With providers challenged to manage information in just one CCD, how can they hope to use these documents to effectively influence care at the population level? The key is to look for ways to use technology to target specific patient information, pinpoint new and relevant information and alert both patients and providers when updated information is available.

Areas of opportunity

For accountable care organizations and others that rely on data analytics and population health management as core components of their healthcare approach, developing an effective and reliable method for exchanging data across the continuum of care is crucial.

As an initial attempt to standardize health information exchange (HIE), CCDs provide some valuable lessons. Healthcare organizations should take into account the following areas for improvement as they search for technology that more efficiently extracts and manages discrete changes in patient data, including:

• Structuring data. Different versions of the CCD — such as CCDR2 and CCDR3 — are used by various electronic health record (EHR) vendors and data warehouses. Some CCDs are compiled in running-text format and include non-discrete, unstructured data — information captured as images, for example. Retrieving specific patient information from non-discrete and unstructured data is difficult, if not impossible. It is particularly important to overcome this challenge for population health management, which requires structured and aggregated data to identify gaps in care for certain risk striations.
• Mapping data. Data terminology varies from organization to organization. For example, one hospital might use “myocardial infarction” and another “heart attack” to identify the same diagnosis. All of this means that when one vendor or healthcare organization acquires data from another, precise apples-to-apples analytics will only be possible if the data is correctly mapped. Without mapping, organizations lose the ability to effectively aggregate data. Better and more consistent semantic standardization and normalization is required to achieve the ability to productively analyze aggregated data.
• Sharing pertinent data. CCDs are comprehensive documents in a compiled data format. In other words, the longer a patient is treated by providers, the longer the document. Typically when CCDs are shared between providers, every component of the record is sent — whether it is useful for the current need or not. A more efficient process would focus on sending only updated or applicable information.

Data for population health management must be collected from disparate sources and then analyzed to determine gaps in care. As illustrated above, this requires discrete, structured and semantically standardized data that can be collected at the transactional level.

Portals offer discrete, relevant data  

Some patient portals offer a way to resolve the population health management concerns not completely addressed by CCDs. Portals that are driven by a data repository platform, for example, are ideal for building the foundation for population health management. This technology gathers all components of the patient record — such as diagnosis, procedure, medications, allergies and history information — in structured formats like HL7 and CCDR3. In other words, these portals have the ability to act as a single data acquisition point.

In addition to being an effective data repository, dynamic portals can also take advantage of discrete data collection and aggregation to monitor events and grab information as it is updated within a patient’s record. This means the most relevant information can be captured and delivered to those providers who need it.

Consider this example: During a patient visit, a family physician notes in the family history section of the chart that the patient’s father just suffered a heart attack. On a transactional basis, this discrete piece of information can be pushed to the patient’s cardiologist through a portal that is capable of collecting, standardizing, normalizing and storing that information.

Although data acquisition is not a glamorous task, it is the essential underpinning for the data analytics necessary for effective population management. With patient portals that take a holistic and comprehensive approach to data compilation, healthcare organizations can access data in a way that allows efficient monitoring and management of patient events.  

The principles behind CCDs are well-intentioned, but the existing format of these documents makes them ineffective for data analytics and population health management. By tapping into patient portals that sit atop a data repository platform, organizations and vendors alike have one consistent access point for acquiring and analyzing meaningful data.

Gary Hamilton is president and founder of InteliChart.