The United States’ healthcare ecosystem — as well as the global healthcare ecosystem — has a data sharing problem. Why? Because data is dispersed across different entities.
With claims data held by payors and clinical data held by providers, obtaining a holistic view of patients and their healthcare needs is nearly impossible. The National Academy of Medicine has questioned how the inability to access real-time data can be considered completely normal in the siloed world of healthcare data, yet is frowned upon in other high-risk industries, like aviation. Their concern closely aligns with the theme of a discussion held at the recent Empowering Beyond Summit 2024 (EBS), where the urgent need for data interoperability to address the siloed nature of healthcare data was highlighted.
Matthew Barlow, Avasant Managing Partner and Healthcare Lead, moderated a discussion, The Future of Health: Harmonizing Cost and Care Through Digitalization, which featured guests Tim Skeen, EVP and Enterprise CIO of Sentara Health, and Rick Hopfer, SVP and CIO of HMSA.
Healthcare payors and providers agree on this need for interoperability. There has been a shift in healthcare providers toward increased data interoperability to improve patient care and “provide real-time access to clinicians making collaborative decisions.” Similarly, the payor side has seen a shift toward increased adherence to interoperability regulations. Examples of this kind of connectedness between payors and providers are evident while looking at the case of UnitedHealthcare, which has developed interoperability APIs to make data sharing between itself and third-party vendors easier and more secure. Though increasing, EBS panelist Skeen noted that attempts at data interoperability are not new, with Health Information Exchanges and Health Information Networks programs having been funded 25 years ago and the passage of previous legislation such as the Cures Act and the Trusted Exchange Framework and Common Agreement (TEFCA).
Research shows that data interoperability in healthcare has the potential to improve the payor, provider, and patient experience and is crucial to better leveraging the benefits of artificial intelligence (AI).
Improves the Experience All Around
Data interoperability will improve experiences across the healthcare spectrum. Because disparate institutions hold different elements of a patient’s data, they are unable to consolidate them into one source where a patient’s full history can be easily accessed. This negatively affects patient outcomes, since patients notoriously struggle to accurately reflect prior treatments and medications. Barlow advocates for an interoperable system in which “when they [the patient] go to this hospital, it knows what they did the last time for treatment, and that information has proper continuity across a lifetime of care, not just a hospital visit.” He adds that complexity of data sharing increases once you add in the numerous small-scale providers contributing to data who each have their own ways of collecting and storing it. Accurate, consolidated histories not only improve treatment precision and efficiency for providers, but also benefit patients directly.
Moreover, insurers stand to gain from access to extensive care data, which according to Forbes, allows them to develop “more accurate risk models, which have the potential to lower premiums.” Panelist Hopfer discussed an added benefit of cost transparency and having all the information “in one site rather than having to go to multiple sites to get all your information.”
Skeen explained that data interoperability is already being realized through the concept of “continuity of care.” He highlighted the impact, stating, “the ability to instrument homes, remote patient monitoring, hospital to home programs, the power of bandwidth and the ability to move data effectively allows continuity of care throughout.” This effective flow of data makes it possible for patients to be monitored from home, reducing the burden on hospital resources and improving patient comfort.
Additionally, data interoperability is essential in addressing global health crises. Structured data that aligns with international health standards simplifies the analysis and exchange of information across borders, creating a more coordinated response to worldwide health challenges.
A report titled, “Why digital medicine depends on interoperability,” published on the NIH’s National Library of Medicine, found that networks such as the European Reference Networks and the Undiagnosed Diseases Network in the US are already facilitating data flow across regions. The adoption of standardized data formats and terminologies, such as the Orphanet rare disease nomenclature, are crucial in harmonizing international efforts to enhance research and care in these domains.
Interoperability is Crucial for Advancing AI
Data interoperability is not only pivotal for enhancing healthcare interactions but also for advancing AI in health research and services. Developing an AI model requires large inputs of data with which to train the model. With so many healthcare institutions collecting and storing data, the field lacks a standardized approach to data collection, a necessity if it is to be shared effectively between these institutions. The authors of “Why digital medicine depends on interoperability” explained that initiatives, such as the International Patient Summary and the European Committee for Standardization, not only help create standards to make data interoperable, but through standardization, create a large source of data from which to train AI models.
Panelist Skeen is already leveraging healthcare data to train language models and generative AI, explaining his tool as “an AI engine taking your entire history and getting data flowing into interoperable spaces. Using LLM and Gen AI to get hundreds of pages during care and creating succinct targeted discharge paperwork that can be used for post-acute care.” The ability to train AI models with vast amounts of standardized health data emphasizes the necessity of data interoperability to harness the full potential of AI in healthcare research and services. In other words, achieving these AI benefits is contingent upon implementing data interoperability.
Moderator Barlow envisions a future where AI and data interoperability serve as a dynamic toolkit for healthcare providers. Patient care data can be used to train AI models which, over time, will create a virtual assistant who “you’ll be able to tell [it] to go do stuff, and the task execution of the APIs will just get better and better.” While he predicts AI’s role in clinical-decision making is a long way away, he anticipates that its advancement in task automation on the administrative side will drastically change the healthcare sector.
The Way Forward
Though there are significant benefits, it is worthwhile to understand the inherent security risks that come with data interoperability. In 2021, 42 million patient records in the US were compromised through data breaches. Interoperability only raises the threat of breaches, because data exists in “many more application program interfaces (APIs) and connections between data networks,” according to a leading healthcare executive. There is a way to enhance security, though. Forbes suggests that by adopting robust encryption, electronic signatures, stricter access rights, and regular updates to security protocols for HIPPA compliance, we can secure our systems. A recent Avasant RadarView report highlights the Regenstreif Institute’s ability to integrate data but omit patient identifiers, setting an example for privacy maintenance. By emulating such practices and strengthening security measures, we can create a secure network connecting payors, providers, and patients.
By embracing data standardization measures, healthcare payors and providers can strive to create an interoperable data ecosystem that facilitates the smooth and secure exchange of patient information. This will not only improve patient care, but also streamline visits for providers and allow payors to assess patient risk more accurately. Additionally, a standardized data repository will be key for training AI systems. As noted by the panelists, an imbalance of incentives between payors and providers poses a challenge to the level of cooperation that interoperability requires. Though there is no perfect path forward, suggested solutions range from Forbes’ idea of imposing fines for information blocking, or the CDC’s push to encourage data exchange through the adoption of standards like FHIR (Fast Healthcare Interoperability Resources), an open-sourced framework for organizing healthcare data. Overall, the future of data interoperability is ripe with benefits for all stakeholders involved if payors and providers are willing to come together.
By Adriana Guzman, Associate Consultant, Avasant