In the coming weeks and months the importance of metadata in healthcare will start to evolve and we will see an increasing amount of discussion and product development start to occur. What is metadata and why will it be so important? That is the point of this post.
Metadata is easily defined as data about data ( Wikipedia ). Metadata describes certain attributes of data ranging from a documentís creation date to diagnostic results. In many cases, metadata can be utilized to provide more information about a particular piece of information.
In healthcare, the explosion of data is dwarfed by the explosion of metadata, however providers are not yet utilizing metadata to its fullest. In fact, many organizations are struggling just to manage the data they have today without considering how to manage the metadata aspects. Once we capitalize on the usefulness of metadata, we will be able to better manage the data that is being generated. After all, we are creating all of this data and metadata, why shouldn’t we use it in our daily decisions?
Let’s look at a simple example of how metadata can help us simplify our lives. Apple’s iTunes is based on the utilization of metadata to drive user friendliness with an ulterior motive of generating new business by being “helpful”. When you launch iTunes, the application generates a list of artists, songs and albums from your library.
The application utilizes metadata to sort the data according to how you have configured your preferences (alphabetical song title, genre…). Now when you go to the iTunes store the application utilizes metadata to present a number of items for your perusal. You can see the top songs, albums and videos based on sales — metadata that ranks the importance of the particular category. When you select one of the top songs, the application takes you to the album and it shows you songs that other people also purchased.
These recommendations are generated by metadata that has been collected by the application when others have purchased the song you are looking at. In fact, on this single page, Apple helps you understand what the top songs of this particular artist are — what they consider to be iTunes Essentials, the Top Rated Playlists with your selected artist and customer reviews. Apple then takes it a step further when you turn on the Genius playlist.
Metadata filters kick in to present a playlist of songs that are similar to what you are listening to. And now iTunes has added Ping , the iTunes Social Network to the mix. Talk about metadata explosion!
So how is this relevant to healthcare? Let’s use radiology as an example. When a patient has some diagnostic imaging tests performed, the results are stored in the local Radiology Information System. The report is then made available to the Hospital Information System as part of the patientís record. The results of that report could potentially be stored two or more times in different systems. Associated with that diagnostic test is metadata about the image (pixel information), about the patient (name, medical record number…), about the exam itself (clinical findings) and general metadata (date of creation, exam type…).
The amount of metadata generated associated with the exam is largely ignored. Now, what if the information systems within the hospital could access this metadata directly?
Instead of having multiple copies of the same report, the system could identify through the metadata the patient’s exam, location of where it is stored, clinical results and more and display all of this data in a single view. Add the “genius” algorithm and the information system also could display other relevant tests, results that the clinician may want to see in addition to the diagnostic images.
Think chronic disease management. An intelligent system that utilizes metadata to its fullest could pull together all relevant data that reflects the state of a congestive heart failure patient — diagnostic images, blood work, ECG results and medications associated with the patient.
Metadata could be utilized as part of a decision support tool whereby the metadata could identify contradicting medications, allergies and other important factors that would affect the patient’s care.
But aren’t there systems out there that already do this? Many of the systems that are out there today are capturing the medication information but they are not tied in with other aspects of the patient’s record. They are solely focused on the medications that the patient is taking and do not reflect the other potential disease states that a patient may have. Aspirin may conflict with the blood thinner, but without the aspirin the patient is prone to chronic headaches. Without that knowledge the physician puts the patient at risk.
How can we manage metadata from all the different systems and provide better patient care? The answer lies in the creation of a metadata repository . A metadata repository indexes all of the metadata from the different documents and information systems and consolidates the data into a single repository. From there, it integrates with the Electronic Health Record to provide context around the metadata. Queries against a patient name will resolve all relevant data for that patient. By applying filters and algorithms against the metadata repository, the “genius” EHR can then display the previously mentioned relevant patient data. We have just taken patient data and made it more relevant in the care of the patient.
Utilizing this metadata to make better clinical decisions will ultimately benefit patients and healthcare providers in the long run but it is going to take some level of investment and development to realize the full benefits. Chronic disease management systems today are trying to accomplish similar results by tying together relevant data points. A metadata repository is the backbone of a knowledge management solution .
In my next post, I will talk more about this knowledge management solution.