How to approach interoperability in healthcare: an implementation guide

10.03.2020
12 min.

Interoperability in healthcare can enhance patient experience, reduce medical error rates, and lower healthcare costs, among other benefits. Health organizations are understanding this and turning to healthcare software development for interoperability solutions. According to MarketsAndMarkets, the global market of healthcare interoperability solutions is expected to jump to $4.2 billion in 2024 from $2.3 billion in 2019, rising at a CAGR of 12.6%.

With all the benefits it offers, healthcare interoperability is not straightforward to implement. Healthcare providers are struggling with too much incoming data, too many standards to choose from, the diversity of existing EHR systems, and partners who are not willing to cooperate.

A modern take on healthcare interoperability

Interoperability has been an important topic in healthcare for over a decade. It all started in 2009, when HITECH Act was introduced and healthcare providers were strongly encouraged to adopt EHR instead of paper-based records.

What does interoperability in healthcare imply?

In March 2019, HIMSS updated the definition of healthcare interoperability, to make it better suited for large healthcare ecosystems. HIMSS added an organizational interoperability level to the existing foundational, structural and semantic levels.

This definition is an aspirational one. It's what HIMSS is trying to accomplish and work toward when it comes to achieving global interoperability. We added the additional level of ‘organizational’ to illustrate the critical need for a robust interoperability infrastructure as well as highlighting the non-technical considerations that play into successful interoperability.

Mari Greenberger

To summarize, interoperability in healthcare is the ability of devices, software, and information systems to connect within or outside the boundaries of healthcare organizations to exchange and access patient data with the purpose of addressing health issues of individual patients and the population in general.

These are the levels of health information technology interoperability:

  • Foundational interoperability: the ability of one information system to receive data from another. The receiving information system is not required to interpret the exchanged data without user intervention.
  • Structural interoperability: this level is more complex. It allows defining the syntax (the structure) for data exchange, so that the receiving information system is able to interpret the data at the data field level.
  • Semantic interoperability: at this level, information systems can exchange and fully interpret received data. This is the most desirable level of interoperability as it supports the electronic exchange of patient information between authorized parties.
  • Organizational interoperability (appended in March 2019): this level adds non-technical components, such as social and organizational aspects, and exchange policies. Those non-technical components allow integrating interoperability into organizational workflows, and together with the previously described technical ones are expected to facilitate secure and seamless data exchange between healthcare organizations.

In 2019, HIMSS conducted a survey among 147 representatives of healthcare organizations to understand where they were in their interoperability quest. The results revealed that almost 75% of the respondents have gone beyond the basic (foundational) interoperability level.

Levels of interoperability reached by the HIMSS study respondents

Another question the HIMSS posed was about the practices the participants were using to achieve interoperability. Leveraging health information exchange (HIE) systems emerged as the most popular answer.

What steps is your organization taking to improve interoperability

Another recent survey by Deloitte discovered that the movement toward value-based care and abidance by regulations are the main drivers of interoperability. Meanwhile, data privacy and security, together with data standards, represent the greatest barriers.

Which of the following items will help drive the industry toward greater interoperability
Top three barriers your organization has faced in becoming more fully interoperable today

Common health IT interoperability standards

Healthcare has every standard we need. They're all out there, but they're not adopted. That’s really the issue.

Michael Gagnon
Fast Healthcare Interoperability Resources (FHIR)

This internet-based standard was developed by HL7 International. FHIR aims at establishing a set of resources that independently or combined can satisfy the majority of common data sharing use cases, and it can accomplish this at a fraction of competitors’ prices.

FHIR is the most recognized interoperability standard on the healthcare market today. It offers structural and sometimes semantic interoperability. FHIR supports several transfer mechanisms, including the common HTTP-based transfer and MLLP (HL7’s own protocol, which is recommended).

This standard is appreciated for its versatility. It can be used with apps, cloud-based systems, EHR-based data sharing, and server-based communication.

FHIR is about making the relevant data more available and more usable directly. When you make data more available, you create opportunities to inform better clinical and administrative decisions.

Wayne Kubick

Use cases:

  • Exchanging documents over a federated repository system
  • Aggregating data into personal health records and permitting patients to view it via patient portals
  • Supporting the integration of clinical decision-making tools into an EHR system

Limitations: Several standards of FHIR are available, which is confusing and, in a way, even hindering interoperability. Additionally, some developers have made documents exchanged over FHIR read-only.

Direct

Direct was developed in a collaboration led by the Office of the National Coordinator for Health Information Technology (ONC). This standard is a part of the strategy for transitioning from paper-based health records to EHR.

In its core, Direct is an email service with an added security layer, which is maintained by a Health Internet Service Provider (HISP). HISP performs encryption/decryption and digital signing of every message and attachment transmitted over Direct.

Use cases:

  • Patient messaging
  • Automated notifications regarding admission, discharge, and transfer
  • Communication of automated patient summaries for emergency departments

Limitations: This technology is not as well-known as FHIR; its costs are on the higher end; and there are some technical difficulties related to particular types of messages.

Cloud fax

A recent survey conducted by the American Medical Association (AMA) shows that most physicians are still relying on fax and phone for completing prior authorization for prescription drugs, while only 21% are using electronic prior authorization built into their EHR systems.

Cloud fax uses internet protocols to transmit faxes without requiring the physical machine. It allows transmission of messages and attachments to both digital and physical fax services. Those fax transfers occur over the Transport Layer Security encryption protocol, making the transfer HIPPA-compliant.

Cloud fax technologies limit data access to authorized actors by using admin privileges and unique IDs.

Use cases: Different types of information sharing, such as related to care transitioning and referrals.

Limitations: It is often confused with the physical fax machine, and healthcare providers are not motivated to use it in the light of the government efforts to remove fax machines from the health data exchange process.

Why interoperability is difficult to achieve

Implementing interoperability is not a straightforward process. However, being aware of the challenges helps you to prepare and tackle them proactively, thus increasing your chances of success.

Heterogenous data from various sources

Nowadays, healthcare organizations are forced to deal with a variety of data which is not limited to EHR and doctors’ notes. It includes data incoming from the internet of medical things used by elderly patients, patients with chronic diseases, and patients in recovery both inside the healthcare facility and on remote patient monitoring programs. All this results in a large amount of non-uniform and unstructured data that is challenging to encapsulate in one format syntactically and semantically.

Healthcare organizations can seek partnership with an external company that can act as a bridge between patient-generated data sources and the clinic. Some technology companies are offering novel solutions to this challenge.

One example is Validic. It specializes in helping healthcare organizations to benefit from patient-generated data. HealthChampion (headquartered in Chicago, IL) partnered with Validic to develop a platform normalizing and connecting personal data incoming from wearable devices with healthcare organizations.

Inconsistent patient records

Patient records are prone to error as they still rely heavily on forms filled out by patients at different stages of care. Wrong information can be entered at any point without anyone noticing. Relying on those forms results in EHR systems being full of inaccuracies and inconsistencies.

When patients switch their healthcare providers, they need to fill yet another form, which they might do differently than the previous one. For now, healthcare organizations cannot access health information of their patients stored by other medical facilities. A report by Pew Trust indicates that 1 out of 5 patients is not matched properly to their records when seeking care.

Opting for a cloud-based EHR can be a partial solution as it facilitates information exchange between healthcare organizations and between physicians within one organization. Using APIs can also be an option. Another viable solution, not limited to any single healthcare organization, is to adopt a national patient identifier (NPID), which can be implemented using blockchain healthcare solutions.

Redox, a healthcare technology company specializing in data sharing, is transforming legacy EHR systems with accessible APIs using its interoperability platform. This platform can be adapted to transform secure messages in the format of the customer’s EHR system.

In another example, MEDITECH has partnered with Google Public Cloud to offer a cloud-based EHR subscription model to its customers. This EHR platform will reduce infrastructure costs and help combat ransomware attacks.

Denied access to claims

As the healthcare industry is shifting toward value-based care, there is an increasing need to aggregate clinical data and claims in a manner that payers and providers can trust and take as a reference. This implies that real-time healthcare information such as admission and discharge records, transfer notifications, lab and examination results need to be synchronized with filed claims. This is complicated by the fact that clinical systems and payer adjudication systems are using different data-sharing standards.

Another aspect hindering aggregation is the unwillingness of insurers to cooperate with healthcare providers and disclose information.

Without having access to claims, healthcare providers will not retain patients’ full medical history and cannot track them throughout their healthcare cycle. Furthermore, being able to access claim data makes it possible to develop population health management capabilities.

To partially solve this problem, the American Medical Group Association advises payers to use APIs to facilitate accessing claims by providers so that they can better follow the value-based care principles.

Standard variability

Not only is the healthcare industry challenged by different EHR formats, but there is also a variety of electronic information exchange standards (as described earlier). Each standard is offering a specific set of features matching particular organizational needs. This might be a benefit for healthcare organizations, but it hinders interoperability in a broader sense.

How to proceed with implementation

Implementing interoperability is not an easy task, challenged by the variety of the technologies and standards available. However, technology is not the only challenge. Interoperability projects can also fail due to misaligned incentives and other human-centered factors. Interoperability is not only about technology, it is about enabling a specific use case of data exchange. As said by a former government official: “Interoperability is 20% technical, and 80% economic and political.”

Identify specific use cases

When you are planning to tackle interoperability, it is recommended to think of specific clinical use cases, the type of data to be exchanged, and the parties involved. This will help you to identify the most appropriate technical approach to data sharing: through a point-to-point connection, a portal, or otherwise.

What data about your patients do your providers need, and at what times? Will the provider pull data about a patient when needed from another organization? Do they even know what that organization is or if there’s even any data out there?”

Daniel Nigrin

While brainstorming your cases, also consider the future and the cases that might arise while your healthcare facility is, for example, moving toward a patient-centered model, meaning you will need to let patients access their data.

Develop your strategy and detailed execution guide

While specific cases are important, the overarching interoperability strategy is essential as well. Align all stakeholders around your vision of interoperability and how you are planning to achieve it by preparing an ‘implementation playbook’. Make sure everyone is on the same page about the value expected from the project, the resources needed, and the timeframe. Inform the senior management that their approval and support will be required throughout the initiative.

In this playbook, you will identify the steps to be taken, outline precise and measurable intermediate goals, assign priorities, and specify how you will comply with regulations. Among other important aspects to consider are the technology to use and the interoperability standard to adopt. Consider if there are any partnerships you need to form, such as with external tech vendors.

This is because interoperability projects typically stretch across entire networks and include many ‘firsts’ for many stakeholders, such as connecting independent practices and presenting patient-specific analytics into the physician workflow. Preparing stakeholders for these firsts with a common playbook and a defined plan ensures unity through the project.

Julie Mann

Begin with an end in mind

Identify the level of interoperability or the maturity state that you want to achieve in the future. Then, conduct gap analysis to have a clear vision of where you are and the distance you need to cross. When starting without a clear vision of the future, it is easy to deviate from the set path. As an old proverb states, if you aim at nothing, you will hit it every time.

Communicate ROI

Another best practice for an interoperability initiative is to clearly document the ROI expectations and what it means for different actors including patients, doctors, and the clinic as a whole. Identify the shot-term and long-term expectations. When everything is defined, communicate it to the relevant stakeholders as this will instill motivation.

For example, telling the management that after interoperability solutions are implemented, patients’ visits will become 5 minutes shorter, will motivate them to cooperate throughout the project.

Prepare your data

The key to interoperability is to aggregate, standardize and otherwise clean the data your organization is using. This will benefit not only your current interoperability initiatives, but also your machine learning and other analytics-focused endeavors. If you are planning to collaborate with integration companies, beware that some of them require clients to conform to predefined data standards before working together.

You need standardization of data collection and reporting with a clean UI/UX. There is no amount of AI, machine learning, data science, natural language processing, or fuzzy matching logic that can make up for starting with good data.

Kevin Mehta

While investing in data cleaning, be mindful of the fact that medical staff will be connecting to and requesting data from external sources. This external communication includes billing systems, lab results, EHR systems, and more.

Still have doubts about interoperability?

Interoperability is quite important today, and will gain even more importance in the future. Organizations that want to use data analytics to its full capacity will need interoperability. Organizations that want to shift to value-based care will need interoperability. Organizations that want to satisfy patients’ demand for accessing healthcare records will need it too. And those are only a few cases.

Deloitte brings forward the concept of radical interoperability and argues that by 2040 it will be an indispensable pillar of any modern healthcare organization. Radical interoperability implies that everyone’s health data will be integrated and available for research and medication purposes. It is time to prioritize interoperability-related initiatives and focus on it in current and prospective partnerships. Make interoperability a part of your long-term strategy rather than just a tactical goal.

Tags: