with a high provider-to-patient ratio to take into account the needs of the family as a whole. Most importantly, a successful family care programme as initiated by the MTCT-plus initiative (Mermin 2005) requires a shift in how providers view their patients as well as the delivery of health care and prevention services (Myer et al. 2005). Family care extends the responsibilities of the provider to include screening, assessment, and referral of parents or children for physical, emotional, or social problems or health risk behaviours that can adversely affect the health and emotional or social well-being of any of the individual members of a family.

Another objective of this family-oriented model of care was to reduce barriers to adherence to care and treatment. It is difficult to make a direct comparison between various models of care in our settings, but the low rate of cumulative loss to follow-up in the MTCT-plus initiative programmes (2.5% in index women and 5.5% in partners) after a median duration of 13 months on ART is a clear indication that this model of care achieved high rates of retention in care in comparison to other models found in the literature (Calmy et al. 2004; Stringer et al. 2006). But the high retention may be attributed to a number of other factors including: a strong peer support programme, attention to psychosocial issues by a multidisciplinary team and relatively robust funding particularly for staffing. Therefore, operational research studies are needed to compare different approaches to address this issue.

Involving family members in care is an evolving field and it is important to monitor how such programmes influence communications within the family and influence decisions regarding care of the family beyond the first two years of follow-up. Family structure can be complex, and economic constraints are enormous – both factors which can prevent family members from accessing care. ‘Family’ needs to be defined flexibly in settings around the world and programmes oriented towards families should not operate from the position of a rigid definition.

The challenges faced in implementing a family-focused programme in the context of poor urban slums with multiple unfavourable social and political concerns were numerous. Limitations in the extent of success were mostly due to disclosure, as non-disclosure rates to partners are still estimated to be very high even in the context of ART access. We must find ways to test children outside the prenatal context to improve the family approach strategy. But the success of pregnant women as entry points to HIV care and services for the family has now been documented around the world (MTCT-Plus Initiative 2004). Most importantly, remarkable retention in the programme was achieved through a multidisciplinary and comprehensive approach. All models of care should become evidence-based as ART and HIV care are rapidly rolled out in poor settings.

Acknowledgements

We thank the Secretariat of the MTCT-Plus Initiative at Columbia University, especially Miriam Rabkin, Chloe Toesdale and David Hoos who all contributed to the Côte d’Ivoire programme’s success. We also wish to thank Ida Viho and Kouadio Bertin (from the ACONDA-VS Côte d’Ivoire team), the CeDReS laboratory team, the ANRS Ditrame Plus study clinic team, and all patients enrolled in the programme. This study is dedicated to the memory of Pierrette Kassi who was a truly loved and respected mid-wife who passed away on the 20th December 2004 at the site while attending her patients. The Abidjan MTCT-Plus care and treatment programme is supported by the MTCT-Plus Initiative through the International Center for AIDS Care and Treatment Programs (ICAP) at the Columbia University Mailman School of Public Health, New York. The MTCT–Plus Initiative is funded by several private US foundations (http://www.mtctplus.org). The ANRS 1201/1202 Ditrame Plus project on which the MTCT-Plus Abidjan Program was built, was funded by the Agence Nationale de Recherches sur le Sida et les Hépatites Virales (Paris, France), with additional support from the Charity Sidaction (Paris, France). Renaud Becquet was funded by the French charity SIDACTION. Didier Koumavi EKOUEVI received grant from the European Developing Clinical Trial Partnership (senior fellowship).

Ancillary