![[ Visit Client Website ]](images/banner.png)
BACKGROUND: Hospital discharge data are typically available by ZIP code of the patient’s residence; however, there are limited sources of population estimates at the ZIP code level to use in calculating rates. Census Bureau ZIP Code Tabulation Area (ZCTA) estimates are only available for Census years and vendor-based estimates may be expensive. The American Community Survey (ACS) recently began publishing 5-year estimates at the ZCTA level. In a partnership between our state’s Asthma and Tracking programs, we sought to calculate rates of asthma hospitalizations by ZIP code using ACS denominators for the Twin Cities metropolitan area to be incorporated into the Tracking program’s web-based interactive maps.
METHODS: We obtained data on asthma hospitalizations for the years 2007-2011 by ZIP code for the 7-county Twin Cities metropolitan area. We obtained ACS estimates for 2007-2011 by ZCTA and age group to use as population denominators. We calculated crude and age-adjusted rates with accompanying 95% confidence intervals incorporating standard errors in both the numerators and denominators. We then determined the cut-points (e.g., natural breaks versus percentiles) and other standards to use in mapping the rates.
RESULTS: We ran into several challenges that became visible when the hospitalization rates were mapped. First, rates for ZIP codes that crossed into counties outside of the metro area were not correctly calculated. Second, because asthma hospitalizations are relatively rare events, we needed to determine how best to map the many ZIP codes with rates based on small numbers of hospitalizations. Third, we needed to ensure that ZIP codes with zero hospitalizations could be distinguished from those with no population living in them (e.g., Minneapolis-St. Paul International Airport). Since ACS estimates by ZCTA are only available for 5-year windows, we will need to determine how best to update the maps when new years of hospitalization data become available.
CONCLUSIONS: Developing maps of hospitalization rates by ZIP code poses many challenges not encountered with county-level rates. ACS 5-year ZCTA estimates are a useful and reliable source of denominator data to use in calculating hospitalization rates by ZIP code.