Improving Visibility through Integrated Information
Health care has made huge changes in how content is managed in the 12 years since I left clinical practice to focus on health IT. As a clinical psychologist, all my intake assessments, progress notes, and treatment plans were hand-written. I moved into health IT to help digitize behavioral health care and have seen a sea change in enterprise content management that spans electronic health records, data warehousing and business intelligence, corporate intranets, cloud computing, and now interoperability and health information exchange.
Building an enterprise data warehouse (EDW) was our first step towards modernizing enterprise content management. Using reporting tools built into an electronic health record (EHR) restricts the queries to a single source system. In contrast, with a EDW queries can span multiple systems. We started building our EDW four years ago when we decided it was time to replace our legacy EHR. Looking forward, we knew we wanted to ask questions looking at metrics collected in both the legacy and the new EHR. In addition to EHR data, our EDW includes information from other key systems such as accounting, general ledger, HRIS, and identity management systems. We wrote our business rules into the extract-transform-load (ETL) processing for our EDW so that analysts wouldn’t have to know how source data had been coded in each system when authoring reports.
Before we had an EDW, analysts built reports using reporting copies of the transactional tables from the source system; each report used complicated tangles of CASE statements to account for programming changes in the EHR. This limited the number of people that could function as an analyst. Out of a staff of 30 people, only three had the depth of understanding to be able to author more complex reports. Now our ETL transforms the source data into clearly labeled and documented information easily translatable to business operations. Now I have over 10 employees authoring reports, and those aren’t limited to Data Sciences staff. Not only are other teams within Information Systems, such as Application Development and Applied Research and Process Analysis, able to retrieve information from the EDW, but clinical informatics and finance staff have self-serve access to business intelligence and the underlying EDW tables. This has driven adoption of data-driven decision-making in our organization.
To successfully manage treatment at a population health level, we have to move beyond exchanging data to sharing treatment plans across health records
Implementation of SharePoint
A second step in enterprise content management has been our implementation of a corporate intranet. We had used SharePoint for our intranet for years, but hadn’t imposed much structure or governance, so over time it became a cluttered, unusable mess. Last year, as we were implementing Microsoft Office 365 across our organization, we took the opportunity to re-work our intranet while migrating SharePoint to the Microsoft cloud. This time around we took care to ensure the intranet was segmented into high-governance and low-governance areas. The high governance areas are locked down, with relatively few people contributing information to key business documents, like our corporate policies and procedures. Other areas are more open, for use by different business units as collaboration spaces.
A key to our successful implementation of SharePoint was installing it in phases. In the initial stages, we mapped out our approach to the architecture of our intranet. An incremental implementation enabled an agile approach, helping us streamline the process as we moved from one department to another, understanding the operations and business needs of each as we migrated their content.
In addition to using SharePoint for our intranet, it also serves as a platform for distributing information and building business applications. Leveraging Microsoft Power Apps and App-It, a SharePoint workflow plug-in from K2, we can create applications both on our intranet and from mobile devices. Microsoft Azure ties information in our EDW to apps built in the Microsoft Cloud.
This cloud transition is just in time for a shift in health care delivery patterns. As payment models shift toward population health, we want to deliver a greater portion of our services “beyond the four walls” of our clinics. Having mobile access to clinical information in the field reduces the amount of time our community-based staff need to spend away from the people we serve. Leveraging the Microsoft tools to build these applications means taking advantage of the built-in security and identity management features within Office 365.
Focusing on Interoperability in Future
The next significant step in enterprise content management will be interoperability, connecting us to other healthcare information on the people we are treating, e.g., their primary healthcare provider, specialty care services or their visits to a hospital. We are implementing health information exchange through the Carequality framework with Denver Health Medical Center, our biggest partner providing both primary care and specialty physical health services. Both organizations are the safety net provider for our respective disciplines; so many people receive services from both organizations. Carequality provides governance and a consensus-based set of technology standards, plus a master provider directory (MPD), to allow EHRs to query one another directly. Queries return information on demographics, insurance coverage, services provided, diagnosis, problem lists, medications, allergies, and laboratory results in a consolidated clinical document architecture (C-CDA) format, allowing for clinical reconciliation of the structured data in the receiving EHR. One advantage of using the Carequality model of health information exchange, rather than submitting data to a central repository, is that our EHR is able to enforce compliance with federal privacy regulations for substance use disorders by verifying that consent to share information is on file before returning the C-CDA file.
Once we are up and live with our first connection using Carequality, it will be easy to connect with any other organization already using the framework. Previously, point-to-point data connections had to be configured one at a time–a costly, time-consuming endeavor. With Carequality, we can leverage its MPD to find and communicate with many different hospitals and health systems. Looking to the future, we are eager to implement Carequality’s roadmap for shared treatment planning. To successfully manage treatment at a population health level, we have to move beyond exchanging data to sharing treatment plans across health records. This will not only enable us to keep track of what we have done but assist us in mapping out what we plan to do to better coordinate care.