Effects of Decreased Emergency g y Department p Access on AMI Health Outcomes Presenter: Yu-Chu Shen Naval Postgraduate School and NBER Co-authors: Renee Hsia, UCSF Laurence Baker, Stanford We thank the Robert Wood Johnson Foundation’s HCFO Initiative grant #63974 for financial support B k d Background Between 1995 and 2005, number of ED went down by 23% while annual ED visits went up by 19% (NHAMCS 2007). Decrease access to ED is well documented in the literature. Anecdotal reports hinting at adverse effects on patient care, little empirical evidence. Th R hQ i Three Research Questions Does increased driving time to ED result in increased mortality rates among AMI patients? Does increased driving time to ED result in changes in health profile of AMI patients who arrived alive in the hospital? If decreased ED access has an adverse effect on patient outcomes, is the adverse effect transitory or permanent? Hypotheses on the Effect of Increased Time to ED on AMI Patients • Patients who survived to have hospital records are younger in areas with decreased ED access than those in areas with no decrease in ED access access. • The observed mortality rates of all AMI patients would increase when ED access decreases decreases, but the observed mortality rates of patients who survived to have a hospital record might not. • Patients in areas with decreased ED access arrive at the hospital with higher probability of needing immediate procedure. A Hypothetical Example Compare between two groups of patients AMI mortality rates Time to ED did not increase b/w 95 and 96 Time to ED Dif increase b/w 95 and 96 1995 1996 Dif-in-diff 14% 9% 16% 14% 2% 5% 3% Data • To T identify id tif ED availability il bilit – AHA surveys – California OSHPD • Patient data – 100% Medicare Provider Analysis and Review (MedPAR) and denominator file • Location data to calculate driving time – Mailer software (patient) – Dr. Jill Horwitz (hospital) • Hospital data – Healthcare Cost Report Information System (HCRIS) Defining AMI Outcomes Mortality rates of different time horizon – 7-day, 30-day, 90-day Age and care intervention – Age at the time of hospital admission – Whether receive PTCA on the day of admission Defining ED Access Change Categories • Control group: people who live in zip codes with no increase in driving time to their closest ED; • Treatment group: people who live in zip codes that experience the following amount of increases in driving time between 1996 and 2005. – <10 minutes – 10-30 minutes – >=30 30 minutes Empirical Approach (1) • Identify the effects of ED access by comparing changes in mortality outcomes and health characteristics between control and TT groups • Mortality rates and PTCA – Fixed-effects linear probability model • Age – Fixed-effects OLS • Robust standard errors to account for clustering at zip code level Empirical Approach (2) • Control variables: – – – – Whether nearest facility has cath lab Patient demographics, comorbidity Admitted hospital characteristics Year trend • Model 2: Add transition yyear indicators – Timing issue: only observe closure on yearly basis – Transitory or permanent effect • Additional analysis – Limit control group to those with limited access to ED in baseline Figure 1. Distribution of ED Access Change on US Map Change in driving time to nearest ED 1996‐2005 No Data Up to 10min increase No increase in driving time Up to 30min increase Up to 5min increase More than 30min increase Main Regression Results (1) AMI Mortality Rates 7-day 30-day 90-day ED Di Distance t Ch Change C Categories t i All yrs after driving time inc by <10 0.0025 0.0034 0.0067* 0.0007 -0.0077 -0.0002 0.0102 -0.0014 -0.0028 -0.0036* -0.0031+ -0.0025 minutes All yrs after driving time inc 10-30 minutes All yrs after driving tie inc by >=30 minutes Closest facility has cath lab Main Regression Results (2) AMI Health Profile Age PTCA on day of admission ED Distance Change Categories All yrs after driving time inc by <10 0.234** 0.234 -0.006* 0.006 0 265 0.265 -0 0.005 005 -0 620* -0.620* 0 035** 0.035** -0.048 0 048 0 003+ 0.003+ minutes All yrs after driving time inc 10-30 minutes All yrs after driving tie inc by >=30 minutes Closest facility has cath lab Transition Year Indicators (Limited ED Access in Baseline) AMI Mortality Rates 7-day 30-day 90-day ED Transition T iti Yr Y Categories C t i (-3, -2) years before change -0.0001 0.0038 -0.0001 (-1, 0, 1) years during change 0.0046+ 0.0097** 0.0085* (2, 3) years after change 0.0068+ 0.0084+ 0.0090+ Discussion • Overall, ED access deterioration only affects a very small percent of the population • In general, results consistent with conceptual y framework’s hypotheses • Cath availability appears to be more important than time at improving short-term short term mortality rates. • Some evidence of transitory effect