Time to first viral load testing among pregnant women living with HIV initiated on option B+ at 5 government clinics in Kampala city, Uganda: Retrospective cohort study.

• Large-scale rollout of viral load monitoring since 2013. • Implementation challenges persist. • Only 1.9% of the women had First Viral load Test (FVLT) done in a timely fashion. • The median time to FVLT post ART initiation was 12 months. • Constant review of EMTCT viral load data to achieve UNAIDS 3rd 95 by 2030 is needed. Timely viral load (VL) testing is critical in the care of pregnant women living with HIV and receiving anti-retroviral therapy (ART). There is paucity of data regarding the Time to First Viral Load (TFVL) testing in resource-limited settings. We extracted clinical and VL test data from records of a cohort of ART-naïve pregnant women living with HIV who initiated Option B + and were retained in care between 01 Jan 2015 and 31 Dec 2015. The data were verified against laboratory VL registers. TFVL (in months) was calculated based on the time difference between the date of ART initiation and FVL test. Descriptive and Cox regression analyses of data up to 30 Sep 2017 (33 months later) were done. Of the 622 records retrieved, 424 women were retained in care. Of 424 women retained in care, 182/424 (43%) had at least one VL result post ART initiation while 242/424 (57%) had no VL performed. Only 30/182 (16.5%) had a second VL. At six, nine, and twelve months, only 8/424 (1.9%), 47/424 (11.1%), and 94/424 (22.2%) had VL testing performed respectively post ART initiation. The median TFVL testing was 12.7 months (95 CI 11.6-13.7) post ART initiation. Across the five clinics, patient factors (age, gravidity, gestational age, marital status, and adherence at 12 months) were not significant predictors. A dismal 1.9% rate of achieving WHO-recommended TFVL testing and a median TFVL testing of twelve months post ART initiation were observed. The non-association of patient factors to these observations may suggest a serious need to review health system factors likely associated with these observations and their effective interventions. [ABSTRACT FROM AUTHOR]