As healthcare organizations begin to apply analytics to care delivery, data has become a hospital's most valuable asset — and one of the most challenging to manage.
This content is sponsored by RelayHealth
Due to advances in digital and cloud technology, more patient health information exists today than ever before. Digital health data has the potential to transform care delivery by helping physicians make evidence-based decisions. With access to health information from different data sources, clinicians gain better insight into patients' conditions, such as heart disease or diabetes, which are complex and costly to manage.
Although the process of capturing, acquiring, validating and storing data is crucial to gaining these insights, many healthcare organizations don't have standard procedures or processes in place to ensure data quality or security. If an organization tries to aggregate and analyze poor-quality data, it may derive useless or even wrong conclusions, which can have dangerous consequences.
To better support strategic decision making in value-based care, hospitals and health systems are establishing data asset strategies to manage the collection and transformation of raw data into actionable insights that support quality improvement efforts.
A data asset strategy outlines the governance framework and sets standard procedures for how a healthcare organization will manage data to ensure it's secure, available, reliable and actionable for an array of hospital staff, including clinicians and administrators. The strategy should span the entire data value continuum — from acquisition to delivering data to end users — and establish best practices at each step that align with strategic goals across the organization.
Without a well-developed data science strategy, healthcare organizations are more likely to struggle to leverage and protect increasing volumes of data and medical knowledge in an organized and efficient manner.
Defining the data challenge
Capturing a diverse range of health data has been a strategic priority across the healthcare industry in the last decade. Provider organizations especially have committed a substantial amount of time and money to building sophisticated health information systems and digital warehouses. They amass exhaustive amounts of patient health information every day through a complex array of internal and external sources, such as inpatient and ambulatory EHRs, electronic prescription databases, laboratory systems and payer claims databases.
Capturing data is the first step toward analytics-driven medicine. But data is only as helpful as the insights it yields, and often it requires a significant amount of rework before it can be combined and used to make clinical, operational or business decisions.
Each individual data source, from fitness trackers to EHRs, provides valuable information of patient health and behaviors, especially when these sources are combined to create a composite medical record. Giving the appropriate caregivers access to this information can positively affect care delivery by reducing the likelihood of duplicative care, discovering gaps in care and avoiding unnecessary testing.
"The more visibility you have [into a situation], the more intelligent decisions you can make," says Tina Foster, vice president of business advisor services for health information technology provider RelayHealth. This is true for both individual patient care as well as high-level business or marketing decisions.
"Executive leaders are asking themselves, should we add more neonatal intensive care unit beds? Does it make sense open an outpatient department? They're looking at information all the time to make strategic decisions," Ms. Foster says.
Integrating data from numerous sources is integral to supporting the business of care delivery in outcomes-based reimbursement. However, it also presents substantial challenges — interoperability, data integrity, usability — as well as privacy and cybersecurity risks that make an organization vulnerable to legal recourse.
"The focus in healthcare has been on automating data capture, acquiring data and getting it together in one spot, not on strategic planning and how we're going to use that data meaningfully on the backend," Ms. Foster says.
The lack of strategic clarity does not go unnoticed among health data scientists. A Stoltenberg Consulting survey found 51 percent of healthcare IT leaders believe the most significant barrier to hospital data analytics is not knowing what data to collect or how much of it, followed by a lack of organizational clarity on what to do with data and what to look for when analyzing it.
Three common data-related issues plague hospitals without data asset strategies.
1. Nonstandard data integration can comprise data quality
Data from disparate health information systems come in different formats with different vocabularies. Common health data formats include HL7, X12, CCR, CCD and CCDA, among others. Combining raw data from various sources results in a hodge-podge of digital information that isn't usable or valuable until it is translated into a standard set of definitions. Before a software algorithm can explore the data for insights, the data must be cleaned up and converted into a unified form the algorithm can understand.
Healthcare organizations have traditionally addressed the challenges of big data by relying on data scientists and IT staff to manually acquire, "clean up" and integrate data from disparate sources. The tedious nature of the job — what data scientists call "data wrangling" or "data janitor work" — requires substantial labor and cost. Data scientists reported spending up to 80 percent of their workday collecting and preparing unruly healthcare data before it can be explored by analysts for useful insights, according a survey conducted by The New York Times.
Nonstandard data processes can make data integration time-intensive. It can also lead to data quality issues, such as nonstandard data terms that make it impossible to combine data sets from different parts of the same enterprise. A data governance strategy can address the problem of data integration by establishing standard data definitions and vocabularies to be used across the enterprise. An example of this would be setting MM/DD/YYYY as the standard format for a date. It can also establish quality assurance controls and data audits throughout acquisition, collection and integration processes.
2. Low-quality data diminishes users' trust
Ensuring only high-quality data is preserved and stored is critical to supporting clinical and strategic decision making. Feeding low-quality data into analytics programs can produce misleading conclusions that negatively affect the accuracy or timeliness of an organization's or physician's decision making. When this happens, end users won't fully trust the data or the insights it offers, and they won't incorporate it into their daily workflow.
As is, fewer than 20 percent of clinicians, healthcare executives and hospital leaders believe their organization's application of data to direct patient care is "extremely effective" or "very effective," according to a 2017 survey conducted by New England Journal of Medicine Catalyst. The majority of respondents rated their organization's data effectiveness somewhere between "effective" (36 percent) or "not very effective" (32 percent), leaving a tremendous amount of room for improvement.
Healthcare organizations must address user trust issues in their data governance strategy. Thoughtful data governance helps healthcare organizations build reliable, valid databases healthcare executives and clinical staff feel comfortable using. The first step in gaining executive and clinician trust is setting data standards that ensure data quality and security across the data value continuum.
3. Data is vulnerable to the threat of breaches
Protecting the privacy and anonymity of patient health data is paramount. That task becomes more complex when an organization uses a patient's data for purposes that go beyond immediate patient care. Storing, transferring and using patient data in multiple capacities increases opportunity for corruption. An inadequate or patchwork security framework may lead to unauthorized access that undermines patients' and providers' trust, among other repercussions.
An organization's data asset strategy should encompass privacy and security safeguards. This includes the processes, policies and technologies an organization will use to protect data and information across the organization from breach, corruption and loss. Protection also ensures that information is appropriately confidential, based on its classification.
A call to action: Establishing a data asset strategy
The aforementioned challenges illustrate the need for comprehensive data asset strategies. "We have to start treating data as one of the most valuable strategic assets an organization has," Ms. Foster says. To help healthcare leaders create a foundation for their own approach, Ms. Foster recommends IT and business executives consider the following:
- Align. "First, understand the strategic imperatives across your organization. Then make sure you organize your data aggregation practices to match," Ms. Foster says. Strategic alignment supports an information-driven, decision-making culture and ensures the hospital workforce at all levels has access to the information it needs to make good decisions in real time, and it supports the expectation that information is used appropriately and strategically. Hospital trustees should work with senior leadership to ensure that the organization has a comprehensive information governance plan in place that aligns with and helps promote the overarching organization strategy, Ms. Foster says.
One way organizations can help to align information management practices is to create a data governance department. These employees can help break down management silos between IT, clinical and business departments and ensure each arm of the organization gets the data it needs to make informed, strategic decisions. - Select and Store. Once an organization knows what information and data it needs, leaders can seek out and select the best possible data source for that project, as well as the most efficient way to process and store it. Technology plays an important role in making sure data is stored securely but is still easily accessible by employees.
Some healthcare organizations see value in using technology platforms to support data asset management across the value continuum. For example, RelayHealth offers a data platform that acquires, aggregates and helps store and organize clinical, financial and operational data across various care settings. The platform also supports a variety of data formats, meaning it can deliver data to end users according to their viewing preferences, such as through a mobile device or EHR. - Enrich and Tell. "CEOs, physicians, nurses, board directors — they all have different competencies, different data needs and different preferences for how they want to look at it," Ms. Foster says. A data asset strategy should consider the most effective ways to deliver the right information to the right stakeholder at the right time. For instance, an executive may prefer viewing data on a mobile device in statistical form, whereas a clinician might prefer looking at a graph through the EHR.
Conclusion
To survive in the changing healthcare landscape, organizations need to define a plan for how to tap into the value of data across their enterprise. This includes building the skills and processes and employing the right enabling technology to transform raw data into information that drives strategic value. With proper support from executive and IT leadership, a data asset strategy helps organizations better position themselves for success under value-based medicine. "When done right, data asset management positively impacts every facet of the business," Ms. Foster said.