Partner Dante Diamente and Engagement Manager Elizabeth Southerlan, both members of Oliver Wyman's Strategic Information Technology & Operations team, explore how healthcare institutions can solve the “data problem” and reap the benefits of good data. They explain the risk of overlooking data planning – separate and apart from IT strategy – and spell out the six essential steps to getting to good data:
Most healthcare organizations today are paying dearly for data that is, on most days, just barely good enough. Commonly, brute force or heroic efforts are required to produce quality and compliance reports, customer reports, or even revenue reports. This inefficient process is fatiguing and distracting organizations right when they need to be upping their data game to meet value-based shifts and achieve the triple aim.
The root of the challenge lies in the fact that most healthcare companies – payers and providers alike – do not have a good handle on their data quality. Both providers and payers struggle with complex and fragmented legacy environments. Each has their own lines of business that are operating in silos, each maintaining their own datasets with limited use of shared language.
In addition, providers and payers are trapped beneath inconsistent formats and data definitions, even within the walls of one organization. This leads to burdensome reconciliation efforts. Recent consolidation of players – via joint ventures, strategic alliances, and full blown M&A activity – has exacerbated the situation, as has the introduction of disruptive health tech companies.
Six Components of Effective Data Strategy
There is no one approach to achieving good data. However, there is value in approaching the challenge from the “top of the house.” This entails defining strategic objectives for data and data quality, developing technology to enable those objectives, and aligning the organization to a single, overarching data strategy. Regardless of an organization’s progress on its journey to good data, here are six components all can factor in:
1) Align data strategy to business needs
Addressing all the needs of an organization from a data perspective can be an overwhelming, potentially never-ending journey. Define the questions that drive your business and articulate where data quality is most valuable.
2) Define what “good data” looks like
The quality of data is highly contextual and depends on what it is used for. Standards for data need to be defined in order to satisfy all key users – and the standards adopted need to conform to the ones required by the most demanding users. Few authoritative standards (e.g., HL7, SNOMED, ICD-10) or frameworks are available – even fewer that span the full ecosystem of payers, providers, and vendor partners. The industry is working hard to mitigate this (e.g., CAQH’s CORE) and, in the meantime, organizations can internally define standards for data that most directly support the business strategy.
3) Establish shared data sources across key users
While it is extremely difficult to create a single shared data source for large, complex organizations, healthcare institutions should aim to designate common authoritative sources for users with a high degree of overlap (e.g., risk and finance). Key users should support these sources, as these users likely know what data best underlies the business strategy. This allows data to be captured, enriched, and aggregated within one authoritative source, and have all users pull from that same source to generate reports and analysis.
4) Establish an authorized path for end-to-end data flow
As institutions grow and new systems are implemented, optimizing data flow may not always be top of mind. But data-lineage analysis is critical to understanding the various sources, transformations, and arrival points across key data elements. Institutions should aim to establish an authorized path for data flow, defining authoritative data sources, homogenizing transformations, and placing data quality controls appropriately.
5) Embed data quality controls at critical data capture and transformation points
Data capture often can be manual and subject to error. Placing more stringent controls upstream to prevent data issues propagating downstream is not just critical, it is now a matter of security. Smart controls for completeness, accuracy, and conformity often can be automated, allowing issues to be addressed directly at the data capture point.
6) Define clear responsibilities for all data ecosystem players
Data is a shared asset and flows across business lines, functions and ecosystem players; thus, assigning ownership is often challenging. We advocate establishing data owners who are heavy users of the data and can attest to its quality at various stages. At the same time, data owners will not solve all data problems; data management-related activities need to be embedded in the job description for all data ecosystem players.
Good data is no longer “nice to have.” It is required for regulatory compliance, accurate reimbursement, and providing a consumer-centric experience. Few healthcare institutions are satisfied with their progress on data quality, and even those that have made significant strides recognize that it is a journey that never ends. By taking a holistic and pragmatic approach to achieving good data, and adhering to the six elements of data strategy, an organization can align and pivot data-quality efforts to link explicitly to business-value creation.