Identifying Risk Factors for Poor Glycemic Control among Emergency Department Patients and Improving Linkage to Outpatient Care

PROJECT SUMMARY Of the millions of Americans who visit an emergency department (ED) each year, there are substantially higher rates of ED use among racial and ethnic minorities and patients with significant socioeconomic barriers. However, EDs rarely provide any sort of preventive care, and the care they provide is often not well coordinated with critically important outpatient care. The consequences are higher rates of undiagnosed diseases like diabetes among ED patients, which then go untreated, exacerbating health disparities, poor glycemic control, diabetic complications, and early diabetes-related death. Given the high volume of ED patients, especially those who are high-risk and have poor access to care, ED-based diabetes screening can help identify undiagnosed cases and presents an important opportunity to link newly diagnosed patients to appropriate outpatient care. Nearly 1 in 5 Americans visit an ED in any given year and more than 50% of ED patients receive some type of blood test. NYU Langone Health has recently launched a system-wide initiative to screen ED patients already receiving blood work using HbA1c tests. Our preliminary data demonstrate that this screening approach can help identify ED patients with previously undiagnosed diabetes, especially among minorities and those without a regular primary care provider. The most critical practical consideration that has emerged from this initiative is how to improve follow-up with outpatient care. Only 23% of these newly diagnosed patients had a follow-up outpatient visit within our healthcare system. While these patients may have had visits elsewhere, this statistic is consistent the low rates of follow-up after ED visits. Therefore, this proposal seeks to address the following questions: (1) which patients are least likely to follow-up after a new diagnosis of diabetes, (2) what are the primary reasons why these patients do not follow-up as outpatients, and (3) whether targeted interventions can help bridge these gaps in care. Therefore, this study will develop clinical algorithms using data from electronic health records to identify which patients are least likely to follow-up for patient care after being newly diagnosed with diabetes. It will also perform detailed qualitative interviews among these newly diagnosed ED patients who fail to follow-up to obtain a comprehensive understanding of their barriers to care. We will also perform a randomized controlled trial of telemedicine to link ED patients identified as having newly diagnosed diabetes to appropriate outpatient care. This study will help improve diabetes surveillance by testing a scalable model to transform EDs into a clinical setting that can effectively screen and refer the millions of high-risk patients that visit EDs in the United States each year, thereby reducing disparities in diagnoses, the frequency of diabetic complications, and early diabetes-related death.