ABSTRACT
Major depressive disorder (MDD) is the leading cause of disability in many countries. Designing and evaluating clinical trials of antidepressants is difficult due to the pronounced and variable placebo response which is poorly defined and may be affected by trial design. Approximately half of recent clinical trials of commonly used antidepressants failed to show statistical superiority for the drug over placebo, which is partly attributable to a marked placebo response. These failures suggest the need for new tools to evaluate placebo response and drug effect in depression, as well as to help design more informative clinical trials. Disease progression modeling is a tool that has been employed for such evaluations and several models have been proposed to describe MDD. Placebo data from three clinical depression trials were used to evaluate three published models: the inverse Bateman (IBM), indirect response (IDR) and transit (TM) models. Each model was used to describe Hamilton Rating Scale for major depression (HAMD) data and results were evaluated. The IBM model had several deficiencies, making it unsuitable. The IDR and TM models performed well on most evaluations and appear suitable. Comparing the IDR and TM models showed less clear distinctions, although overall the TM was found to be somewhat better than the IDR model. Model based evaluation can provide a useful tool for evaluating the time course of MDD and detecting drug effect. However, the models used should be robust, with well estimated parameters.
