The system provides an estimate of in vivo efficacy of commonly used antiretroviral drug combinations based on user-defined information of HIV genotype (mandatory) and additional patient information (recommended). The engine is not intended as a replacement for standard of care but can be used by HIV specialists as an additional treatment decision support tool. Although analysis of a large data set of medical records has shown a good performance of the system, there is no warranty that its use will improve patient health. HIV care must rely on a solid knowledge of the complex host-virus interaction and proper consideration of patient status and commitment.


In November 2010 the system has been trained and tested on a set of data derived from clinical practice records of more than 48,000 HIV patients collected in Europe. Treatment success was defined as achieving a decrease of at least 2 log or an undetectable HIV RNA load at week 8 after treatment start. More precisely, the system has been trained on 6.221 Treatment Change Episodes (TCEs) of which 252 are RACs (DRV, TPV, ETR). The CV performance is AUROC(Area Under the ROc Curve)=0.76-0.78 (0.74-0.75 on RAC) and the test performance AUROC=0.77-0.78 (0.66-0.69 on RACs).

Plot1 Plot2 Plot3 Plot4 Plot5 Plot6 Plot7 Plot8