The Use of Decision-Analytic Models in Parkinsonʼs Disease

Abstract

The aims of this review were to review decision-analytic models used to evaluate interventions in idiopathic Parkinson's disease (PD), and to consider the future directions for development of methods to model the progression of PD over time. A systematic search of the healthcare literature up to June 2010 identified model-based economic evaluations in PD. The modelling methods used in the identified studies were appraised using good practice guidelines for decision-analytic modelling. The review identified 18 model-based evaluations of interventions in PD. All models evaluated treatments targeted towards the motor symptoms of PD or the motor complications of PD treatment. There were no models identified that evaluated interventions targeted towards the non-motor symptoms of PD, such as neuropsychiatric problems or autonomic dysfunction. Consequently, models characterized disease progression in PD using clinical measures of motor functioning. Most studies (n = 13) evaluated medications, three evaluated diagnostic technologies and two examined surgical procedures. Overall, the models reported structural components and data inputs appropriately and clearly, although limited evidence was provided to support choices made on the structures used in the models or the data synthesis reported. Models did not adequately consider structural uncertainty or internal/external consistency. Modelling methods used to date do not capture the full impact of PD. The emphasis in the current literature is on the motor symptoms of PD, characterizing the clinical nature of disease progression, largely neglecting the important impacts of non-motor symptoms. Modelling methods reported for the motor symptoms of PD may not be suitable for future interventions targeted towards modifying disease progression in PD across the entire spectrum of PD. More comprehensive models of disease progression, including both motor and non-motor symptoms will be needed where it is important to capture the effects of interventions more broadly.