5 Steps To a “Learning Health System”

Michael weilert MD Amy Abernethy

Amy Abernethy

I recently came across an article by Dr. Amy Abernethy, who delivered the opening keynote address at the American Medical Informatics Association’s (AMIA) annual symposium.  Without good data to back it up, she says, patient-centeredness is nothing more than a buzzword.  And without a patient-centered focus and proper organization, data tends to be pretty useless.  Comparing the flood of data now available to the Amazon River, Abernethy declared the need for a vision for a way forward, called a learning health system.  This needs to be powered by a “river of data”, with a “North Star” to guide operations.  To Janet, this North Star is a 37 year-old patient with melanoma named Janet.

While speaking with Janet on the risks of interferon treatment, Abernethy pulled in records from various data streams that formed rivulets, which dumped into a large river of data.  Informaticians, as members of the AMIA call themselves, face two challenges: exploring all of this data now available to them and explaining it to practitioners, payers, administrators and patients.  According to Abernethy, the vernacular in informatics doesn’t align with that in other parts of healthcare, meaning that there needs to be a common language.  As healthcare professionals think about trying to bridge the communication gap, people find it safer to keep on their sides and not talk to each other, despite the fact that they could be learning so much.  Abernethy spoke of the need to get better at bringing data to standard clinical practice, and outlined five recommendations to do so:

1. Putting the patient at the center.  Patients are the anchors of Medical Informatics, and their stories help clinicians work better with data.

2. Having meaningful data by putting the information into the right context.  This information, says Abernethy, should be taken as more than just a snapshot in time, so that it can be repurposed for research.

3. Improve data quality by putting that data to use.  Informaticians need to use data to make sure that it’s accurate.  They care deeply about data’s quality, sanctity, security and validity.

4. Data needs to be trustworthy.  To get information from patients over time, there needs to be a system of trust, so that the data is valid.

5. Data must be interoperable.  Talking about Janet, Abernethy said that even if her disease does return, through her data, Informaticians can learn more about these diseases.  Abernethy noted that the Cancer Biomedical Informatics Grid (CaBIG) failed partially due to a breakdown in communications, since communicating the language and culture of informatics to clinical care at the time was nearly impossible.