178350 Geography of primary care users of the Bellevue Emergency Department

Monday, October 27, 2008: 11:45 AM

Arkaprava Deb , NYU School of Medicine, New York, NY
Jonathan E. Soverow, MPH , NYU School of Medicine, New York, NY
Ann Nguyen, MD , NYU School of Medicine, New York, NY
Joseph Rabatin, MD , Brown University School of Medicine, Providence, RI

Neighborhoods may make heavy use of the emergency department for primary care. Identification of such communities may facilitate interventions to provide these patients with appropriate primary care resources.


To identify and map New York City neighborhoods that disproportionately use the emergency department for non-urgent care.


The resident physician assigned to each patient was asked whether the case was “urgent” or “non-urgent,” and inter-rater reliability was measured with standardized cases. Chart review was used to procure address and additional information (chief complaint, age, gender, time of presentation, insurance status, ethnicity, and primary language). Data collection occurred during August 2006 and 2007. All cases were mapped using geographical software (ArcGIS v9.0). Clusters were investigated spatially using the scan statistic (SatScan v7.0.3) to compare urgent (controls) versus non-urgent (cases) presentations. Associations between covariates and outcomes were analyzed using logistic regression (STATA v9.0).


Out of a total 1001 visits, 49.1% (491) were deemed “urgent” and 50.9% (510) “non-urgent.” Inter-rater agreement on urgency was high [kappa: 0.89; 95% CI (0.87, 0.91)]. The scan statistic identified one significant cluster in the 2007 dataset (relative risk: 1.60; p-value=0.017) and one in the combined 2006/2007 dataset (relative risk: 1.50; p-value=0.007), each centered at distinct points in Brooklyn, New York. Logistic regression found no associations between the covariates and urgency status.


This study identified two areas that merit primary care attention and demonstrated that geography influences emergency department use. Surveying an urban emergency department can yield knowledge about the landscape of primary care resources.

Learning Objectives:
1. Demonstrate a method to identify primary care-deficient neighborhoods 2. Describe how geographical statistics and software can be used to analyze non-urgent users of an emergency department 3. List two neighborhoods worthy of primary care attention in New York City

Keywords: Geographic Information Systems, Primary Care

Presenting author's disclosure statement:

Qualified on the content I am responsible for because: I conceived, planned, and conducted the research being presented.
Any relevant financial relationships? No

I agree to comply with the American Public Health Association Conflict of Interest and Commercial Support Guidelines, and to disclose to the participants any off-label or experimental uses of a commercial product or service discussed in my presentation.