What will it take to turn electronic health records data into valuable insights? Mark Scott, CMO of Apixio, gives his view.
There’s a wealth of information buried in electronic health records (EHRs). Valuable details about patients’ medical history, such as current conditions and successful treatments, exist across multiple systems in the form of scanned images, digital files and handwritten notes and charts. This data could change healthcare by improving individual care and informing treatment protocols more broadly. However, it is essentially sits in EHRs, inaccessible for analysis and use.
The urgency behind EHR data use
It is essential for care providers to have timely access to the full spectrum of patient data in order to provide accurate, individualized care. For example, when faced with a 30-year-old African-American woman with a new diagnosis of type 2 diabetes and a history of high blood pressure, a physician would need to draw on her full patient profile to determine the best initial medication to control her blood sugar.
On a larger scale, having real-time access to patient data allows for earlier intervention, effective matching of procedures and physicians, more personalized care and, ultimately, improved outcomes and a healthier population.
Messy healthcare data is the problem, but EHRs aren’t the solution
To make headway in solving the EHR data use problem, we must first appreciate the scope of the challenge.
The high expectations for EHR data have been unmet. However, it’s important to remember that EHRs were not really designed with care management in mind. Rather, they were designed to serve fee-for-service incentives.
In 2009, the government passed the Health Information Technology for Economic and Clinical Health Act (HITECH), which gave providers subsidies to transition to EHRs quickly. Providers raced to do so, but they had to make sure that the transition did not disrupt their entire operations. Since the move happened at a time when providers still operated under a fee-for-service payment model, EHRs were built to deliver fee-for-service results. They essentially became point-of-sale systems.
Instead of improving evidence-based medicine, EHR primary competency was capturing information for billing and claim submissions. It’s no surprise that healthcare organizations struggle to tease out meaningful clinical insights from EHRs today.
New analytics solutions pave the way to extract data from EHRs
The inability of EHRs to deliver comprehensive clinical insights creates the need for an independent software workaround. However, there are significant difficulties in developing such a solution.
Big data experts generally agree that 60 to 80 percent of the clinical data in EHRs is found in the unstructured (free text) archives of patient records and physician encounter notes. Accessing and analyzing this unstructured data is a tedious and expensive process at best.
Why is acquiring this data so difficult? Well for one, accessing data has to be done in a secure and Health Insurance Portability and Accountability Act 1996 (HIPAA) compliant way. Every person who touches the EHR has to go through HIPAA training and the files have to be encrypted and decrypted several times throughout the process.
“With the help of cognitive computing technologies, EHR data will become a valuable asset rather than a perennial stumbling block to providing patient care.”
Second, we lack a universal language or format. There are different EHR systems across an organization, or different versions of the same EHR system, and relevant data is located in a different place in each one. Last, computers can’t automatically read scanned data and make it actionable. When the computer looks at a scanned document, instead of English text, all it sees is a series of images or symbols.
Over the past decade the sophistication of data analytics technologies, such as cognitive computing and machine learning, has made it easier to extract data from EHRs. Cognitive computing is a combination of technologies that enable a computer to learn from its experiences and improve its performance over time. It was popularized by IBM’s Watson supercomputer, which famously used cognitive computing to win US TV quiz show Jeopardy. Cognitive computing can serve an important need in patient data analytics. If we apply these technologies to EHRs, then the industry will be one step closer to realizing the meaningful use of the data they contain.
The data-driven doctor paradox
After extracting clinical data from EHRs, one hurdle remains: physicians’ time. We have the ability to utilize all the data we collect on patients to drastically improve care, but the burden remains on physicians to make use of it.
In a short time, doctors’ offices and hospitals went from manila folders with handwritten notes to electronic systems, and the industry is struggling with how to best make use of the data without draining physician time. Not only do physicians need to spend time nurturing the patient-doctor relationship, but now they are being asked to also become data entry experts.
The move to a data-driven future is daunting to physicians who want to invest their time in delivering patient care rather than inputting data behind a computer screen.
With the increasing adoption of EHRs, there have been studies which show decreasing productivity among physicians, which is exactly the opposite of what was assured with the implementation of EHRs.
A recent study in the Annals of Internal Medicine found that for every hour physicians spent with patients, they spent nearly two additional hours on administrative work.
Among the non-clinical activities which physicians are mandated to perform is the proper selection of the codes that indicate pertinent patient diagnoses and the treatments provided during an encounter. In an era of payment-for-value rather than payment-for-services, these codes become an important set of data and are increasingly used for data analytics related to risk prediction, quality of care, practice patterns and resource allocation.
Improper diagnosis or procedure code selection following a patient visit in the clinic or hospital can result in inappropriate payment (too much or too little), inaccurate risk or quality performance measure results or poor targeting for proactive management of costly patients.
We should work in conjunction with cognitive computing to read medical charts to allow physicians and their office staff to spend less time doing taxing back-office work and more time treating patients. It will enhance physician satisfaction and elevate the long-term sustainability of the medical profession for many doctors. With the help of cognitive computing technologies, EHR data will become a valuable asset rather than a perennial stumbling block to providing patient care.
The perks of a data-savvy healthcare system
Making sense of patient data seems daunting, but it’s a challenge worth facing. Everyone can agree that having a complete picture of patient health opens dramatic possibilities for improved individual care.
Beyond benefiting individual patients, access to this data will also create a living laboratory of clinical data to better inform healthcare decisions. Rather than depending upon narrowly designed studies that do not directly relate individual patients, healthcare organizations and researchers can learn about healthcare delivery from everyday data. Doing so enables them to deliver personalized care based on what works best for patients according to real-world clinical data generated from millions of patient care profiles.
Healthcare is currently living in the dark ages, but the sophistication of technologies, such as artificial intelligence, machine learning and wearable devices, brings us closer to true industry disruption through greater access to patient information. While EHRs weren’t designed to single-handedly improve care and usher in more evidenced-based medicine, coupled with other technologies they can help move the healthcare industry closer to that reality.
Mark Scott is CMO of Apixio.
Apixio, US, Mark Scott, Healthcare, Risk management, Electronic health records, Data analytics, Technology, IT, Insurance