Darren is the chief executive at Apixio.
Despite spending more than $3 trillion a year on health care in the U.S., we do not yet have a way to easily access your complete medical history. The average hospital reinvests in MRI machines every five years or so at a cost of millions of dollars. Yet, comparatively little has been directed toward unlocking some of the most valuable information health insurers, physicians, and hospitals already have about you.
The U.S. produces 1.2 billion clinical care documents annually, but nearly 80 percent of the data they contain is unstructured. This information is difficult for entities to understand and use. The medical chart contains a record of your health care — visits with doctors and hospitals, treatments, procedures, medications, diagnoses, and the results of workups. It is the key to understanding your health and improving the care provided to you. Yet, the challenge of accessing and making that information useable is immense.
The typical medical chart is stored in various fragments across different locations and systems. Your primary care physician has their record of you but not the record from your cardiologist or gynecologist or from the emergency department doctor you saw six months ago for bronchitis, for example. Imagine your entire medical record as a jigsaw puzzle in which the pieces are scattered and stored in different locations and different types of boxes, each of which is hard to open. No wonder people feel as if they are repeating themselves every time they visit a medical facility.
Luckily, technologies that make sense of the immense amount of data and preserve the patient narrative are rapidly emerging. With the rise of cognitive computing, natural language processing (NLP), and data science in health care, we now have the power to unlock untold value in health care data and drive proactive, targeted health care.
The first step is being able to make sense of the rich narrative in the medical chart whether from a primary care physician or specialist practicing in different organizations, different regions or both. While your doctor is familiar with your record, the medical system as a whole is not. So medical care continues to be reactive rather than proactive.
This is where cognitive computing and NLP enter the picture. NLP tools can help extract data from free text found in the patient record creating valuable material for big data technologies to analyze. Cognitive computing platforms use NLP along with pattern recognition models and data mining techniques to simulate human thought processes in a self-learning computerized system. A cognitive computing platform refines the way it looks for patterns as well as the way it processes data so it becomes capable of anticipating new problems and modeling possible solutions.
Once doctors have access to patient data, the question is how can they use it to accurately predict what treatments are most effective. Data science gives rise to a better understanding of the relationship between treatments, outcomes and patients. Now health care organizations have the tools to combine data from different sources and paint a more complete picture of the patient to personalize their treatment.
A data-rich health care future
These technologies give health care organizations the ability to access the previously untapped 80 percent of health care data so providers have real-time access to information and a deeper understanding of patients. If doctors know more about patients, then they can make more intelligent decisions that will result in quicker recoveries, fewer readmissions, lower infection rates and fewer medical errors. Ultimately, it will supercharge the value of care.
Beyond benefiting individual patients, access to this data will also create a living laboratory of clinical data to better inform health care decisions. Now that information about clinical care can be machine read, physicians can access it and base research on the everyday clinical care of millions of patients. Rather than depending on narrowly designed studies that do not directly relate to individual patients, health care organizations and researchers can learn about health care delivery from everyday real-world data.
Big data technologies can make use of information that is locked up in our medical charts in different systems and locations so we can transform how we look at and interpret patient health. With access to the untapped 80 percent of patient data and tools that put the data to use, we can change the delivery and consumption of health care as we know it, and usher in a new data revolution in health care that will improve patient care and result in high-quality outcomes.