The Institute of Medicine defines a learning health system as “one in which progress in science, informatics and care culture align to generate new knowledge as an ongoing, natural by-product of the care experience and seamlessly refine and deliver best practices for continuous improvement in health and health care.”

Stated simply, in a learning health system, deidentified patient care data is used to identify evidence-based clinical practices to improve outcomes continually.

Anyone who works in healthcare has a desire to help others. And it’s this desire that pushes us to make the next day better than the last. But it’s also important that changes or improvements be based on reliable data.

One of the unique opportunities we have as a learning health system at HCA, is our ability to access data across 164 hospitals and 20 million patient encounters a year.

As a result of our scale, we see trends that smaller health systems may not see and are able to make improvements that positively impact more lives.

Here are two examples.

  1. REDUCE MRSA – we, along with partners at the CDC, the Agency for Healthcare Research and Quality, the University of California, Irvine’s School of Medicine and Harvard Medical School and Harvard Pilgrim Health Care Institute wanted to know the best approach to preventing MRSA in the intensive care unit. Our results were published in the May 29, 2013 issue of the New England Journal of Medicine.In just 18 months, we were able to implement three approaches to treatment at 43 HCA-affiliated hospitals. We proved that universal decolonization (ie: bathing patients with an antiseptic wipe and using nasal ointment for five days) not only reduced MRSA, but all bloodstream infections by 44 percent. This will be a new standard of clinical excellence.
  2.  My 39 Weeks – Traditionally, a full term pregnancy had been defined as 37 – 42 weeks gestation and deliveries at less than 39 weeks were not considered a risk for mothers or newborns. We wanted to know, scientifically, which length of gestation was best for baby.Since HCA delivers around 225,000 babies each year, we were able to gather data from 18,000 deliveries from 27 HCA-affiliated hospitals in only three months. We found that there is an increased risk of illness, such as lung immaturity leading to respiratory distress, for infants delivered electively at 37 and 38 weeks, as compared to those delivered at 39 weeks or later. As a result of creating the evidence demonstrating that early elective delivery is not as good as full-term delivery, many of our hospitals developed their own policies preventing early elective deliveries (known as a “hard stop”) while others used more indirect methods like education and quality monitoring.  I’m glad to say that approximately 96% of deliveries at all HCA-affiliated hospitals occur at 39 weeks or beyond.

These examples, and others like them, are the result of technology and big data enabling healthcare providers capture and implement new knowledge efficiently across a large system. There are many reasons a healthcare system might choose to operate as a learning health system. Chief among them for HCA is better patient care and improved patient safety.