Loss to follow up is high in pregnant women not otherwise indicated for treatment starting ART in Malawi's Option B+ programme.
Lyson Tenthani showed findings from an analysis of retention in care of HIV positive pregnant and breastfeeding women started on ART in the Malawi Option B+ programme by the Ministry of Health Malawi and IeDEA South Africa.1 This was a late breaker presentation at the 5th International Workshop on HIV Pediatrics.
Malawi pioneered the Option B+ strategy -- i.e., all pregnant and breastfeeding women starting lifelong ART -- in 2011. It has led to a massive 763% increase in HIV positive pregnant women receiving ART.2 But there are concerns about acceptability, retention and adherence particularly among asymptomatic women.
This cohort study analysed facility- and patient-level data from sites enrolled in the national electronic ART register across the country. The investigators used meta-analyses, logistic regression and competing risk survival models to examine at site- and patient-level predictors of loss to follow-up.
The evaluation included facility level data on a total of 21939 women from 540 sites and patient-level data on 28428 women from 19 ART sites with electronic medical record system.
At 6 months after starting ART 82% of women were alive on ART and 17% were lost to follow up -- a small proportion had either died or stopped ART.
In multivariate analysis controlling for age, facility type and random effects for site Option B+ women starting ART in pregnancy were 5 times more likely to be lost to follow up with no follow up visit than patients initiating treatment with low CD4 and/or WHO stage 3/4. Option B+ women starting during breastfeeding were twice as likely to have no follow up visit. See table 1.
Notably 41% of patients had no CD4 data -- one of the reasons for starting B+ Malawi was because of lack of access to CD4 testing.
|Table 1. Option B+: Predictors for Loss to Follow Up|
|No follow up visit||Lost at second visit||Lost during 3-8 months on ART|
|Indication||OR (95%CI)||p-value||OR (95%CI)||p-value||SHR (95% CI)||p-value|
|Low CD4 or stage 3/4||ref||<0.0001||ref||0.1972||ref||0.0091|
|B+ pregnant||5.18 (4.35-6.17)||1.90 (1.46-2.48)||0.91 (0.67-1.24)|
|B+ breast-feeding||2.26 (1.83-2.80)||0.90 (0.64-1.27)||0.60 (0.42-0.85)|
|Pregnant and low CD4 or stage 3/4||1.88 (1.23-2.90)||0.45 (0.18-1.10)||0.61 (0.31-1.19)|
SHR: Substitution hazard ratio
Performance across sites was variable, overall 37% of sites did well with less than 10% of all patients lost to follow up six months after starting ART but 33% of the sites had greater than 20% loss to follow up.
Analyses of site level predictors for loss to follow up in 368 sites revealed Ministry of Health sites did less well (faith based as reference) OR 1.19 (0.94-1.50), p=0.04 as did central hospitals (health centre as reference) OR 2.7 (0.92-7.94), p=0.47, that is larger sites with higher proportions of B+ women.
If ever there was a job for the community!
Most Option B+ patients who were lost to follow up started ART on the day of their diagnosis and never came back. Unsurprisingly these women were less likely to return to clinics than pregnant women who started treatment after a little time to come to terms with their HIV status (OR 1·7, 95% CI 1·4-2·2).
Much of the discussion following the presentation focused on understanding the barriers to retention in care and including preparation and support for women in Option B+ programmes. The sites with a low rate of loss to follow up -- demonstrating that good retention in care can be achieved -- need to be scrutinised for generalisable strategies.
The crucial role of treatment literacy and preparation for women, starting from the first visit, cannot be emphasised more.
There are many excellent models of community support, treatment education, adherence clubs and one to one peer support. These programmes need to be taken seriously, communities must be included in their design and funding must cover their successful execution.
A related presentation at IAS 2013 from Uganda also showed high loss to follow up among women associated with rapid initiation of ART.3
This study compared women starting ART in ANC to those starting in labour. In this programme 190 women tested positive and 92% started ART in the ANC; 82% returned to receive their CD4 results. There were also 162 women who started in labour and only 12% of these women returned.
Although there is no opportunity to delay ART initiation at this very late stage of pregnancy, the same attention to support and education is likely to improve this tiny proportion of women returning to the facility.
Links to other websites are current at date of posting but not maintained.