A question you might have at this point is, with all those data available, how can I make it useful to me? How can I use these data to understand my population, my town, or my specific research interest? Now, I will return to the where of the data. Going back to the graphic from the CDC, one way to customize data or make it more useful or targeted is to use maps, the where to convey information contained in the data. There are several resources for this kind of data exploration. These several resources exist online to map data. Some of these ArcGIS, UDS Mapper allow data to be uploaded and mapped. The others are mapped and mappable repositories of existing data. ArcGIS is a software system for mapping data, while some data such as county boundaries and some socioeconomic data exists within the system. It is primarily a tool for geospatial analysis. UDS Mapper is also a system for mapping data, but it concentrates on the geographic reach of federal health center programs and showcases healthcare reach and coverage. County health rankings. A project of the Robert Wood Johnson Foundation uses maps to display county rankings of a variety of social determinants and outcomes of help. The final two, City Health Dashboard and the CARES Engagement Network are mapped repositories of existing data. I will speak in detail about these two. City Health Dashboard developed by the public health and population health departments at New York University, display survey data from the Behavioral Risk Factor Surveillance System, or BRFSS, to allow users to explore health and social determinant data for the largest 500 cities in the United States. Many of these datasets are available at the census track level, and comparison of similar cities based on a number of characteristics as possible. Let's take a look at some of these available data. Since I live in Rochester, New York, I'm going to select that city. Data are available by different levels of geography. For example, city level, or census track, and by category, such as health behavior and clinical care. As an example of the data available, I will select health behavior and binge drinking. This brings me to a screen which shows Rochester, New York, and the metric. In reading more about the metric, shows details about the data collection and availability. Since these data are available by census track, scrolling down, you can see the variability by tract in binge drinking in Rochester. The information at the top tells you where Rochester falls among the 500 cities in terms of the value, and we can see that Rochester is slightly lower or better than average in terms of binge drinking. Another nice feature of this site is the comparison across cities. We can search for comparable cities. We can either select a city from the 500 cities, or we can filter to find comparable comparison cities. If I select to compare on general characteristics such as population and location, I see that one city has matched Rochester and that's Yonkers, New York. Selecting Yonkers. I can see that Yonkers too is slightly better than average than the 500 cities. The CARES Engagement Network allows users to customize by area and topic, both health needs, assessment reports, and maps. The site houses tens of thousands of sets of data. Mapping layers, and many are nationally embedded datasets, such as purpose and the census tract data mentioned previously. Let's explore the resources available here. First, I will build a report. I can select a state, and selecting from the list of states, I select Texas. This brings up a list of counties, and then I will select Austin County. Having set my location, I now look at the data indicators. From here I want to select some social and economic factors. I will select no high school diploma. If you select the eye next to any of the indicators, it gives you information about the census track data availability, and how the data were collected, and what data are available. Closing this, I want to select an additional health behavior to go with this in my report, so I select alcohol consumption. Again, note the eyes giving you information about all of the data indicators. Having selected my county and my data indicators, I then see what's available in the reports. First, if we open up quick facts, we see a general list with information with the location summary and the state average for these variables. If I go back to Reports and pick comprehensive report, I get much more detailed information about these factors. The location, data tables, census track information, a dial indicator telling me how Austin County compares with Texas and Texas at large and the United States overall. There's also information by gender, ethnicity, and race. Notice that the first sediment formation is the education, no high school diploma, and the second is health behaviors. I go back up to the information you see that any of this is available to download in PDF form to share a link, and to export an Excel format. Going back to the CARES Engagement Network, now, I'd like to make a map. The care site has literally tens of thousands of sets of data available, as I mentioned, and I can search by keyword. I would like to select alcoholic beverages by the Nielsen folks. Then to make this the same set of data available by year, I'm going to select binge drinking from 2014 from the 500 Cities Project that we saw in our earlier example with the City Health Dashboard. I'd like to add these to the map. From the resulting map, you can see that cities in the upper Midwest have high percentages of binge drinking. Because of the geographic specificity of the second set of data, it's not placed on the map. I need to zoom in. If I zoom in to say Milwaukee, Wisconsin, here, you can see that I have the 500 cities data in the purple circles overlaid with a census track map of alcoholic beverage expenditures. Now, you can see that this overlay is a bit complicated because it's difficult to see what's going on. Let's make the binge drinking not the place, but also the census track. This gives us more data and more information, more specificity. But again, it's a little bit difficult to see. CARES Engagement Network makes a tool available that you can look at the two layers simultaneously so that you can swipe one layer and the other layer to see the variability by tract. Now let's go back to one of the first slides, this time with health improvement and health survey added to the Google Trends. I think this reflects two current realities in population health assessment and improvement. First, interest in health improvement appears to be lagging behind the interest in data. Do we have so much data that the implementation to actually improve health is complicated or cumbersome? Perhaps. Second, the reliance on survey data to reflect health data is waning, as methodologies used in survey research, such as calling landline phones to get participants, are becoming outmoded. Finally, I leave you with some international resources of health data statistics and downloadable datasets. The World Health Organization Europe site contains information on topics ranging from violent crime to nutrition to infectious disease. The Organization of Economic Cooperation and Development, or OECD site, contains information about 34 member country organizations, including the US, Mexico, Canada, Japan, Korea, and many European countries. The Georgetown Library site is a compendium of international health data and includes datasets for specific countries. I hope this session has helped you identify some data sources for your own use and given you an overview of some population health data exploration tools. Thank you.
