Background: Routine viral load (VL) to monitor the response to ART has been recommended by WHO since 2013. From 2012 routine VL testing to monitor ART was introduced in MSF projects in Lesotho, Malawi, Mozambique and Zimbabwe. All districts except Changara were rural settings where ART had been extensively decentralised. VL is performed annually in all sites except Malawi ( 2 yearly). To assess programmatic implementation of routine VL an analysis was carried to assess performance at each step of the VL algorithm.
Methods: Analyses were performed between January and November 2015 across six districts in four countries. Reviews of clinical and laboratory records of representative samples of patients were used to determine how each step of the routine VL algorithm (coverage of VL, uptake of enhanced adherence counselling, repeat VL testing (within 2-9 months), re-suppression, and appropriate switch to second-line ART) was implemented within a defined period according to local guidelines (18 months preceding date of analysis in Lesotho, Mozambique and Zimbabwe and 30 months in Malawi). Results were presented to programme staff and barriers for implementation identified.

SiteBuhera, ZimbabweGutu, ZimbabweThyolo, MalawiNsanje, MalawiRoma, LesothoMoz Changara (3000 copies/ml is threshold for action throughout algorithm)
Year routine VL testing started201220132012201320142013
Number of patients in the analysis476029787576278530693095
Coverage of routine VL testing (VL1)91%74%56%32%70%62%
VL > 1000 copies/ml14%15%9%20%10%40%
EAC documented for patients with VL >1000 copies/ml57%76%62%56%70%70%
Repeat VL test performed (VL2)68%67%55%40%42%23%
Resuppressed to <1000copies/ml43%39%46%32%8%22%
VL threshold for switch to second-line ART (copies/ml)100010005000500010003000
Eligible patients switched to second-line ART37%35%15%38%37%10%
[Outcomes of VL cascade analysis]

In those sites with low coverage of VL1 and VL2 challenges included lack of human resources to draw blood, dedicated staff to perform enhanced adherence counselling and lack of effective appointment and tracing mechanisms. Across all sites reluctance to task shift and decentralise second line ART care was cited as a barrier to switching.
Conclusions: This analysis demonstrated limited compliance with a routine VL algorithm based on WHO recommendations. Scale up plans for VL monitoring must address human resource issues and make implementation plans for provision of second-line in sites where ART care has been decentralised.