A probabilistic approach for the evaluation of pharmacological effect induced by patient irregular drug intake.

Fine individual drug intake data, generally collected by electronic monitoring devices, reveal that individual marked random patterns are likely to persist through long therapeutic periods. This work aims to establish the relationship between irregularity in drug intake and its potential impact on therapeutic outcomes, which will also serve as a basis for more objective interventions. First we proposed a direct way to extract the necessary information representing the patient drug intake history. To provide a fair evaluation of the pharmacological performance, we revisited several classical pharmacological indices and proposed new ones in the stochastic context of patient's drug intake irregularity. To illustrate our procedure, we have considered two cases of HIV treatment using a combination of lopinavir/ritonavir (Kaletra@) for once daily and twice daily regimens. We have quantified the impact on therapeutic effect of various characteristics in dosing histories, namely missing doses and deviations from nominal times. Using our newly defined pharmacological indices, we clearly showed the ability of our probabilistic approach in measuring the impact of noncompliance. As a direct fallout, we have discussed strategies to attenuate the impact of noncompliance through an optimal design of dosing regimen.