An empirical analysis of physician prescription behavior

My dissertation consists of two essays. In the first essay of my dissertation I model the relative impact of key drivers such as product, promotions, patient, and payor on the physician's prescription choice at the individual patient level. The estimates from this model can be used to appropriately target the physicians and to allocate resources across different components of the marketing mix more effectively. I incorporate observed and unobserved physician heterogeneity and patient history induced physician state dependence in our model.; I develop a physician choice model at the patient level and find significant state dependence in the prescriptions written for a patient. Patient history explains about 45% of the variance in physician choice. This results in a rather large bias while estimating detailing responsiveness and ROI without taking patient treatment history into account. A response based resource allocation generates much higher revenue for the pharmaceutical firm as compared with the current practice in the industry.; The high state dependence in patient prescriptions also highlights the importance of new patient acquisition. I find that re-allocating the marketing resources to target physicians with higher prescription volume and higher percentage of new patients increases market shares significantly over the currently used prescription volume based allocation and increases the ROI of marketing resources.; The second essay of my dissertation models the influence of the introduction of a generic drug on the switching behavior of physicians. The first essay shows that physicians typically do not switch patients currently taking chronic drugs if the drugs seem to be having the expected result. Physicians change this prescription behavior when a generic drug is introduced and switch patients to the generic drug due to the large difference in price between the branded drugs and the generic drug.; I study the change in this physician behavior before and after the introduction of a generic drug for the leading drug in a drug category treating a widely prevalent chronic disease using a basic model of physician behavior. I find a large and significant difference in the price sensitivity and the responsiveness to marketing activity before and after the introduction of the generic drug. I also find that introduction of generic versions of a leading drug leads to switching not just for patients on the same chemical compound but also from some patients currently taking different compounds. I explore further with an in-depth model of physician behavior for new patients and for existing patients separately after the introduction of the generic drug which physicians switch patients from other chemical compounds to the generic and which patients are most vulnerable to this type of switching.