Reasons for loss to follow up in the Malawi Option B+ programme
Polly Clayden, HIV i-Base
The Option B+ programme in Malawi has reported considerable loss to follow up. Two presentations at AIDS2014 explored some of the reasons for this and offered some recommendations. [1, 2]
Malawi began Option B+ (universal lifelong ART for pregnant and breastfeeding women) in 2011. The programme’s introduction led to in a 7-fold increase in women starting ART for PMTCT in just over a year. But loss to follow up is considerable: 27% reported by 12 months from starting ART and 24% by 6 months in facilities with high volume of patients.
Hannock Tweyla presented findings from a study that looked at outcomes and reasons for loss to follow up of women from the Malawi Option B+ programme.
This retrospective cohort study, conducted at Bwaila Hospital, Lilongwe, used data from electronic medical records (EMR) and a patient-tracing programme. The hospital has the largest antenatal clinic (ANC) and maternity wing in the country with over 14,000 registrations annually. The Lilongwe District Health Office and partners provide the PMTCT service, which starts approximately 110 pregnant and breastfeeding women on ART every month.
Pregnant women with unknown HIV status accessing the service are offered a group counselling service and opt out HIV testing. Expert mothers provide support to the women for the first and all follow up visits. All HIV positive women are registered in the EMR system and started on ART on the day of diagnosis.
The tracing programme staff list women who miss an appointment by three weeks or more and women who consent (during registration) are traced up to three times by phone or visit. Tracing outcomes are categorised as: dead, uninterrupted treatment, treatment interruptions, self transfer out, stopped ART, never started ART and not traced.
Between September 2011 and September 2013, the investigators identified 2930 HIV positive women who started ART for PMTCT Option B+; of these 2,458 (84%) were pregnant (the remainder were breastfeeding). The women’s median age at ART initiation was 26 years (IQR 22-30) and follow up was 8.2 months (IQR 3.1-16.7).
Out of 2,930 women, 577 (20%) missed a scheduled appointment for at least three weeks; 272 only collected their antiretrovirals at the start of treatment and did not return. The overall incidence of loss to follow up was 23.5 % per year. Retention was 85%, 82% and 79% at three, six and 12 months respectively.
In multivariate analysis, factors associated with loss to follow up were: younger age, 13 to 24 vs 25+, adjusted rate ratio (ARR) 1.29 (95% CI 1.09-1.52); breastfeeding vs pregnant, ARR 0.63 (95% CI 0.49-0.89); and earlier year of Option B+ implementation, 2011 vs 2012 ARR 1.25 (95% CI 1.06-1.49), all p<0.001. Of note more recent data showed further decline in loss to follow up, 2013 vs 2012 ARR 0.41 (95% CI 0.29-0.58) – likely due to stabilisation of the programme.
Of 577 women, the investigators successfully traced 228 (40%) and established that 9 (4%) had died. Of the 219 women found alive: 67 (30%) had self-transferred to another ART clinic, 118 (77%) had stopped taking ART, 13 (9%) were on ART uninterrupted, 9 (6%) had treatment interruptions, 7 (5%) had not started ART and 5 (3%) declined to be interviewed.
Reasons given by women (n=111) for stopping ART were: travelled away (38%), transport costs (16%), limited understanding of ART (10%), suspected side effects (10%), very weak/sick (10%), non disclosure to husband (8%), religious belief (5%), forgotten (5%) and other reasons 44%.
The investigators noted that at 23.5% per year the loss to follow up rate among women started on ART in this Option B+ programme is greater than that reported in the general HIV population accessing ART for their own health of 9.3% per year. Almost half (47%) of women who were lost to follow-up received ART once and never returned for their appointment, leaving them at risk of vertical transmission. A considerable proportion of women could not be traced due to incorrect addresses documented in their clinic files – the investigators suggested that women could give false physical addresses because of fear of stigma and discrimination. A third of the women self-transferred to another clinic, which suggests national retention in PMTCT programme is underestimated.
The investigators concluded with a number of recommendations:
- ANC/ART clinics should further enhance post-test counselling by engaging HIV testing counsellors and expert mothers for ongoing counselling and psychosocial support.
- Targeted programmes for young women need to be established.
- ART clinics need to establish data linkages so information on patients that transfer can be shared.
- Further decentralisation of PMTCT services with good ANC/maternity services is needed.
A related presentation by Joep van Oosterhout showed results from a survey conducted across all health facilities providing PMTCT/ART services in the South East Health Zone of Malawi. The survey was undertaken to identify approaches to Option B+ service delivery (models of care) adopted in the national programme, in which great variation between retention rates has been reported (42-100%).
The investigators explored associations of the diverse models of care with programme performance indicators: uptake of HIV testing in ANC, uptake of ART, and retention on ART.
The South East Health Zone comprises 6 of 28 districts with 3.5 million inhabitants. There are approximately 154,000 pregnancies in the South East Zone per year and 22,500 (14.6%) of these are among HIV positive women. By June 2013 (when the investigators conducted the study), the South East Zone had 153 health facilities with integrated HIV care services.
The investigators used a structured questionnaire with questions covering: the availability of services, staff involved in PMTCT/ART service provision, the location where newly infected pregnant women are started on ART, the timing of adherence counselling for ART initiation, and the timing of transfer to ART or mother-infant clinic.
They used routinely collected health facility reports to determine uptake of HIV testing and counselling, and ART initiation for newly identified HIV positive pregnant women. They then evaluated 6-month outcomes for women registered as having started ART under Option B+ between July 2012 and December 2012. High HIV testing uptake was defined as greater than 85% in this evaluation and high retention on ART as greater than 92%.
Of 153 health facilities, 141 were included in the study of which the investigators identified four models of care:
- Facilities where women are started and followed on ART at ANC clinic until birth (n=75).
- Facilities where women receive only the first dose of ART at ANC clinic with follow up at ART clinic (n=38).
- Facilities where women are referred from ANC to the ART clinic for ART initiation and follow-up (n=18).
- Facilities serving as ART referral sites and that do not provide ANC (n=9).
They found that the proportion of women tested for HIV during ANC was highest in model A facilities 82% (95%CI 78-85) and lowest in model B facilities 68% (95% CI 61-74). The proportion of women starting ART was 81% (95% CI 78-85), across all four models. The highest 6-month retention rates were found in models C and D, 90% (95% CI 86-94) and lowest in model B facilities, 78% (95% CI 74-84).
In multivariate analysis, factors significantly associated with ART retention were district location, volume of patients (lower retention with higher volume) and the model of care. Model C facilities were 5 times more likely than model B to have high 6-month retention rates. Facilities with fewer than 31 women in the 6-month cohort were 5 times more likely to have high retention rates than facilities with the most women.
Dr van Oosterhout noted that approximately a quarter of pregnant women (18-32%) were not tested for HIV at ANC. HIV testing uptake was associated with ratio of women to testing staff, test kit stock outs and model of care.
In the survey 7-20% of women had defaulted Option B+ by 6 months and retention was associated with district location, patient volume and model of care.
The investigators concluded that overall the worse programme indicators were in model B facilities, where women only receive the first dose of ART at ANC.
An article in JAIDS earlier this year illustrated the “complex personal, societal, and structural barriers to continued HIV care facing postpartum women receiving ART through Option B+” revealed in a South African study.  The authors noted that although some of these barriers are true of all non-pregnant adults in the same setting, postpartum HIV positive women need targeted support and adapted programmes.
The study also highlighted the need for increased treatment literacy, stressing the importance of a return to HIV care for the mother’s own health after delivery.
“As countries seek to expand ART access and eligibility through Option B+, they must ensure that the unique needs of HIV positive postpartum women are addressed to retain them in care and facilitate adherence to ART”, the authors wrote.
James McIntyre gave an excellent update of Option B+ successes and challenges at the 2014 paediatric workshop. 
It is worth noting that most data so far is not looking at the plus part of Option B+ and refers to pregnancy, breastfeeding and immediately postpartum.
- Tweya H et al. Loss to follow-up among women in PMTCT Option B+ programme in Lilongwe, Malawi: understanding outcomes and reasons. 20th International AIDS Conference. Melbourne. 20-24 July 2014. Oral abstract THAX0101.
- van Oosterhout JJ et al. Elimination of mother to child transmission of HIV: performance of different models of care for initiating lifelong antiretroviral therapy for pregnant women in Malawi (Option B+). Oral abstract THAX0102.
- Clouse K et al. “What they wanted was to give birth; nothing else”: Barriers to retention in Option B+ HIV care among postpartum women in South Africa. JAIDS. 1 September 2014. Volume 67, issue 1 – p e12-e18.
- McIntyre J et al. Update on Option B+ successes and challenges. 6th International Workshop on HIV Pediatrics .18 -19 July 2014, Melbourne, Australia. Session: Emerging issues in PMTCT.