Understanding and Risk-Adjusting Return Visits to the Emergency Department

Investigator: Reena Duseja, MD
Sponsor: PHS Agency for Healthcare Research and Quality

Location(s): United States


Findings from the proposed plan will be relevant to the academic and policy arenas and ultimately will enhance the quality of emergency medicine and improve patient outcomes. Emergency department (ED) return visits may be a useful indicator of the need to change ED processes or structure. Many questions remain regarding what leads to ED return visits and therefore how to intervene to prevent them. The proposed plan will help answer some of these questions.

We hypothesize that patient characteristics, ED structure and processes, and post-ED primary care influence rates of return visits. However, there has been little systematic study of these issues using population-based data (most studies are from single institutions). I propose to address these knowledge gaps in 3 ways: 
Aim 1 : To describe variation in risk adjusted 3-day revisit rates in all general acute hospitals in Arizona, California, Florida, Hawaii, Nebraska, and Utah. This includes the development of a risk adjustment model, using administrative data, which will allow me to perform the first study identifying revisits to all EDs (rather than just the index ED). Creating the risk model will permit us to identify high and low performing EDs for the next aim. 
Aim 2 : To use qualitative methods to identify patient, ED, and primary care factors or interventions that may be associated with variations in ED return visit rates. Qualitative methods may enable us to elicit from interviewees factors contributing to revisit rates that have not yet been described. 
Aim 3 : To determine the association between the presence of ED-level interventions to prevent ED revisits and the risk adjusted rate of ED revisits. In this aim, we will develop and administer a survey to assess for the presence of interventions to prevent ED revisits. Survey topics will be selected based on our initial conceptual model and findings from Aim 2. We will use the survey data to assess whether there is an association between ED interventions and risk adjusted revisit rates as calculated in Aim 1. These three aims will provide important insights into variations in ED revisit rates as well as possible interventions to help lower them.