By Clinical Informatics News Staff

November 26, 2013 | A new computational model released by the HIV Resistance Response Database Initiative (RDI) tailors HIV drug cocktails to individual patients, without the need to perform the expensive genotyping used in resource-rich settings to identify HIV strains. The new model, which is intended to direct changes in drug therapy and not first-line interventions, takes into account a patient's viral load before intervention, treatment history, and viral load after intervention, measurements more appropriate for resource-limited environments. Drawing on a database of therapy changes in HIV case histories, the model then recommends the most promising course of action. In a large test set of retroactive cases, this model, called HIV-TRePS, chose the best intervention more frequently than a model based on genotyping. The program, which is intended for research purposes only, is freely available at www.hivrdi.org/treps, and if further validated and refined, could be used to guide more effective therapies for HIV while controlling drug costs at financially strained clinics. The development of HIV-TRePS resulted in a paper published yesterday in the Journal of Antimicrobial Chemotherapy, with key contributions from scientists at the Desmond Tutu HIV Centre at the University of Capetown.