Data and Interoperability as the Engine of Digital Transformation
Technologists often emphasize the significance of facts and evidences, but nowhere is this concept more vital than in healthcare technology, as the COVID-19 outbreak has proven. Data and interoperability are the two most crucial components of next-gen digital transformation agenda. Without them, innovation is certain to fail. In an era when healthcare is coping with a rush of new data sets and ramifications, this crucial part will explore the progression of data as evidence and where it is heading to. Healthcare is a white whale and interoperability remains a mystery for most businesses.
It’s important to remember that competitiveness is at the core of evolution. It necessitates constant innovation and new ways to increase revenue. Digital transformation enables businesses to develop more quickly and effectively. And therein lies one of the most difficult obstacles: pursuing digital transformation is certainly like constructing a race car and competing in an endless race. You’re mending, altering, and modifying your car for a mid-track switch in this competition with no pit stops, since speed is everything.
Hence, healthcare incumbents have to keep data compartmentalized in various systems, while the recent Fast Healthcare Interoperability Resources (FHIR) guidelines 1 encourage innovation, interoperability, and data flow. For years, Americans have desired more disclosure in healthcare, and now there are regulatory organizations in place to back it. However, utilizing new interoperability mandates necessitates specialist understanding in order to make data useful, secure, and comply with changing rules.
Barriers to Healthcare Digital Transformation
As even more digital technologies enter the fray, the future of healthcare technology will become even more complex. Are healthcare software systems and advanced analytics prepared for the onslaught of data and web-based health applications? The journey of modernization for organizations employing traditional software testing methodologies entails a large amount of human effort in the preparation and management of test data. When modern technology, such as AI, engages with all these systems, the very same data quality issues arise, which might result in prejudiced models that fail to produce results.
A data availability crisis is also brewing, threatening to derail digital transformation, as 97 percent of healthcare leaders polled said they’re pressing for increased data interoperability. Three out of four healthcare executives polled stated they require clean, interoperable data that works across all of their systems. Nearly half of respondents (42%) said their organizations’ data is severely disjointed and siloed, 41% named data interoperability as the biggest technical impediment to innovation, and 58% stated their EHR provider couldn’t implement their corporate data vision.
To summarize, the five most significant interoperability concerns in healthcare are.
In healthcare, legacy systems are still prevalent. According to reports, Microsoft would not support 70% of Windows-based medical legacy equipment 2 as of January 2020. Many legacy systems were built during a period when network connectivity was scarce. As a result, they lack contemporary security mechanisms and are difficult to keep up with. To fulfill current interoperability norms, they must be modernized.
Healthcare firms can use a hybrid cloud as an intermediary solution to retrieve data from legacy systems and make it accessible to contemporary programs. Furthermore, some interoperability rules for healthcare are better suited to legacy systems than others. For legacy standards, FHIR, for example, provides well-documented mappings.
Information exchange is frowned upon
Some healthcare professionals are hesitant to share patient information that they have. For example, hospitals contend for patients with walk-in clinics, so institutions aren’t inclined to comply when the clinic requests patient records. Whenever it comes to providing health information for research reasons, the situation is identical. Due to rivalry among organizations and some legal concerns, a group of scientists from various Swiss universities 3 claimed difficulty getting important medical statistics for research purposes.
Pew Charitable Trusts 4, for example, has urged the Biden government to alter data-sharing policies. “Through its standards relating to electronic health records (EHRs), the federal government may assist make data-sharing among healthcare providers and public health organizations more straightforward,” Pew added.
There is no conventional method for identifying patients
Most healthcare centers use their patients’ names, birthdates, and social security numbers to identify them. However, there is no standard structure for this conjunction, and not all clinics use it as identification. There is no interoperability when there is no accepted agreed-upon way of referring to a patient.
Many people believe that the universal patient identification UPI 5 is the answer. A UPI is a one-of-a-kind medical identification number that can only be used with healthcare data. As a result, obtaining a person’s UPI will prevent them from accessing their bank information.
Requests for patient information sharing needs approval
Healthcare institutions prioritize the safety and security of patient data. As a result, the request validation and approval process must be regulated. Keep in mind that patients are the proprietors of their own information. Make sure you have a good consent management strategy in place, which includes cases where consent is gained from someone who is accountable for the patient, such as parents consenting on behalf of their children.
Existing standards are diverse
There is no single healthcare interoperability guideline on which the medical world can rely. Unfortunately, the tool that is designed to help streamline the process is really hampering it. Even within the same criteria, says Michael Gagnon, Executive Director of HealtHIE Nevada, there might be differences. “I can tell the difference between two C-CDAs from two different suppliers by looking at them. It complicates everything. As an HIE, we have to figure out the distinctions between each vendor and try to produce something usable out of the data we can get.”
“The EHR suppliers don’t really want to come together and develop one way of accomplishing anything,” Gagnon said, adding that interoperability of electronic health records is also a concern. This puts pressure on them to improve everything in a unified manner. They’d have to come to an agreement on a standard, and when you attempt to do it in healthcare, you usually end up with something that is really dumbed down."
Despite the barriers, and in the face of the COVID-19 disease outbreak, the digital transformation of the health care system has expedited and boomed since 2020, according to the Surescripts 2020 National Progress Report, owing to the mass adoption of interoperable systems to facilitate innovations in patient care, cost containment, and specialized treatments.
Using an interoperable advanced analytics for transformation, striving to maximize data
To get the most out of comprehensive patient data, health systems and HIEs need to take a platform approach to data management—one that is accessible, interoperable, scalable, and capable of applying intelligence to data at any speed, either structured or unstructured.
Imagine the following scenario: an iPhone arrives with the iOS mobile operating system, which includes several native features (e.g., GPS, cellular or Wi-Fi connectivity, a camera, biometric security, a compass, a clock, etc.). iOS is an independent app development ecosystem that is compatible with other platforms like an android. We could have over 200 apps on our phones from various software businesses, which all use native iOS features. Just to cite an example, when we tour, we could use Apple Maps for directions, Waze for traffic analysis, and Citymapper for traversing new European cities. All three of these apps use open APIs (application programming interfaces) to access the built-in GPS in iOS, eliminating the need for each developer to recreate the GPS wheel. The iPhone also syncs data from the Apple Watch, including heart rate, sleep patterns, and physical activity, which can all be shared with other fitness apps (e.g., Strava). These apps offer a variety of metrics, recommendations, and forecasts to help people keep healthy and active.
An interoperable analytics platform must give extensive health system and HIE information that healthcare practitioners need to properly comprehend a patient’s status and make the most educated health choices, following the iOS interoperability paradigm. The system receives HIE data, which captures health information from primary care doctors and experts to laboratories, skilled nursing facilities, visitation caregivers, and emergency care. Consider what would happen if healthcare organizations used a platform model for managing data analytics at an empirical level, allowing them to incorporate and interoperate with data from all over the healthcare ecosystem, as well as the 80 percent of social determinants of health not discovered in medical care settings:
Clinical and business reasoning that is adaptable
Registries, value sets, as well as other data algorithms are layered on top of raw data and may be retrieved, repurposed, and modified via open APIs, allowing for third-party application development.
Pouring Data in real-time
Near-real-time data streaming from the source through the data operating system to the information’s representation, allowing for both transaction-level information exchange and analytic computation.
Combines structured and unstructured information
Textual and structured data, including graphics, are combined in the same ecosystem.
Integration of workflows
The capacity to transmit such understanding at the time of decision-making, even directly into the pipeline of data sources, such as an EHR, is one of the strategies for defining information in the data OS.
Architecture of microservices:
Standardized microservice APIs are available for data operating system functions such as identification, authentication and authorization, data pipeline monitoring, and software platform telemetry, in conjunction with detached data processing. These microservices were created with the goal of allowing third-party applications to run on the data os.
Artificial intelligence and machine learning
The data operating system runs deep learning and AI models natively, allowing for rapid development and implementation of these concepts across all apps.
Data lake with no bias
Any healthcare information lake, HIE, or data warehouse can be used by the data operating system. Several computing systems can use the recyclable concept structures (e.g., SQL, Spark SQL, SQL on Hadoop, et al.).
Many businesses, including hospitals, are enthusiastic about digital transformation — the use of digital technology to build or improve company operations, culture, and consumer engagement — in order to thrive and continue competitive. When individuals think about digitalization in the healthcare system, they usually think of using data to make better clinical decisions. Artificial intelligence (AI) systems, for example, are rapidly being employed in domains including radiology, cosmetology, gastroenterology, ophthalmology, and pathophysiology to enhance the visual identification of disease indications. Nevertheless, it’d be a misconception to use digital transformation only to enhance clinical decision-making.
We believe that the digital shift and evolution have a significant part to perform in optimizing hospitals’ organizational and operational decision-making, which can ultimately lead to enhancements in the quality and effectiveness of care and patients’ access to it, based on the HBR studies 6 and those of others, and the rapid expansion as well as innovation breakthroughs as far as how hospitals are now using data interoperability, information, and analytics.