Breaking down the walls: Liquidating data silos for enhanced insight extraction

This is the first of three articles, powered by CHIME Digital Health Insights and sponsored by Philips Healthcare, exploring how AI can help improve patient outcomes and operational efficiencies by addressing the challenges of data silos and overload and fragmented systems in healthcare provider organizations.

The healthcare industry generates nearly one-third of the world’s data. This data, especially at the personal and population health level, holds tremendous potential to provide insights into what causes disease and how they can be treated or even prevented. However, large portions of healthcare data remain unstructured or uncaptured.

Unfortunately, much of that insight is either elusive as it is unstructured or uncaptured, or locked away because healthcare data is notoriously fragmented. For example, many clinical and claims records are still stored on mainframes. Often, research data is only available to small teams, and device data is a stream of numerical values, without context, left to stagnate in isolated pools.

Aggregating and harmonizing siloed data sets, let alone analyzing them, is a monumental undertaking. However, if data can be “liquified,” it can more freely flow across the healthcare ecosystem to entities with the appropriate privileges. This can give stakeholders a fully longitudinal view of the patient care journey – and the insight it provides is poised to revolutionize care delivery.

Data silos have consequences

Data silos are everywhere in healthcare. Providers have electronic health records (EHRs), inpatient monitoring devices, and imaging scans. Insurers and employers have medical claims. Public health agencies maintain immunization and infectious disease data. Life science organizations have research datasets from Phase I trials to post-market studies. Pharmacies store information on prescription drugs. Patients have digital health applications and wearable devices. Social determinants of health (SDOH) may be housed in social services and other types of providers and agencies. The list goes on.

Silos exist for several reasons. Institutions use legacy on-premises systems to store data sets, believing it’s too costly or complex to migrate them to the cloud. Amid concerns about data privacy and cybersecurity, stakeholders remain reluctant to share data. Data sets are rarely compatible; some use different variables to identify individuals, while others contain unstructured data that can’t be expressed in rows and columns.

The consequences of data silos are well documented. Clinical staff may lack the insights needed to recommend care plans or therapies tailored to individual needs. As a result, patients may receive fragmented care. Inpatients may require expensive transport to another hospital in the network simply because the data was siloed. Research and development efforts also suffer, as incomplete data sets make it harder to identify unmet needs in underserved or vulnerable populations. All told, 51% of healthcare leaders say silos make it difficult to use data effectively, which then raises the cost of care in an era of value-based care.

Connect the dots with data liquification

Data liquification is the process of making data easier to share among disparate applications and entities. The idea is that points of data, once recorded in one system, can be used by other systems downstream.

Successful data liquification involves several initiatives that occur in parallel:

• Standardization and aggregation. Create common data formats for important variables and leverage common platforms that ease data aggregation among applications.
• Interoperability. Adopt technical standards such as Fast Healthcare Interoperability Resources (FHIR) that facilitate data exchange among multiple institutions.
• Connected networks. Integrate data from enterprise applications (such as EHRs and claims) along with Internet of Medical Things devices and third-party healthcare data platforms.
• Governance and privacy. Develop frameworks that spell out who can access data, how data should be accessed securely, and how it may (and may not) be used.

Apply AI to liquidate data and unleash insights

Data liquidity makes it possible for healthcare organizations to access integrated, longitudinal data sets. Applying proven machine learning and AI models to this data can empower clinical staff with real-time insights at the point of care.

This can make many things possible for the future of healthcare. One AI model could identify possible signs of a chronic disease diagnosis or identify risk factors and recommend a screening for that disease. Another model could create personalized treatment plans based on a patient’s family history, genetics, lifestyle, and sociographic factors. A third model could help care teams create care plans for managing a chronic condition with therapeutics and lifestyle changes.

In each case, clinicians and patients are empowered to make proactive decisions that aim to avoid preventable complications, comorbid diagnoses, or hospitalizations, thereby improving health outcomes and reducing the cost of care.

Practical steps for moving forward

While organizations can’t achieve data liquification overnight, there are three important steps for getting started:

• Prioritizing data governance helps foster the culture of data collaboration that’s necessary for sharing data. Data from the 2023 Digital Health Most Wired survey revealed only 58% of healthcare organizations have AI and machine learning governance in place within their organizations.
• Identifying key data silos and prioritizing their integration with the larger data platform to shorten the timeline for achieving return on investment. As more and more distributed data is on the edge, there will be an increase in “invisible” silos.
• Bringing the customer into the process will ensure their pain points are identified and addressed early on, leading to streamlined solutions that are easier to understand and implement, especially given the trend of enhanced patient-facing portals.
• Investing in the right technology and infrastructure to move away from legacy hardware, software, and data management strategies. The next article will explain this topic in greater detail.

Breaking down data silos through “liquification” offers tremendous potential for healthcare providers. Standardizing, connecting, and governing data can unlock the insights hidden within disparate systems. Applying AI to this liquid data empowers clinicians, personalizes treatment, and ultimately improves patient outcomes and reduces costs. While the journey to data liquification requires strategic planning and investment, the potential rewards for patients, providers, and the healthcare system are undeniable.

As we explore this topic further in the next article, remember: the future of healthcare lies in connecting the dots, one data point at a time.

About Philips

Royal Philips is a leading global health technology company focused on improving people’s health and well-being through meaningful innovation, employing about 74,000 employees in over 100 countries. Our mission is to provide or partner with others for meaningful innovation across all care settings for precision diagnosis, treatment, and recovery, supported by seamless data flow and with one consistent belief: there’s always a way to make life better. For more information, please visit Philips.com/Global.