Healthcare delivery has always relied upon data, from Hippocrates’ time when the hand alone was used to gauge the temperature of the human body (hot or cold) to the ultra-precise digital outputs of countless medical devices today. Enormous, ever-increasing amounts and types of data, as well as its expanding and evolving methods of (and need for) capture, dissemination and decision support make imperative the governance of that data–starting in our own backyards. Whether payer or provider, patient or public policy-maker– insistence on consistent high-quality data by means of robust policies, procedures and tools will build upon and continually enhance industry standards, measurement of healthcare outcomes and member experiences, in support of the growing predilection toward pay for performance in healthcare.
Don’t Get Seduced by Big Data
While summary narratives and trends based on aggregated data are important (and increasingly popular and depended upon), don’t forget to persistently focus on the fundamentals. Although these efforts can be tedious (and “thankless”), they pay off by providing more complete and up-to-date encounter reporting–helping alleviate today’s frustrations that include— data that is missing or inaccurate, not granular nor timely enough to be useful–and often episodic and unlinked–all potentially impeding that underlying data’s raison d’être. The consumer can also play a role, often knowing their health better than it is reflected in a “system”–another reason to not get too far removed from the fundamental importance of each member’s data. Having a grasp of population health is great; individual health, even better.
Despite remarkable advances in medical equipment, diagnostic capabilities, drugs, treatments, etc., IT is just now catching up with Electronic Health Record (EHR) systems– e.g., the ‘informational’ aspects of medical technologies. With 2 million members, L.A. Care is the nation’s largest publicly operated health plan, having helped more than 3,000 primary care providers reach meaningful use with EHRs. However, EHRs are only as good as the data they contain, which historically further lags behind in quality, the result of decades of disparate practices and non-integrated systems, and, arguably, the general apathy of the consumer, as third parties were responsible for sorting out the details. Rapid changes in attitude, knowledge/awareness of medical topics and issues, regulations, etc. and much deeper engagement in the direct management and scrutiny of their own healthcare by ever-larger numbers of people in this country substantially and continually increase attention to how and where health information is defined, collected, stored, shared, and used.
Consider the banking industry and your credit card statement, for example. There exists zero tolerance for errors or excessive delays in transactions posted to your account. Real-time, 100 percent accuracy is assumed for financial transactions, no matter how trivial or extensive. A credit card issuer expects to address discrepancies on an immediate basis with its consumers upon their request, and instantaneously correct errors, even proactively notify its customers of unusual (suspicious) activity. By contrast, just obtaining one’s medical records can take weeks, often at one’s own expense, with the further anticipation of those records being overly-complicated, incomplete, error-prone, indecipherable, out of date, etc.–e.g., not “user friendly. A patient is also left with little or very protracted (at best) eventual recourse with (or even discussion/explanation from) providers, much less potential dispute resolution or ultimate satisfaction. At the core of striving to improve healthcare information, industry efficiency, cost-effectiveness, member outcomes and overall satisfaction is the accuracy and timeliness of patient data. Of course accessibility–sharing, interoperability, confidentiality and regulatory issues are essential parts of ultimate IT solutions; however, let’s continue our focus on improving data quality through effective data governance.
Principles of a Data Governance Program
Data Definition–a shared, common set of definitions for all data elements–what each element means, where it is created, how it can be changed, and how it should be used–so it can be consistently understood, managed, processed, and secured.
Data Source– an extension to data definitions to identify the systems and databases that contain a data element, including the best source based on the purpose for accessing the data. There should be one system of record for any data element, with the number of systems allowed to modify data elements kept to a minimum, while ensuring the data is available as required in the timeframe necessary to cost-efficiently support the business.
Data Quality–extension to data definitions to define validation rules for each critical data element for users and developers alike.
Data Security and Compliance–the final extension to data definitions is clear identification of the rules around granting access to data elements and ensuring the rules are properly applied and enforced, providing clear instruction for provisioning and requesting access to data elements in accordance with requirements established by all regulatory and governing entities.
Successfully putting each of these principles to use involves:
•Metrics (e.g., quantitative and qualitative adoption/ achievement & exception measures, etc.)
•Oversight (roles & responsibilities for data guardians/ stewards, collaborated and coordinated throughout the enterprise as appropriate)
•Processes (e.g., issue logging, tracking, prioritization & resolution)
•Tools (Data Dictionary, architecture diagrams, meta-data management, policies & procedures, etc.)
Data Governance Maturity Model
Further compelling the development of their data governance models are organizations’ evolution from silos of information to true enterprise views, shifting efforts from predominantly those of capturing, reviewing and cleansing data to productively analyzing it, once high-quality data is routinized.
Effective data governance is business-driven, not just an IT endeavor. As such, C-level sponsorship of and commitment to these goals is vital for enterprise success.
Benefits resulting from widespread data quality improvements include significantly better:
•Measurement of healthcare outcomes, both for individuals and populations (e.g., more accurate and meaningful Big Data built from the bottom-up)
•Reliability of enterprise data analytics used internally and reported externally
•Platform for evaluating member experience
•Not only measurement of, but potentially positively impacting healthcare outcomes
Under the leadership of the CIO, with an informed, and participative organization, good data governance directly impacts the ultimate objective of our industry–better healthcare outcomes.