I2B2 TRAINING VERSION 1.6.0.3 Informatics for Integrating Biology and the Bedside Reference Guide - 10/30/2012
Table of Contents i2b2 3 Will i2b2 benefit research...3 Web Site...3 Using the web site.. 3-9 Page 2 of 10
i2b2 i2b2 Informatics for Integrating Biology and the Bedside supports research by putting data in the hands of the user. Users can query de-identified patient data from data sources such as EPIC, Invision, and Signature. UTMB has loaded 10 years worth of data from these three data sources. Other data sources will be added to the centralized data repository as requested, assuming the resources are available to provide and implement the feeds. Once access is granted to i2b2, users are allowed to independently query de-identified data, thus eliminating bottlenecks such as resource limitations, access restrictions, and other commonly found constraints. Access will grant users the ability of determining patient counts and demographic information pertaining to a particular condition; such as age, ethnicity, disease, medication, etc. Will i2b2 benefit research? There are many benefits to the research community; most notably having open access to de-identified patient data. This tool provides the unique ability of determining patient counts and patient lists (sets). All queries ran within i2b2 are saved, and if authorization is provided, that same query can be used to retrieve identified data, thus providing specific details regarding the patient population. Retrieval of identified data is not available at this time, but will be offered in the near future. Web Site The online i2b2 application allows users to access i2b2 using a standard browser such as Internet Explorer, Firefox, or even Safari. Using the online application, queries can be processed that return patient lists (sets) and patient counts. The web interface is behind UTMB s firewall, so it is only available on campus or via VPN. Using the web site To log onto i2b2, please use the supplied user name and password. 1. Go to http://i2b2prod01new.utmb.edu/webclient and login with your supplied credentials. Figure 1. Login screen for i2b2 Page 3 of 10
After successfully logging into i2b2 the main page appears, and queries can be created, processed, and interpreted. If permissions are granted, further analysis can be performed on self-generated queries by clicking on the Analysis Tools on the top right hand side of the page. The main page has several containers; the ontology panel, query panel, work place panel, previous query panel, and the query status panel. Each panel contains different data elements, but they all interact together. Figure 2. Web interface for i2b2 Page 4 of 10
Figure 3. Dragging ontology sets to the query tool The groups in the query panel create the query. If you add more than one item to a single group you are utilizing the OR clause, whereas if you add items to other groups you are utilizing the AND clause. Once an ontology selection is dragged into a group, further restrictions can be applied to that group, such as date restrictions, occurrences, and exclusions. Click the titles in the desired group and supply the parameters to apply restrictions to data. If the date range title is selected, a popup similar to Figure 5 will appear. If the occurs > 0x title is selected, a popup similar to Figure 6 will appear, and users will have the ability of limiting their results to a defined number of occurrences. Exclusions can also be applied to sets of data, similar to saying not equals to; this feature is shown in Figure 7 and 8. Figure 4. Occurrences Figure 5. Including a date range on the parameters If a researcher wants to query all patients that were given Tylenol at least six times during their visit, using the occurs > 0x feature will allow this. This option allows the users to further refine the query to only include desired results. Page 5 of 10
Figure 6. Occurrences Figure 7. Exclude The exclude button within the query panel allows users to exclude certain results from being returned. Figure 8. Exclude Before running the query, users can select which type of results they want returned by clicking their checkbox next to their preference, as shown in Figure 9. Many users are only looking for the patient count to identify if we have a patient population to support a clinical study; other users require a little more analysis and will desire the patient list (set), which is a collection of data elements found in i2b2 such as age, race, and gender. The patient list (set) can be used to build demographic reports from within the analysis section of the web site. If further data elements are needed from the patient list, that are not found within i2b2, users will have to go to the governance body to get approval. When queries are processed, the query status screen, as shown in Figure 10, will display a summarized view of query results, such as count and approximate execution time. Page 6 of 10
Figure 9. Run Query While processing queries, a popup will appear where the name can be modified or kept, also two checkboxes for either retrieving patient counts, sets, or both. Once OK is clicked, the query will start running in the Query Status Panel; both a count of patients and the process time will appear. Figure 10. Running the query status results To clear the query window and to initiate another query, click the New Query button. This will erase the previously displayed output. Deleting a query from the Previous Query panel will permanently remove the query from both the database storing the query and from the interface. Page 7 of 10
Figure 11. Initializing new query Figure 12. Searching for codes An ontology set can be selected from the drop down, but if the specific ontology is not found, the search feature might be able to assist in locating the right one; click the Find Terms tab and enter the search terms container. This feature can perform conditional searches by the name, which includes conditional statements such as contains, exact, left, and right. Figure 13. Using filters Page 8 of 10
Figure 14. Analysis Tools Link Plug-ins assist in analyzing queries, but permissions must be granted before they can be used. One example, using the demographics plug-in as shown in Figure 16, displays a demographic report after dragging a patient list to the empty container. The demographic report is a graphical interpretation of the patient list (set). Figure 15. Plug-in Figure 16. Dragging and Dropping patient set to create report To utilize one of the plug-ins, you ll first need to click the plug-in (the plug-in will then occupy the empty container), and then drag the desired patient set to the text box that says Drop a Patient Set here. After dropping the patient set, click the View Results tab. Page 9 of 10
Figure 17. Plug-in for the patient set If the error in Figure 18 is displayed when trying to use the plug-in, then no permissions have been granted for the user to run analysis on patient lists (sets). Figure 18. Error received if you don t have permissions In Figure 19, demographic information pertaining to the patient list (set) is displayed. The patient list (set) report for the selected query displays age, sex, race, language, religion, and vital status. To retrieve this type of result set, or report: 1) Create a patient list (set) query 2) Click the Analysis link on the top right of the main page 3) Click Demographics (1 patient set) Simple Counts from the plug-in panel 4) Drag a patient list (set) from the previous queries panel to the empty text field 5) Click View Results (if you have the correct permissions, this will be allowed) Figure 19. Demographic report for patient list (set) Page 10 of 10