Posted on Jan 02,2018
The patient enters the hospital and the electronic medical record quickly begins to grow.
Vital signs. Lab results. Medication schedules. Physicians’ notes.
The information compounds at an overwhelming rate.
“Over 24 hours, you’re going to have about 800 megabits of data,” says Mohamed Rehman, M.D., chairman of the Department of Anesthesia and director of perioperative informatics at Johns Hopkins All Children’s Hospital in St. Petersburg, Florida. “If I tell you to look at the data, that’s a lot of data. But if I show you the very small segment of the data that is important to you based on the machine-learning algorithm and data visualization it may point out a medical problem in a timely fashion that may be lifesaving at times.”
Enter the emerging medical subspecialty of health informatics, a discipline that organizes and analyzes data to improve quality of care, safety and cost containment. Areas of computer science known as machine-learning algorithms, visual analytic techniques and natural language processing identify trends and patterns.
Informatics can do simple things such as creating an alert to remind the medical team to redose a child with antibiotics during a 12-hour surgery or flagging if someone enters a medication to which the patient is allergic. But it also crunches complex information to shape guidelines that might eliminate costly blood typing and matching for patients who don’t need it or to predict the best course of treatment based on genomics, metabolomics and other data, Rehman says.
“Precision medicine is all about trying to predict the best course and decrease risks and hazards,” says Sharon Ghazarian, Ph.D., senior director of the Health Informatics Core at Johns Hopkins All Children’s.
Even with computers crunching the numbers, big data still can be hard to comprehend. Health informatics has evolved beyond spreadsheets into visual analytics, the science of displaying information from large data sets in a comprehensible manner. For instance, Rehman says if you are confronted with a computer screen full of numbers and asked to pick out the 9s, it will be difficult. But if the computer displays the 9s in a different color, it is easy to focus on the information the algorithm is flagging as important.
“Visual analytics and machine learning,” Rehman says, “are going to be the future of medicine.”
Diving Into the Future
Johns Hopkins All Children’s started its Health Informatics Core in 2016. The hospital is making a broad effort to attract scientists, research assistants and informaticians—physicians trained in biomedical informatics—with the vision of becoming a leader in the field.
Part of that effort was recruiting Rehman, an early adopter about 20 years ago of electronic medical records who saw the potential in using the data collected for higher-level analysis. Rehman joined Johns Hopkins All Children’s in April 2017. Previously, he started the first biomedical informatics group in anesthesia and the critical care unit at The Children’s Hospital of Philadelphia.
“‘Big data’ is the big buzzword now,” Rehman says. “My teams found success in data analytics before it became popular.”
Ghazarian already has systems in place to manage clinical, financial and other data and analysis for the whole organization. The hospital recently launched DataLine, a streamlined process for all staff to request data and analytics on operations and administration, patient care, quality improvement, research or outreach/advocacy. Johns Hopkins All Children’s also is implementing its Enterprise Data Warehouse, a single platform to create patient records that aggregate everything from electronic medical, financial, patient satisfaction, affiliate and other records to allow more precise analysis.
"Johns Hopkins All Children’s is changing in a dynamic way," Ghazarian says. "Everyone is in a state of growth right now, so your expectations have to be fluid and your finger kept on the pulse."
Transforming Children’s Lives
The potential for advances in health care based on harnessing data is enormous. Improvements made in pediatric care could save a life or alter the course of a person’s health over a lifetime.
“The holy grail of development that we're looking for in the future is to spot patients who are trending toward critical conditions before it is apparent,” Rehman says. “The data warehouse platform and the algorithms running on it in real time will make this a reality in the future.”
Rehman was ahead of the curve as an early adopter of data analysis. He sees only great things ahead for the field and for Johns Hopkins All Children’s.
“I think we’re just at the very early phase of big data, visual analytics and machine learning,” Rehman says. “There are not many hospitals that have the vision to invest into this. Visual analytics and machine learning are going to be a core competence of Johns Hopkins All Children’s Hospital and help it become one of the premier institutions in the country.”
This story first appeared in Leading Care magazine.