Blog

Deeper Analytic Insights Through Shared Data

Information in the palm of your hand

Several weeks ago during a conversation with a health system chief information officer (CIO), the topic of data sharing came up. We talked about the types of data that each of our organizations were sharing today and what we hoped to share in the future.

The current state of the CIO’s health system: physicians and staff query the electronic health records (EHR) vendor’s network for data when it’s clinically appropriate for a patient encounter – e.g., new patient history, lab results, consult reports and so on. The catch is that only providers using the same EHR have data available in the network.

I asked when the health system planned to connect to a Health Information Organization so they can get data pushed for all patients from an EHR-agnostic network of providers. The CIO – also a practicing physician – responded, “We [he and the system’s chief medical officer] don’t find any value in exchanging data like that. We prefer to decide when we want to look for patient data and choose which data we bring into our EHR.”

Having spent the better part of 15 years moving data between healthcare organizations and turning it into actionable information through analytics and decision support tools at the point-of-care, I was surprised. Surely an organization of this size, with a fairly mature analytics program, realized the value in leveraging untapped data to enrich the insights of its patients.

Three weeks later, I was speaking with another CIO and chief medical information officer and learned that their hospital had taken a similar stance regarding data sharing – the value wasn’t there for what it would cost to build and maintain interoperability. They also didn’t see value in sharing their data with external organizations.

For years, we have been hearing about the challenges of interoperability:

EHR vendors do not want to play nicely in the sandbox. Standards for data exchange aren’t really standard. Organizations focus on the minimum necessary data to meet Meaningful Use or value-based arrangements. More interfaces means information overload. Organizations block data.

In that same time span, proponents of data sharing have been touting its benefits. External data fills in the gaps of what we don’t know about a patient. Problems, medications and allergies are no longer a mystery when a person presents at a new healthcare provider. Primary care physicians would know about important events (e.g., visit to an ER) in near real time without having to wait for the patient to mention it at the next visit. Care plans can be shared among disparate care teams. Knowledge of services already provided (e.g., claims) can cut down on overutilization. Clinical quality care gaps can be closed with actual data versus manual reporting.

These are all great points to discuss data sharing. However, in many discussions around the value of data sharing, we often overlook the downstream benefits through reporting and analytics.

Beyond the EHR

The benefits of shared data within the clinical workflow are well documented. From an analytics perspective, data sharing opens up a whole new world of possibilities to support better healthcare at lower costs. Here are a few ways data sharing can create value for organizations and patients.

Generates a 360-degree view of a person based on clinical, claims and financial data.

Robust longitudinal and episodic views of a person are only possible through the integration of disparate data and the delivery of information outside of isolated application environments in siloed organizations.

Bringing data together can be especially important in value-based arrangements where a physician bears the responsibility (and risk) of care for a patient to ensure that clinical quality outcomes improve and cost and utilization metrics are reduced.

If the payer side of a value-based contract is collecting data and subsequently sharing that with providers, the arrangement has a better chance of success through more robust reporting on key metrics. Likewise, health plan quality reporting will be more complete when supplemented by external clinical data that wasn’t previously available in claims. The value of reducing provider and patient abrasion (i.e., lower customer satisfaction rates; conflicts between payer, provider, and/or patient; etc.) due to the lack of complete data cannot be understated.

Builds better analytic models.

Whether predicting readmissions, assessing the effectiveness of treatment, or stratifying populations based on risk scores, more data can translate into more effective models.

Providers have a wealth of clinical data. Health plans have claims with cost and utilization information. Open exchange among clinical organizations and between providers and payers should improve analytical views with a broader set of historical data which can be used to develop stronger, more accurate predictive models. Feed the model further with other data such as social determinants of health, open data and genomics.

Enables discovery of undiagnosed conditions.

Advances in natural language processing capabilities have opened up doors to the discovery of conditions that remain hidden within the depths of a patient’s scattered medical history.

Consider Geisinger Health System’s NLP project to identify patients with a dangerously large, undiagnosed (i.e., untreated) abdominal aortic aneurysm based on thousands of radiology reports. Over the course of two years, they claim to have saved 12 patients through surgical intervention. While not every condition is life-threatening, advanced analytic projects have shown that leveraging internal and external data can save lives.

Strengthens efforts to reduce fraud and waste.

That there is a lot of fraud and waste in healthcare is perhaps the elephant in the room of interoperability. The reasons behind this are probably best saved for another article, but suffice to say, services are performed and medications are prescribed more than necessary or for no justifiable medical reason. Whether organizations knowingly contribute to these issues or not, there is no doubt some concern that data sharing can open up a door that can’t be shut again.

If we ever hope to be successful in reducing healthcare costs, we have to find effective ways to reduce unnecessary services. Payers do not have to hold the monopoly on tackling this issue; providers can use data and analytics to attack fraud, waste and abuse. These types of analyses are limited in effectiveness if using only claims or only clinical data. By using both data types, likely shared by external organizations (e.g., another clinical provider or a health plan), a number of challenges can be addressed: identification of improper billing, identification of over-utilized services or over-prescribed medications, calculation of provider risk scores and analytic models to predict future fraud, waste or abuse.

How Can You Help?

At the end of the day, healthcare is about the patient, the member, the person – not market share, strategic advantage or intellectual property. Technical, financial, procedural and cultural barriers to interoperability all need to be addressed throughout the healthcare industry to build momentum and make open data sharing the norm. Those of us who are firm supporters of data sharing owe it to the industry to socialize the benefits and advocate a culture of proactive change.

Be an advocate for your organization and help answer and address these questions:

  • Does your organization share data today?
  • Do you maximize the value through analytics?
  • What are your barriers to sharing data? What will break those barriers down?

Be an advocate for the industry. Federally mandated interoperability improvements and calls to end data blocking in the 21st Century Cures Act are starting to gain momentum. The Office of National Coordinator or Health IT recently published a draft Trusted Exchange Framework and Common Agreement which “outlines a common set of principles for trusted exchange and minimum terms and conditions for trusted exchange.” HIMSS’s Interoperability Call to Action was published in October 2017, and seeks to support and advocate interoperability efforts by setting “guiding principles to inform health policy and spur our nation’s health sector to action.” Organizations interested in supporting the cause can find more information and links to sign up as a champion or supporter of the call to action via the HIMSS website.

Join the HIMSS C&BI Community today to stay up-to-date and learn more about the tools and educational opportunities focused on the value of data sharing and other critical topics that will help you on your journey to turn data to action. #PutData2Work | #PopHealthIT | #PrecisionHIT

Keywords: 
Data sharing; Analytics; Fraud and Waste; Value; Payers; Clinical & Business Intelligence; HIMSS