Engagement in care among women and their infants lost to follow-up under Option B+ in eSwatini

Abstract

Background: Universal antiretroviral treatment (ART) for pregnant women with HIV, Option B+, has been adopted widely for prevention of mother-to-child HIV transmission (PMTCT). Some evidence shows high loss to follow-up (LTF) under this model. However, gaps in data systems limit this evidence. We collected additional information for women and infants LTF from Option B+ in Eswatini to assess more accurate outcomes.

Methods: LTF at 6-months postpartum was assessed using facility data. Additional data was gathered from: 1) the national ART database and paper records; 2) patient tracing; and 3) interviews and abstraction from patient-held records. Engagement in care was defined as any clinic visit within 91 days before or after 6-months postpartum or completion of a documented transfer; or, for those traced but not completing study interviews, visits at 6-months postpartum or later (for infants), or visits within 3-months of tracing (for women). Multivariable loglinear models were used to identify correlates of engagement.

Results: One-hundred-ninety-four (44.7%) of 434 LTF women had outcomes ascertained, including 122 (62.9%) women engaged in care. Among 510 LTF infants, 265 (52.0%) had ascertained outcomes, including 143 (54.0%) engaged in care, 47 (17.7%) pregnancy losses, and 18 (6.8%) deaths. Seventy-two of 189 live infants (38.1%) with ascertained outcomes had a 6-week early infant diagnostic (EID) test. Among women with ascertained outcomes, gestational age of 20+ weeks (vs. fewer than 20 weeks, aRR 0.80; 95% CI 0.68-0.94) and age 25-29 years (vs. 15-24 years, aRR 0.81; 95% CI 0.67-0.97), were associated with lower engagement; initiating ART after first ANC visit was associated with higher engagement (vs. at first ANC visit, aRR 1.12; 95% CI 1.04-1.21). Among infants with ascertained outcomes, mother not initiating ART was associated with lower engagement (vs. ART at first ANC visit, aRR 0.71; 95% CI 0.54-0.91).

Conclusion: Substantial numbers of women and infants classified as LTF under Option B+ were engaged in care, though a suboptimal level of 6-week EID testing was observed. These findings highlight a need to improve coverage of routine EID testing, and improve data systems to better capture PMTCT patient outcomes.

Fig 1. Outcomes from data abstraction, patient tracing, and study interviews for women.

Fig 2. Outcomes from data abstraction, patient tracing, and study interviews for infants.