Date and Time
March 31, 2010
1:30 PM - 3:00 PM

Availability
Open to the public
No RSVP required

We examine whether decreased geographical access to the nearest emergency department (ED) results in changes in the mortality rates and health profile of patients with acute myocardial infarction (AMI), a group that has relatively homogenous patient characteristics and is sensitive to the availability of ED care. The primary data sources for ED availability are the American Hospital Association annual surveys and Medicare Provider Analysis and Review (MedPAR) for 1995-2005. Patient data were obtained from MedPAR. We identify the effects of ED access by comparing mortality outcomes (1-day, 7-day, 30-day) and patient health characteristics (age, comorbidity counts) between the following groups: (1) people who live in zip codes with no increase in driving time to their closest ED (the control group); and (2) people who live in zip codes that experience <10, 10-30, or >30 minute increases in driving time. We implement zip codes fixed-effects models and include year dummies and full interaction of patient demographics information. We estimate the models on all patients, as well as on sub population such as sicker patients and patients that share similar geographical characteristics.

In the general Medicare AMI population, there appeared to be no differences in mortality rates among patients in different categories of ED access. However, the health profiles of the patients changed if they lived in locations that experienced >30 minute increases in driving time: they were younger (by 0.77 year, p<0.01), arrived at the hospital with higher comorbidity counts (by 3%, p<0.05), and had a higher probability of needing cardiac catherization or angioplasty (by 4.6 percentage points, p<0.05) compared to the control group. Among the sicker AMI patients (defined as those that need angioplasty within 1 day of admission), an increase of driving time by 10-30 minutes had an adverse effect on mortality: such change increased the 1-day, 7-day, and 30-day mortality rates by 2.21, 3.5, and 4.4 percentage points, respectively (all p<0.01). This adverse effect disappeared by 9 months. The health profiles of the sicker patients did not change, except that those in areas with > 30 minutes increase arrived with higher comorbidity counts (by 9%, p<0.05).