OBJECTIVES: What happens when technologies leave Randomized Controlled Trials and approach the real world? Sometimes no answers are provided to this question. The aim of the COALA study was to evaluate, taking the Italian NHS (payer) perspective, the lifetime cost-utility of lopinavir/ritonavir vs. atazanavir + ritonavir, with a tenofovir-emtricitabine backbone, in a sample of 319 naïve patients starting from, and developing, the assumptions of the Broder (2011) Markov microsimulation model. METHODS: The health and monetary outcomes associated with ATV+r and LPV/r regimens, were analyzed considering: 8 health states, incidence of diarrhea and/or hyperbilirubinemia, AIDS events, opportunistic infections, and CHD events and, for the first time in an economic evaluation concerning HIV+ patients, CKD. Patients were enrolled in one of two scenarios: LPV/r or ATV+r (ATV1), and followed first or second line treatment. The two lines were both modelled at an individual patient level. A sensitivity analysis regarding ATV+r arm (ATV2) was assessed, in order to replicate the results of the CASTLE trial study, and to compare them with an ATV1 arm. The total cost per patient was calculated considering drugs costs, clinical effectiveness data, as well as institutional guidelines, protocols and the Lombardy Region reimbursement tariffs. RESULTS: LPV was assessed with a gain in terms of QALY equal to 0.150 to 0.447. The model results show a higher incidence of cardiovascular events (6.5 per 1,000 inhabitants) for the LPV arm, and a lower incidence of CKD (27.1 vs. 110.9-135.3) for LPV. CONCLUSIONS: Considering the economic performance, there was an annual economic per capita advantage between €73 and €138. This, in terms of QALY and reduction of costs, reveals that LPV/r treatment gives a higher economical performance and is more cost-effective compared with ATV+r.