Open AccessFactors associated with loss to follow‐up among women in Option B+ PMTCT programme in northeast Ethiopia: a retrospective cohort study
Author(s) -
Mitiku Israel,
Arefayne Mastewal,
Mesfin Yonatal,
Gizaw Muluken
Publication year - 2016
Publication title -
journal of the international aids society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.724
H-Index - 62
ISSN - 1758-2652
DOI - 10.7448/ias.19.1.20662
Subject(s) - medicine , breastfeeding , hazard ratio , retrospective cohort study , proportional hazards model , confidence interval , cohort , survival analysis , cohort study , demography , obstetrics , pediatrics , sociology
Ethiopia has recently adopted lifelong antiretroviral therapy (ART) for all HIV‐positive pregnant and breastfeeding women (Option B+ strategy), regardless of CD4 count or clinical stage. However, the exact timing and predictors of loss to follow‐up (LFU) are unknown. Thus, we examined the levels and determinants of LFU under Option B+ among pregnant and breastfeeding women initiated on lifelong ART for prevention of mother‐to‐child transmission (PMTCT) in Ethiopia. Methods We conducted a retrospective cohort study among 346 pregnant and breastfeeding women who started ART at 14 public health facilities in northeast Ethiopia from March 2013 to April 2015. We defined LFU as 90 days since the last clinic visit among those not known to have died or transferred out. We used Kaplan‐Meier and Cox proportional hazards regression to estimate cumulative LFU and identify the predictors of LFU, respectively. Results Of the 346 women included, 88.4% were pregnant and the median follow‐up was 13.7 months. Overall, 57 (16.5%) women were LFU. The cumulative proportions of LFU at 6, 12 and 24 months were 11.9, 15.7 and 22.6%, respectively. The risk of LFU was higher in younger women (adjusted hazard ratio (aHR) 18 to 24 years/30 to 40 years: 2.3; 95% confidence interval (CI): 1.2 to 4.5), in those attending hospitals compared to those attending health centres (aHR: 1.8; 95% CI: 1.1 to 3.2), in patients starting ART on the same day of diagnosis (aHR: 1.85; 95% CI: 1.1 to 3.2) and missing CD4 cell counts at ART initiation (aHR: 2.3; 95% CI: 1.2 to 4.4). Conclusions The level of LFU we found in this study is comparable with previous findings from other resource‐limited settings. However, high early LFU shortly after ART initiation is still a major problem. LFU was high among younger women, those initiating ART on the day of HIV diagnosis, those missing baseline CD4 count and those attending hospitals. Thus, targeted HIV care and treatment programmes for these patients should be part of future interventions to improve retention in care under the Option B+ PMTCT programme.