Figure 6: Cost-effectiveness planes showing the effect of viral load measurement frequency, format and threshold, all in the context of viral-load-informed differentiated care.

a, Viral load monitoring every 12-months is compared with every 6 months (every 2-year monitoring is excluded from the cost-effectiveness frontier due to unproven ability to base differentiated care on a 2-yearly value; however, if less frequent monitoring could be implemented without adverse health outcomes this would be cost-effective). b, Laboratory whole blood corresponds to dried blood spot (DBS). c, Alternative thresholds to define failure (viral load >200, >1,000 and >5,000 cps ml^-1) are compared in the context of laboratory monitoring every 12 months using plasma.