You don't think twice about sending money from your bank account to another person's bank account. This could even be a different bank. But the money gets transferred seamlessly and promptly. This is interoperability, and it doesn't happen only in the financial industry anymore. The healthcare sector has been taking fair advantage of this trend for some time now as well.
Interoperability has now been effectively paired with Artificial Intelligence. Artificial Intelligence or AI so far played a very crucial role by offering analytical insights into huge powerhouses of data, giving healthcare professionals more power than ever before, including access to potentially life-saving information without losing time. An individual may have hundreds of thousands of healthcare data points, and all this data will have to be collected and used appropriately through machine learning. This is where AI works with interoperability.
Now, let's examine the two basic components of true interoperability:
1. There are two or more systems that could be interacting with each other for exchanging information.
2. The collected information would then be used effectively to serve its purpose.
AI can help with healthcare in a number of ways
The broad advances made in the field of Artificial Intelligence have enabled it to give a major boost in the field of healthcare. Here are a few areas in which that can be applied:
1. The limitations with legacy systems - One of the curses of healthcare organizations is that the systems and environment become quickly outdated and obsolete. Some of the applications in such systems would be pertinent to daily activity, and some stay unused for a long time and eventually affect the confidentiality and integrity of the healthcare systems. Such complications in legacy systems prompt hospitals to move to modern technology systems and depend on AI to use these systems optimally.
2. Gather the best from a multitude of smart devices - AI helps in the ubiquity of interconnected smart devices. Many patients have more than just a smartphone, they have plenty of smart devices that could help them clinically, and the data that's coming in from all directions is massive. AI helps in analyzing the data, making useful insights from them all so healthcare professionals can take timely action and treatment.
3. Smart devices are available to patients - In-home, voice-enabled devices are getting popular, and this opens up a realm of possibilities in healthcare with patients having their own devices that tell them what's happening with their body. Artificial Intelligence helps in providing better patient-centric services on a timely basis.
According to a report by Markets and Markets, the market for AI is going to grow really high. With an annual compounded rate of 62.9%, the market could reach $16.6 billion by 2022. That really does sound impressive, doesn't it? This makes AI the driving force behind healthcare. And it is so important to have some sort of technology to sort, manage and sort all the data that's coming because it's going to be really massive. This technology really has the potential to touch some major areas in healthcare:
- The quality of care can be significantly improved when you have a data-driven approach at your disposal.
- The doctors can make credible decisions because they are armed with real-time insights in their hands.
- AI-powered healthcare will definitely boost the human element in it, thanks to clinical expertise, diagnostic brilliance, and incredible predictive abilities.
AI plays a crucial role in different aspects of healthcare, be it precision medicine, imaging, diagnostics and many more. Once the AI models receive the right data, the impact it can deliver will be tremendous. The technology can even help in conducting surgeries with very little room for error; robot-assisted surgeries will soon be the norm. And you can connect several systems at the same time, and monitor the data that flows in constantly. The efficacy levels would be 100% with absolutely no loss. AI even helps in ensuring accuracy in lab tests.
Why EHRs were not enough
When Electronic Health Records were first introduced in the early 1970s, it came as a major boon. And with the growth in technology, there was noticeable efficiency in the way physicians managed the health records of their patients. However, with big data, EHRs began to face challenges The great promise of EHRs to promote better-coordinated care to patients came to be affected. Digitization was one reason for this. Earlier, physicians had to spend a huge chunk of their time managing the paperwork of their patients or decoding the information recorded by previous physicians. This was majorly where EHRs helped, including looking out for transactions where diagnosis, treatments, hospital visits, physician notes, and other details are noted. This was all accessible to patients as well, so they have access to accurate data whenever they need it.
However, this is not so good at all times, because there are limitations in EHRs that need to be addressed. The security of data that's shared across different platforms is a serious concern. Details of the patient's information like bodily function, substance abuse or psychological problems, if any, are some information that shouldn't be shared freely, and lack of security measures could force them upon the wrong hands. In fact, hackers make use of all this medical information for their own purposes. The abuse is almost like abusing financial information.
Countries across the world are digitizing health records by setting up an Integrated Health Information Platform or IHIP ensure there would be interoperability among all the health records and multiple healthcare systems.
Digitization was a new concept even when EHR was the standard practice, so it is natural that some amount of disruption can occur.
Why interoperability is difficult to achieve between EMRs
The best way to harness the total potential of EHR is through interoperability. It helps healthcare providers provide better-coordinated care at lower costs, but data is easily accessible to different physicians in a safe manner. However, it is not easy to achieve interoperability among the different devices because there is a large number of fragmentation among the different vendors.
The large companies integrated with EHR systems cater to the interests of specialized providers. Some cater to the demands of large hospitals, while some serve the needs of smaller physician practices. The challenges could be pertinent while sharing patient records and patient information across organizations, between providers who are geographically dispersed, between different EMR vendors, and even among vendors using different instances of the same EHR or in competition with each other.
62% of physicians surveyed would like their current systems to be more interoperable. (Image Source: Deloitte)
Here are some commonly faced challenges in the usage of EHR:
- Inability to share information across systems - There is a lack of secure, compliant and private communication system that allows individuals to share medical information, but at the same time support enterprise needs.
- Issues in getting 100% interoperability while sharing electronic records - Achieving 100% interoperability when physicians share their patient records with other providers irrespective of the software they use is so challenging.
- Providers get competitive or are geographically dispersed - One of the main reasons why the data is not getting shared properly is the competitive spirit of the vendors in a negative way. They don't like to give out user security and permission, and they try to limit the data sharing so the patient information remains in their system.
- High technology costs - There was a huge spike in technology costs, perhaps due to the high-demand in physician-owned multi-specialty practices.
Why interoperability should be paired with AI
EHR developers use Artificial Intelligence to perfect the digitization of medical records. Intuitive interfaces, voice recognition, dictation and automating routine processes are some of the perks associated with it. Just like we are used to Siri and Alexa at home, virtual assistants will also soon make their way to the patients' bedside to provide clinical assistance using embedded intelligence. AI can also use routine requests from the inbox and even prioritize tasks like a clinician does.
These are exactly the reasons why you need interoperability. Processing big data, precision medicine, and all other advantages that come with AI can be optimally used only when interoperability is there. The competition is also mostly about delivering care that's precise, accurate and timely.
Apple is a great example of this scenario. The company is making huge waves in clinical studies, using data with ResearchKit (medical research collected data from every source) and CareKit (where patients can manage their own medical conditions). CareKit can be integrated with hospital medical records, have an activity scorecard, symptom checker and monitor the progress of a patient. Since data is the main thing in AI applications, it can even provide medical researchers with information that wasn't easily available.
The American Medical Association plans to unleash a new dimension to patient care through Integrated Health Model Initiative (IHMI), with an aim to use the patient-generated data from devices and mobile applications and help in providing long term care.
According to HIMSS (Healthcare Information and Management Systems Society), interoperability happens in three stages:
- Foundational interoperability - This is the basic tier where data is received from one HIT (Health Information Technology) system, but not interpreted. You can consider this as the first step in the communication pyramid.
- Structural interoperability - This tier forms the middle layer and talks about the format of data that's exchanged between different systems. The structuring of the messages is important and aids in the uniform movement of health data.
- Semantic interoperability - The final tier of interoperability has two or more systems exchanging information (both sending and receiving). The receiving HIT system can interpret data accurately.
The explanation of the different kinds of interoperability is quite basic here. We will follow this up with a detailed article later.
By looking at the above, we can conclude that there are two basic components to true interoperability:
1. The capability of two or more health systems to impart and receive clinical information.
2. The ability of HIT systems to use the information for data analytics and patient care purposes.
How AI can help achieve interoperability between EHRs
AI capabilities for EHR systems are continually being explored, and the integrated delivery networks using the technology with EHR help to make it more flexible and intelligent. Through interoperability, AI makes the EHR to be more user-friendly, apart from providing personalized care options and gleanable insights through data discovery and extraction.
AI can collect both structured and unstructured data silos in the healthcare industry through interoperability. Achieving interoperability is very important since it would enable improved access to records and pertinent health information. Here are some ways in which AI works with EHR systems to make them flexible and accessible to healthcare professionals and to patients:
1. Extracting tons of data from faxes, clinical notes - Procuring data from clinical notes and faxes can be a hard task. But now you have service providers that can help you collect the data and gain meaningful insights from the same. For example, Amazon Web Services has a cloud-based service that helps you collect and index data from clinical notes.
2. Algorithms that aid in timely diagnosis - It would be greatly helpful to patients if clinicians can detect diseases and certain conditions before they actually occur, for example, heart failure. AI collected from various sources can identify patients with serious medical conditions and how they would respond to it. This would then be fed into the EHR system to help with treatment strategies.
3. Documenting clinical reports accurately - Clinicians might prefer to be spared the mundane details of worrying about their screens and keyboards. Through natural language processing, it is possible to capture clinical notes to help with data collection and preparing reports. For example, Nuance is a healthcare solutions company that integrates EHRs with AI-supported tools for this purpose.
4. Top-level accuracy in clinical decisions - Earlier clinical decisions were more generic and based on rules. But machines have become smarter and by using this kind of artificial intelligence, it is possible to offer higher levels of personalized care. IBM Watson, Change Healthcare and AllScripts are examples of such services.
Conclusion
Interoperability and AI do have a symbiotic relationship because of the challenges involved in that too. These come with data governance, interagency relationships, and developing common data de�finitions across the state healthcare enterprise.
Want to know more about how interoperability and AI can transform your healthcare system? We can help!
