Download PDF Download Word Document Access module resources

MODULE 7: Creating an Informatics Center and Accountability / Feedback

In this module, we will

• (1) Review the development of the Community Care Informatics Center.
• (2) Provide an overview of user applications hosted within the Informatics Center, including: the Case Management Information System (CMIS): Pharmacy Home, Reports Site; Chart Review System; and Provider Portal.
• (3) Discuss the functions served by our informatics capacity, including: quality measurement and feedback; population needs assessment; risk stratification and identification of patients for targeted initiatives; analysis of hospital and emergency department utilization patterns; clinical decision support; care team communication, and program evaluation and accountability.
• (4) Discuss staffing, technical environment, privacy, and security.

Informatics is a term that broadly encompasses all aspects of electronic data use within the Community Care program, including the collection, processing and representation of healthcare data from multiple sources for multiple stakeholders. From the very beginning, Community Care has recognized the importance of data to guide program priorities; to identify the needs of the population; to monitor costs, utilization, and quality of care; to drive quality improvement efforts; and to evaluate the effectiveness of interventions. As the Community Care program has expanded in scope and evolved over time, so have our systems for collecting, managing, processing, and reporting data. Diligent interpretation of data and efficient exchange of information has proven to be critical to the program’s success.

Information systems in the early years of Community Care

Medicaid Claims Data Warehouse
Our principal source of health information about the enrolled population has always been Medicaid claims data. The North Carolina Division of Medical Assistance (DMA, the state Medicaid agency) allowed Community Care-employed analysts to directly access the Medicaid data warehouse. DRIVE, the Medicaid data warehouse, contains six years of claims data as well as provider, eligibility, reference file, prior authorization and drug data. DRIVE is updated weekly from the Medicaid Management Information System (MMIS), which processes claims for Medicaid. Community Care analysts were able to query Medicaid claims to ascertain disease prevalence and utilization patterns, to calculate quality measures and to identify patients appropriate for specific initiatives.

Community Care provided networks with quarterly or semi-annually “Case ID” reports in Microsoft Excel® or Access® formats that would list patients meeting specified criteria (e.g., patients with diabetes or asthma diagnoses to whom the networks could outreach for their disease management initiatives; or patients with frequent emergency department visits for whom the networks could provide education or assistance about using the primary care medical home). DMA provided quarterly per-member-per-month (PMPM) cost and utilization statistics for Carolina Access practices, which Community Care analysts were able to augment with peer group comparison data and quality metrics related to the asthma and diabetes initiatives. Network leadership could use the resultant Practice Profile reports to identify cost or utilization patterns over time and address any concerns directly with practices (Figure 1).

Figure 1: Example of Early Reporting – Community Care Practice Profile

Case Management Information System (CMIS)
Begun in 2001 as a Microsoft Access database in a single Community Care network, CMIS was designed for care managers to make administrative and therapeutic notes. It evolved into a Web-based portal accessible to all networks, allowing care managers to maintain a health record and single care plan that stays with the patient as he or she moves from one area of the state to another. Community Care contracted with an external developer to establish a monthly import of Medicaid enrollment and claims data to populate the CMIS patient record with demographic and primary care provider information as well as a view of the individual’s hospital, emergency department and pharmacy claims.

In addition, networks were able to utilize CMIS to manage enrollment, eligibility and care management services for HealthNet projects (Community Care networks that were also managing the care of the uninsured residents) that in 2010 were serving 12,500 enrolled individuals. Thus CMIS enabled a continuity of care record for patients as they migrated “in and out” of Medicaid, Health Choice (North Carolina’s State Children’s Health Insurance Program) and a lack of any insurance. CMIS provided a standardized framework for care manager workflow management and documentation, incorporating tools for patient assessment, goal setting and health coaching.

2008-2010: Informatics Center Development

By 2007, Community Care had grown significantly, serving more than 900,000 patients across the state. Recognizing the cost savings and quality improvement achieved by the Community Care model for Medicaid families and children, the North Carolina Legislature mandated an expansion in Community Care’s scope to serve the elderly and disabled Medicaid populations. Community Care networks, at that point, had engaged the participation of primary care providers in all 100 North Carolina counties, including increasing participation by family medicine and general internal medicine practices.

In addition, Community Care sought a CMS demonstration program waiver to manage the care of Medicaid-Medicare dually eligible and Medicare-only populations. While our existing information systems were helpful, they were insufficient to support this expanded mission. Coordination between the externally maintained DRIVE system and Community Care’s CMIS was cumbersome and difficult to manage as the program grew. Recognizing that primary care physicians (PCPs) and networks needed more timely, reliable data to assist them in their work, we convened a panel of care managers, clinicians and technical experts to help us define and meet our program’s emerging information needs.

Several themes emerged from these discussions, which provided the overarching vision for Informatics Center development.

• In contrast to the pediatric Medicaid population, the elderly and disabled Medicaid population has a high prevalence of chronic conditions with multiple comorbidities, multiple medications and multiple providers (Figure 2). In the face of limited network resources, the medical complexity of this population would necessitate a more sophisticated approach to population needs assessment, risk stratification, quality measurement and identification of patients who could benefit from care management interventions.

Figure 2: Chronic Care Population

• Reliance on Medicaid claims data alone has inherent limitations, necessitating the incorporation of additional data sources.

• Delays in claims processing time meant that service utilization information was not visible to the care managers and providers until weeks or months after an event. The older the data, the less likely it is to be useful to care managers and providers.
• Claims data reflects only the coding used for billing purposes. Many clinically important services and diagnoses are “under-coded” in administrative data (examples include depression, hyperlipidemia, obesity and chronic kidney disease). Other diagnosis codes may appear inaccurately (e.g., asthma often is over-coded during emergency department visits, and heart failure is over-coded on procedure claims conducted for diagnostic purposes). Claims data may not reliably indicate that a service was received (such as blood pressure screening or depression screening), or the data may indicate that a service was received but not provide the more clinically relevant information (e.g., we can see that an A1C test was performed for a patient with diabetes, but we can’t see the result of that test).
• Medicaid claims data are particularly incomplete for dually eligibles. Pharmacy claims data were no longer available in the Medicaid system after the implementation of Medicare Part D in 2006. Claims from laboratories and other providers do not appear if submitted to Medicare only; and inpatient stays within 60 days of a prior discharge typically do not cross over into the North Carolina Medicaid claims warehouse once the Medicare deductible has been met.

• Aggregated reports of cost, utilization and quality data (like that presented in the Practice Profile example) have limited utility without drilldown capability to the individual patient level. Network managers and clinical directors wanted the ability to manipulate data and customize reports, to better address local concerns, support local initiatives, engage the interest of providers and assist both providers and care managers with “actionable” information about specific patient needs.
• Rapid growth in the number of participating patients and providers made the management of information, and the manual process of creating and distributing reports, very cumbersome.
• Information available to network personnel within the CMIS system could be leveraged to greater benefit if more easily accessible to other members of the patient care team, particularly the primary care provider and medical home staff, hospital-based providers and mental health providers.

The conceptual framework that emerged from these discussions was the need for a comprehensive, integrated database of healthcare information from multiple data sources, capable of feeding a variety of end-user applications to efficiently meet a variety of informational needs for the patient care team (Figure 3). With the overarching objective of getting the right information into the right hands at the right time, the Informatics Center development began.

Figure 3: Informatics Center Conceptual Diagram

Informatics Center Functions and Front-End Applications

Pharmacy Home
The Pharmacy Home Project was created to support Community Care’s pharmacy management initiatives, and address the need for aggregating information on drug use and translating it for the network pharmacist, case manager and primary care provider in a manner best suiting their care delivery needs. To accomplish this charge, we initially set up a monthly extract of pharmacy claims history from the DRIVE system to be warehoused within our own Community Care environment. Extraction, Transformation, and Load (ETL) processes were written to prepare data and load it into the application database from which prepared data could be readily pulled into front-end views of patient prescription history or user-generated population-level reports.

In addition, ETL processes applied logic to create derived variables indicating adherence calculations, gaps in therapy (days elapsed since the most recently dispensed pill supply would have expired) and other clinical care alerts (e.g. indicator of beta agonist overuse, which may indicate poor asthma control). From the application database, the system was set up to provide a patient-level profile and medication history for point-of-care activities, as well as a population-based reports system to identify patients that may benefit from pharmaceutical care support via pharmacists, case managers and PCPs in the medical home (Figure 4). The Pharmacy Home drug use information database is used both prospectively (for identification of care gaps and problem alerts, targeting of at-risk patients, and development of the pharmaceutical care plan); and retrospectively (for continuous quality improvement and program evaluation) (Figure 5).

Figure 4: Pharmacy Home – Example of Prospective Population Management Report

Figure 5: Pharmacy Home – Example of Retrospective Project Evaluation

• Quality Measurement and Feedback Chart Review System
  Chart audit, quality measurement and performance feedback was an integral component of Community Care’s early quality improvement initiatives related to asthma and diabetes. Despite rapid growth in member and provider enrollment, Community Care clinical leaders have remained committed to the monitoring of quality at the individual practice level, to engage providers in the quality improvement process and to monitor progress at the practice, county, network, and statewide level. As the program expanded to serve a larger population with multiple complex comorbidites, a broader array of quality measures was adopted (based on evidence-based care guidelines for diabetes, asthma, hypertension, cardiovascular disease, and heart failure). Community Care now conducts over 26,000 medical record reviews in over 1,250 primary care practices statewide on an annual basis.
  
  To manage the expanding scope of the chart review process, we moved from a paper chart abstraction tool to a fully electronic, streamlined system in 2009. Medicaid claims data is used to generate a random sample of eligible patients, and to pre-populate audit tool elements according to an individual’s identified chronic conditions (Figure 6).
  
  Secure client-server software allows independent auditors to work offline when Internet access is not available in the clinic location. When access to Internet is available, the system automatically synchronizes data with the server. Data is fully encrypted offline and in transit. Data sent to the server automatically updates a variety of process, progress, and analysis web-based reports. Practices and CCNC networks then have immediate access to chart review results through a secure web-based report site, with patient-level information as well as practice, county, network, and statewide results with national comparative benchmarks (Figures 7 and 8).

Figure 6: QMAF Chart Review – Data Abstraction Tool

Figure 7: QMAF Chart Review – Patient Report

Figure 8: QMAF Chart Review: Practice Report with Benchmarks

• Informatics Center Reports Site 
  The IC Reports Site was created to allow the efficient and secure distribution of reports through a secured web-based report access and management application, with report access permissions determined by the appropriate scope of access of individual users. Network-level administrators authorize their own employees and providers by customizing their scope of access by practice or region. A report built at the statewide level can be readily distributed according to the permission tree structure, such that only the appropriate patient information is visible to each end user.
  
  Initially, most reports distributed through the Reports Site were created by data analyst staff querying the DRIVE or IC data warehouse with SAS to create an underlying data table for the specific report. We could then create a Web-based report in RDL (report definition language – Microsoft standard format for web-based reports) format, pointing to the data in that underlying data table. Publishing the RDL reports through the Report Publisher (a custom-made Windows-based application) would then automate the process of separating the data and publishing report instances customized to networks, regions, and practices according to the permission tree. Over time, we are translating report code into SQL, and establishing ETL processes to allow extracting data directly from the IC data warehouse and loading it into the reports underlying data tables. All reports are printable and can be exported into .pdf or Excel format.
  
  Sample reports are provided here, to illustrate various functions served by our analytics and reporting capacity:
  
  
  • Population Needs Assessment: Identification of demographic, cost, utilization, and disease prevalence patterns by service area. Figure 9 shows a “Chronic Care Summary Statistics” report. This database is updated quarterly to reflect the current aged, blind, and disabled enrolled population, and contains 79 data elements. Network leaders can readily obtain information about the demographic characteristics, prevalence of chronic medical and mental health conditions, spending by category of service, and rates of hospital, ED and other service use within their county-level service areas. This aids in program planning and resource allocation; identification of outlier patterns (such as unusually high rates of personal care services); and tracking of local utilization patterns over time.

Figure 9: Chronic Care Summary Statistics

• Risk Stratification, Identification of High-Opportunity Patients, Patient-level Information. The size and complexity of the Medicaid population, in terms of physical health, mental health, and socioeconomic needs, necessitates intelligent mechanisms for identifying patients most appropriate for care management interventions, particularly in the face of limited resources. The use of historical claims data to screen patients for care management intervention can greatly improve the efficiency the care team. Figure 10 gives an example of our Chronic Care Patient Database, which updates quarterly with cost, utilization, and diagnosis data (over 70 data elements) for all aged, blind, and disabled enrollees.

Figure 10: Chronic Care Patient List



Within that database, we are able to flag patients who meet specified criteria for further screening by a care manager, according to patterns of service use over the prior 12 months (such as multiple ED and inpatient visits, multiple medications, lack of PCP contact, target medical conditions, and high cost). We flag approximately 17 percent of the population for screening, and approximately five percent as highest priority for screening. In addition, we flag patients who meet criteria for mental health case management based on mental health service use, in order to best leverage external case management resources and coordinate care. A comprehensive patient-level view of this information is available in a searchable Chronic Care patient snapshot database (Figure 11), which facilitates triage when referrals are made for care management by providers or at the time of hospital discharge. Similar reports are generated for specific initiatives or pilot programs (e.g., identification of patients with newly diagnosed asthma, heart failure, and diabetes; identification of patients receiving controlled substance prescriptions from multiple sources; identification of patients with poor adherence to their blood pressure medications for a telephonic health coaching intervention).

Figure 11: Chronic Care Patient Snapshot

• Monitoring of ED and Inpatient Visits. Figure 12 is an example of a detailed utilization report that can be easily navigated by local care managers and clinicians who may not be technically savvy. The user can readily access this database of ED visits by their enrolled population, which updates weekly with every claims payment cycle. The report can be parameterized by hospital, PCP, or patient or visit characteristics; and can tally visit counts by patient or practice. A similar report is available for inpatient hospitalizations. These reports are very flexible for answering a variety of questions (e.g.: Are patients from my clinic having a high number of non-emergent ED visits during regular office hours?  How many heart failure discharges were readmitted within 30 days, and did they bounce back to the same facility or a different location?); and for identifying at-risk patients in a timely fashion (e.g., “Here is a list of all patients with an asthma-related ED visit, let’s make sure they have a follow-up PCP visit scheduled.”).

Figure 12: ED Visit Report

• Tracking of Care Quality Indicators. In addition to the quality measures tracked in the annual chart review process, we are able to track a number of quality measures using claims data alone, with quarterly updates. Measures can be aggregated to the practice, county, network, or statewide level. Results are distributed in both spreadsheet format for an easy comparative view across practices (Figure 13), and a comprehensive practice-level report with trend information (Figure 14). Reports include measures related to diabetes, asthma, heart failure, cardiovascular disease, pediatric well visits and dental care, and adult breast, cervical, and colorectal cancer screening.
  
  Figure 13: QMAF Claims Measures: Spreadsheet View

Figure 14: QMAF Claims Measures: Practice Report

• Program Evaluation and Tracking of Key Performance Indicators. Figure 15 is an example of program performance tracking for monthly reporting to the state Medicaid agency and state legislature. Tracking of key metrics provides stakeholders with assurance that efforts are aligned toward the overarching goals of cost savings and quality improvement, and that all networks are held accountable for the overall performance of the program. Key indicators include both process measures (such as percent of targeted hospitalized patients receiving medication reconciliation) and outcome measures (such as readmission rates).

Figure 15: “Scorecard” Report: Monthly Tracking of Key Program Performance Measure

• Provider Portal
  
  The Informatics Center Provider Portal was released in August of 2010. This portal was built with the treating provider in mind, offering elements of CMIS, Pharmacy Home, and the Reports Site, tailored to the target user. Through a secure web portal, treating providers in the primary care medical home, hospital, emergency room, or mental health system can access a Medicaid patient health record which includes patient information, care team contact information, visit history, pharmacy claims history, and clinical care alerts. Importantly, the use of Medicaid claims data provides key information typically unavailable within the provider chart or electronic health record. For example, providers are able to see encounter information (hospitalizations, ED visits, primary care and specialist visits, laboratory and imaging) that occurred outside of their local clinic or health system. Contact information for the patient’s case manager, pharmacy, mental health therapy provider, durable medical equipment supplier, home health or personal care service provider is readily available. Providers can discern whether prior prescriptions were ever filled, and what medications have been prescribed for the patient by others. (Figure 16)   Built-in clinical alerts appear if the claims history indicates patient may be overdue for recommended care (e.g. diabetes eye exam, mammography).

Figure 16: Provider Portal Patient Record

The Provider Portal also contains key resources for assisting providers in the management of Medicaid patients, such as a compendium of low-literacy patient education materials, and practice tools for risk assessment and disease management. Through a seamless link into a licensed service maintained by an outside partner, providers can retrieve medication information for patients in multiple languages, in video or print format. Medical home providers may directly access population management reports and quality metrics for their own patient population through a seamless link into the Informatics Center Reports Site. (Figure 17)

Figure 17: Practice Reports Available Through Provider Portal

• Care Management Information System (CMIS) Enhancements
  In 2009, we transitioned the CMIS system to be maintained within the NCCCN, Inc. Informatics Center environment and to be sourced from the IC data warehouse. This allowed for greater developmental flexibility and the opportunity to exchange information across IC applications (e.g., care management data fields may be visible through the Provider Portal or available for reporting in the Reports Site, while chart audit reports may be retrieved within the CMIS patient record). As the Community Care Chronic Care program evolved, care management tools could readily be incorporated into the CMIS system. Important enhancements included a comprehensive health assessment and functional assessment tools as well as disease-specific screening and monitoring modules (Figures 18 and 19). Bulk task capacity was added to allow for population-level interventions (e.g., to send a flu shot reminder to all patients with diabetes). A secure messaging feature allowing 20 MB attachments was added to allow care managers to communicate patient health information securely to PCPs or others involved in the patient’s care outside of the CMIS system. Report-designing capacity was built within CMIS to allow managers to more closely monitor the caseload and activities of the care management workforce.
  

Figure 19: CMIS Heart Failure Assessment Tool

Buy it or build it?

Throughout the evolution of the Informatics Center, we have repeatedly encountered the question of “buy it” vs. “build it.”  Over the past few years, the Health IT market has exploded with vendors and products offering software solutions for healthcare management entities, particularly for predictive modeling/risk stratification, quality measurement and reporting, and clinical decision support. Our experience over the years in developing CMIS and our practice- and community-based infrastructure for quality improvement had taught us that a technical product is not a “solution” in and of itself, but rather the solution is what people do with the information. Thus the end users—care managers, network administrators and practicing providers—have been integrally involved in the development of all IC applications, assuring that whatever we built would be exactly what they were asking for.

In general, we have chosen not to incorporate vendor products with proprietary or “hidden” logic into our Informatics System. We also felt it important to develop and maintain our own user interfaces for greatest flexibility in design over time, thus staying away from vendor products that relied on an off-the-shelf or standalone user interface. We have identified a number of opportunities for enhancing our systems with third-party tools and development partners, however. Examples include the use of a clinical decision support rules engine to maintain a weekly-updated patient-level clinical alert database to identify patients needing specific services (that data is then available for display within our reports, CMIS and Provider Portal applications). We will also be incorporating third-party software to help with risk assessment for care management interventions and to allow for risk adjustment of our performance measures.

As we continue to build the Informatics Center, we are most interested in the seamless incorporation of other data sources into our environment. On the near horizon, this will include Surescripts pharmacy data for dual eligibles, lab values from a large independent laboratory provider, LabCorp, Medicare claims for our 646 population from CMS, and real-time hospital admission, discharge and transfer (ADT) data from 48 North Carolina hospitals through a cooperative initiative with the state hospital association. Over time, we will be exploring opportunities to incorporate a shadow claims system for uninsured patients in the HealthNet programs, administrative data from private health plans and direct exchange of information with provider electronic health record resources.

Technical environment

Warehouse build and the evolving data model
The Informatics Center began to build its own data warehouse of Medicaid claims, eligibility and provider data in May 2009. When this effort began, Pharmacy Home and many reports being shared with the networks were created by data analysts manually querying DRIVE, the Medicaid data warehouse. The results were exported to Excel or used to load claims to display in Pharmacy Home. CMIS, which migrated to the IC from an outside vendor in July 2009, was manually loaded from monthly claims and enrollment extracts. Beginning in May 2009, NCCCN obtained from the Medicaid data warehouse and back-loaded three years of Medicaid claims. This initial load contained approximately 500,000 million rows and 250 columns of data. As the warehouse development continued, NCCCN arranged to receive weekly extract claims delivered via FTP to the IC. However, the 20 additional files of eligibility, provider and reference data continued to be manually pulled each week by a data analyst from the IC using SAS queries.

In early 2010 the manual extraction was replaced by an automated FTP of files created from the weekly load files of the Medicaid data warehouse. This automation of files from the Medicaid warehouse to the NCCCN data warehouse enabled the IC to automate its own processes. The NCCCN warehouse now sources weekly updates for CMIS, Pharmacy Home and the databases, refreshing utilization reports. The evolution of processes of time was the only option to continue to meet the existing business needs while developing new structures.

As the process to load the warehouse has evolved, so too has the NCCCN warehouse data model. The Medicaid data warehouse uses Sybase ASIQ relational database technology. This product performs well with a model containing a large fact table. The fact table holds six years of Medicaid claims. NCCCN initially mimicked the Medicaid data model. This was a familiar structure for the data analysts and allowed NCCCN to move forward with development while the warehouse was being created. In choosing Microsoft SQL Server, NCCCN leveraged the databases already residing in the applications and staff expertise with the product. But Microsoft requires a much narrower data model for best performance. The IC database administrator continues to remodel the IC claims tables to reduce the number of columns of data and maximize performance. Tuning continues as does the creation of data marts to minimize the queries against the main warehouse.

Web-based technology
The Informatics Center continues the integration of four separately built web-based applications. These applications, the Case Management Information System (CMIS), Pharmacy Home, a chart audit tool and the IC web reports sites were built with Microsoft technologies. These products include SSRS Report Builder for the creation web-based reports, SSIS or SQL Server Integration Services for extraction, transformation and load (ETL) and Microsoft SQL Server as the relational database. The use of common tools by staff with like backgrounds and experience enabled faster integration and allowed NCCCN to leverage the work already accomplished in the creation of these applications. As an example, in December 2009 NCCCN consolidated the active directory authentication of users for three of the four applications to achieve a single sign on. In another recent integration, the IC now displays to care managers in their CMIS application patient profiles that were built for and reside in the IC report site. As the integration effort continues, staff have migrated all underlying databases to the same more recent version of Microsoft SQL Server. Microsoft’s dominance in the market makes it a choice for many of the venders working with Community Care. Fortunately, several of Community Care’s development partners whose products will be used or displayed in the IC web portal also rely on Microsoft products. As with the internal integration, the shared technology enhanced the efforts as the IC integrates new third-party applications.

Data Security
The IC locates its servers at Hosted Solutions, a SAS 70 Type II certified commercial data center in Raleigh, N.C. Hosted Solutions offers a facility that meets National Institute of Standards and Technology physical plant security standards with 24-hour security, redundant power and Internet connections, complex firewalls and encrypted telecommunication lines. The IC hardware is located in locked set of racks in the facility, and services are scalable allowing upgrades in bandwidth as traffic in the IC web sites increases. Storage is handled by a disk array configured with redundant storage (raid 10). All servers are backed up regularly as are all applications. Using Hosted Solutions eliminates the need to upgrade leased office space to meet the power, temperature and security requirements of a data center and reduces the number of staff required for the computing infrastructure. For a small newly created organization such as NCCCN, use of a commercial data center represents a lower investment of resources and greater security and reliability.

Informatics Center applications are built using Microsoft .net technology, and they are available to authorized users via the Internet using a secure socket layer. User access to applications and reports is handled in several ways to insure multi-level security. With Windows Active Directory, the IC manages a complex permission tree to restrict user access to the applications or reports they are entitled to see. All sites maintain audit logs of the users’ access to the application, the date reviewed or comments modified.

A final note: privacy, security and legal arrangements

When IC development was first contemplated, we did not fully anticipate the time and financial commitment necessary to address the many legal concerns related to data sharing, data protection and compliance with state and federal laws. All IC operations must assure adherence to provisions under federal Health Insurance Portability and Accountability Act of 1996 (HIPAA) and the Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 (which specifies the purposes directly connected to the administration of the Medicaid or the CHIP programs), and federal regulations related to the disclosure of substance abuse information (found at 42 C.F.R. § 431.302). North Carolina also has state laws governing the confidentiality of mental health, developmental disabilities and substance abuse information as well as the confidentiality of information related to HIV and other communicable diseases.

All of this required the involvement of legal counsel from multiple stakeholder groups and thorough analysis of statutory authority and contractual relationships among the Medicaid agency, the N.C. Department of Health and Human Services, NCCCN, Community Care networks, providers, local mental health management agencies and local health departments. In many instances, the legal framework for health information exchange had not yet “caught up” with the technical opportunities. As health information exchange efforts gain steam nationwide, it is likely that the barriers and misperceptions that we encountered will diminish over time.

Table of Figures
Figure 1. Example of early reporting: Community Care Practice Profile
Figure 2. Chronic Care Population
Figure 3. Informatics Center Conceptual Diagram
Figure 4: Pharmacy Home: Example of Prospective Population Management Report
Figure 5: Pharmacy Home: Example of Retrospective Project Evaluation
Figure 6: QMAF Chart Review: Data Abstraction Tool
Figure 7: QMAF Chart Review: Patient Report
Figure 8: QMAF Chart Review: Practice Report with Benchmarks
Figure 9:  Chronic Care Summary Statistics
Figure 10: Chronic Care Patient List
Figure 11: Chronic Care Patient Snapshot
Figure 12: ED Visit Report
Figure 13: QMAF Claims Measures: Spreadsheet view
Figure 14: QMAF Claims Measures: Practice Report
Figure 15: “Scorecard” Report:  Monthly tracking of key program performance measure
Figure 16: Provider Portal Patient Record
Figure 17: Practice Reports available through Provider Portal
Figure 18:  CMIS User Interface
Figure 19:  CMIS Heart Failure Assessment