The United States' health care system has faced growing per capita use, and subsequent crowding, of the emergency department since the 1990's, evidenced by longer wait times and the institution of ambulance diversions. Emergency departments are but one element of a complex system, with difficulties which persist despite policy efforts.

This dissertation asserts the endogenous point of view is a useful and practical approach. The initial proposition is there exists an endogenous mechanism contributing to ED crowding. A specific mechanism was proposed after a thorough review of the literature. This mechanism proposed primary care physicians were using the ED as a backdoor for hospital admissions to avoid barriers raised by payers created to reduce unnecessary hospitalizations.

Disconfirmatory interviews with ED doctors, recruited via connections with colleagues, were used to evaluate the initial model. Participants rejected the specific proposed mechanism, but all offered an alternative endogenous mechanism: the rise of hospitalists leading the ED becoming the front-door for hospital admissions. This resulted in patients who would have been admitted directly under their PCP in the old model to instead be directed to the ED.

Participants cited many advantages of this method, such as faster testing in the ED, shorter stays under hospitalists and PCPs avoiding travel. All of these may support a hospital administrator encouraging the use of hospitalists. Administrators should be aware, however, of the increased load this will create for the ED. The hospital must be prepared for the additional patients and would benefit from considering how to best streamline admission through the ED.

The process worked wonderfully to elicit not just a disconfirmation, but alternate proposals, driving the research forward naturally.

The interview results provoked a return to the literature for links between hospitalists and the ED. A new model was created based on the participants' feedback and the literature review. Future research should make a new interview booklet and conduct further interviews with hospitalists and hospital administrators to challenge and flesh out the new model.