Mothers starting antiretrovirals in Malawi lost from care more frequently, revealing weaknesses in Option B+ implementation

Although Malawi’s policy of offering lifelong antiretroviral therapy (ART) to women living with HIV who are pregnant or breastfeeding resulted in a sevenfold increase in women receiving ART in 15 months, implementers are concerned by high rates of loss to follow-up, researchers reported at the 20th International AIDS Conference (AIDS 2014) in Melbourne last week.

The lifelong offer of treatment regardless of CD4 cell count, known as 'Option B+', was pioneered by Malawi’s Ministry of Health in order to simply the implementation of ART for prevention of vertical transmission (from mother to child).

Evaluation of retention at Malawi’s largest antenatal clinic found that 23.5% of mothers who initiated ART at the clinic were lost to follow-up after one year. Rates of loss to follow-up were higher in mothers aged 24 and under and in those who initiated ART while breastfeeding or during the first year of the programme (2011), Hannock Tweya of the Lighthouse Trust told delegates. In comparison, 9% of adults who started treatment on general health grounds were lost to follow-up during the same period.

Between September 2011 and September 2013, 2930 women started ART, of whom 84% (2458) were pregnant and 14% (410) breastfeeding. Median age at ART initiation was 26 years (IQR: 22-30).

Of the 20% (577) who missed a scheduled clinic appointment, 272 (47%) collected ART only at initiation and never returned.

Of those women lost to follow-up and subsequently traced, half had stopped antiretroviral therapy and one third had transferred to another clinic, suggesting that while loss to follow-up is higher among women who initiate treatment under the Option B+ guidance, retention may be underestimated.

Among the 40% of women (228) successfully traced, over half of those who stopped taking ART gave travel and lack of transport as reasons for stopping treatment. Not understanding the information that medication was to be taken for life in the initial antiretroviral education session accounted for a further 10% of discontinuations, as did suspected side-effects (10%). Ten per cent were too weak or sick to attend the clinic again. The sizeable proportion of women not able to be traced is likely due to them deliberately giving a false physical address because of fear of stigma and discrimination if their HIV status is inadvertently disclosed, Dr Tweya suggested.

Dr Tweya told delegates that these findings indicate a need for improved post-test counselling in antenatal care and ART clinics; the establishment of targeted programmes and youth-friendly clinics for younger women; and further decentralisation of services working to prevent vertical transmission.

Joep van Oosterhout, presenting on behalf of Dignitas International, the International Union Against Tuberculosis and Lung Disease and the Malawi Ministry of Health, reported on which health system factors support or hinder uptake and retention among women starting Option B+ in Malawi.  

The study evaluated the relationship between health facility characteristics and retention in 141 facilities in the south-east health district. Health facility surveys and health facility cohort reports using routinely collected data were undertaken to determine uptake of testing in antenatal care, ART initiation and six-month retention.

The 141 health facilities comprised four district hospitals, eight community hospitals, 120 health centres and nine private clinics.

Findings from this cross-sectional analysis showed that health facilities had integrated Option B+ into routine service delivery in diverse ways, with variations in location, timing of ART initiation, counselling and referral. While all health facilities had to implement Option B+ in 2011 into antenatal care service delivery, no specific guidance was given on how to do it.

Among the 141 health facilities surveyed the four models of care identified were: 

• A: facilities (n = 75) where women newly diagnosed with HIV are initiated and followed on ART at the antenatal clinic until delivery
• B: facilities (n = 38) where women receive only the first ART dose at the antenatal clinic with subsequent follow-up at the ART clinic
• C: facilities (n = 18) where women are referred from the antenatal clinic to the ART clinic for ART initiation and follow-up; and
• D: facilities (n = 9) serving as ART referral sites (not providing antenatal care).

Multiple variable analysis showed health facility factors significantly associated with ART retention included district location, patient volume (lower retention with high volume) and the model of care applied.

Facilities (model C) where women are referred from the antenatal care clinic to an ART clinic for ART initiation and follow-up were five times more likely to have high six-month retention rates than facilities (model B) where women receive only the first ART dose at the antenatal clinic with subsequent follow-up at the ART clinic.

There were no differences between the models in the proportion of women newly identified in antenatal care initiating ART, 81% (95% CI: 78-85).

However, there was a difference in the proportion of women not tested during antenatal care. Model B facilities had the highest proportion (32%), whereas model A had the lowest (18%). This was associated with client to HIV testing staff ratio, test kit stock-outs as well as model of care.

There were no differences in the number of women in the six-month cohort who had started ART under option B+. However, six-month loss to follow-up ranged from 7-20% with model D facilities having the highest retention rates and model B facilities the lowest.

Joep van Oosterhout told delegates that, while this study may not be representative of all of Malawi, it offers a creative approach to operational research. Use of high-quality routine government data and a large data set provides real-world findings.