The following is a guest article by Mike Noshay, Chief Strategy & Marketing Officer, Board Member at Verinovum.
Within the data sphere, patient information is the most important and crucial healthcare data. It has got to be right. New legislation and technology are changing the way data is handled, and payers and providers are upping their commitment to clean, curated, quality information for patient safety and positive outcomes.
There’s been a lot of recent press around poor data quality and its tangible impact on healthcare. Even large software companies aren’t immune to data quality problems, and there have been several lawsuits brought against EMR and EHR companies for using bad data in their advanced algorithms. While artificial intelligence and predictive modeling present attractive “pie in the sky” projections of better healthcare worldwide, even the most sophisticated models and tools can’t provide value if the data they use lacks in quality. In short, even the best algorithms can result in poor outcomes if the underlying data is flawed.
When bad data happens, the wasted time and resources, lost revenues, claims denials, penalties, and unacceptable patient outcomes can be staggering. On rare occasions, patients have paid with their lives.
According to an article in the Journal of Technology Research, the cumulative financial impact of errors in healthcare records can reach up to $40 million. Moreover, determining pre-authorization and ensuring clean patient data upfront are critical in preventing claim denials. Whether the mistake is a misspelled name or a more complex coding issue, a claim could be denied. According to Black Book Market Research, an estimated 33% of all denied claims result from inaccurate patient identification or information, costing the average hospital $1.5 million in 2017 and the U.S. healthcare system over $6 billion annually.
But how do healthcare organizations and technology companies ensure the data behind their decisions is of the highest quality possible, every time?
In a recent webinar, Verinovum and HealthLX shared real-world experiences about how data quality is foundational to maximizing the benefit of modern healthcare IT. And the key questions were evident:
What is the vision of big tech in healthcare?
By “big tech,” we refer to companies like Amazon and Microsoft with a vision that all available evidence on an entire patient picture is actionable: That more data and insights mean better outcomes. The idea is to take all available research and marry it with the human factor in the clinical world, producing the best quality outcomes and reducing costs.
What causes data quality problems in healthcare?
In recent years, payers and providers have come together more. They’ve shone a light on the importance of data quality. But the reality is that there are widespread issues of data quality. In truth, payers aren’t embracing this at the necessary level and “big tech” isn’t entrenched in reality – we’re not getting the adoption that we need.
Data is flawed and the human body is complex. Attempting to measure and understand the human body using data points isn’t easy. Data continues to face a range of problems–from use of different tools for data collection to missing data, untimely reporting, and human resource constraints.
There are too many detours, including:
- EMR upgrades and conversions
- Aggregation of disparate EMRs
- Changes in workflow
- Centralization of data in the data lake
- Shortsighted data manipulation
- Evolving interoperability frameworks
Why can even the most sophisticated algorithms underperform and put patients at risk?
Data quality, fundamentally, in healthcare has a greater impact. If there are issues with patient identifiers, or with observations, or if there’s a desire for Natural Language Processing (NLP) clinical notes, tools like NLP won’t work correctly if the data is bad, and the assertions made from those analyses won’t be good.
There are companies, one large big tech firm in particular, that’s investing millions of dollars into building out the tools around NLP for provider data. The hypothesis is that it will only work if the clinical data being fed into this FHIR data repository is of high quality. If it is not high quality, then any assertions that are going to be made from the analytics out of this platform will fail.
And there’s the human factor – the majority of data originates from a manual entry in any number of disparate systems. Mistakes can, and are, made.
What is necessary to shift outcomes?
The only way to get to the future of high-quality data is an iterative approach to:
- prioritizing desired outcomes
- analyzing data requirements
- curating and enriching that data
- taking action on the data you have available
The industry is taking steps to improve the process. The HL7® FHIR® (Fast Healthcare Interoperability Resources) standard defines how healthcare information can be exchanged between different computer systems regardless of how it is stored in those systems. FHIR is based on internet standards widely used by industries outside of healthcare.
A key provision of the latest Cures Act ruling is the requirement that developers use FHIR as the technical standard underpinning the application program interfaces (APIs) that healthcare applications use to exchange data with other applications and information systems.
Because of the FHIR mandates, a variety of applications will become a part of the ecosystem. The member portal experience will improve because more information will be available to members. It’s a huge turning point in the history of healthcare in the U.S. and will be a fundamental part of doing business moving forward. FHIR is a modern technical standard that enables applications to plug directly into EHRs or claims databases to obtain patient health data.
Although the deadline for compliance with the FHIR patient access rules is fast approaching, most payers are not prepared, according to a June report from Gartner. That’s because implementing these new standards presents several challenges to payers.
What can organizations do to ensure that their expensive technology investments are optimized with high-quality and accurate data?
Data integration unto itself is no longer enough. Healthcare needs to begin with a circular approach rather than a linear approach to prioritizing the things most important to each organization.
Big tech (or small tech) enabled analytics, AI, and BI are all feasible today but require a few key principles to reach their full potential, such as:
- plan with the end in mind
- establish a strong foundation of data quality and data governance
- select tools that empower people and processes
- build an iterative approach to solving the data quality conundrum with targeted curation and enrichment
This is something that organizations need to be thoughtful about and plan for the long term. Each team needs to be prepared to continuously evaluate the data available, tied to the use cases they’re trying to drive and the quality requirements necessary to make them happen. Organizations who think about this journey as an iterative approach from the start will be poised for long-term success.
To learn more about this process, view the full webinar, Self-Actualizing Big Tech’s Healthcare Vision
About Mike Noshay
Mike Noshay, MSE, is Chief Strategy & Marketing Officer, Board Member at Verinovum. As Chief Strategy & Marketing Officer, Mike is responsible for all prospective client interactions, including client success and support, strategic planning, marketing, and business development.
About Verinovum
Verinovum is a market leader in clinical data curation and enrichment, enabling payers, providers, and partner organizations to improve operating performance and quality with actionable data. By delivering clean, complete, and accurate data, Verinovum supports healthcare organizations in their efforts to access the right information, in the right format, at the right time, so it may be tailored and curated to fulfill specific use cases and achieve mission-critical clinical and business goals. Discover more at Verinovum.com.
