In my last blog I wrote about high-level best practices when attempting an integration between an EHR (Electronic Health Records system) and a HIE (Health Information Exchange) which can be found here.
In this blog post, I will discuss the first technical steps of the integration that need to be considered and will provide some resources that I found helpful to aid you in your project as you get started.
If you are just starting your integration (and haven’t done one before), you may feel absolutely overwhelmed at the scope, vastness and complexity of the medical industry – and the data exchange efforts therein – that are underway. Where do you start? What do you have to do to your EHR system to prepare? What are a few resources available to help you along the way? I’ll answer these questions and more.
Where do I start?
You start with the requirements of your project; these requirements have either been given to you, or will become self-evident as you meet with HIE representatives and discover what capabilities they offer. To guide you in your next steps, you’ll need clear answers to the following:
- What data needs to be exchanged? For example, do you need to exchange an ADT (Admissions, Discharge and Transfer) message, a CCD (Continuity of Care Document) document, or an immunization message?
- Will the content you send or receive need to contain standardized values for such things as medications, problems or even lists of races and ethnicities?
- Has a data exchange methodology already been prescribed? For example, do you need bi-directional exchange using XCA profiles; a one way feed using a web service to deliver the message, or sending a secured email?
Let’s start by exploring the first two questions.
What data needs to be exchanged and why is this important?
The type of message (ADT, CCD, Immunizations, DICOM, etc.) that you need to exchange (and its contents) greatly affects how you proceed in your next steps. I have found that every piece of formal data content type has a history of evolution behind its purpose and creation. And each type has been curated by different entities, at different times in the past, and reached different levels of maturity and acceptance in the industry. Not to mention that each one has a different standard in how its data should be formatted and organized. For example, ADT messages have been around a lot longer than messages containing CCD documents. ADT messages use the HL7 formatting standard for their message content instead of the CDA R2 specification and ADTs may also use different code sets (explained later) to describe standard lists of items (like a persons’ gender).
Here are a few links that helped me get started. Note: for HL7.org links, you will need to get a login to download the actual standards. Be sure to check the state you are working in or HIE you are working with to see if they have supplemented the federal standards with their own specific modifications to those standards.
These links listed above are just a few of the different message types that can be exchanged.
Why do I need to care about data standards, and what are code sets?
We need standards of all kinds to effectively communicate between systems; otherwise it would be extremely difficult to make a single system exchange data with many other systems if each system used its own “vocabulary”. Similar to people in different countries – it certainly would be easier if we all spoke the same language. For a simple example illustrating the problem:
System A has defined patient gender lists to include (M=Male, F=Female, U=Unknown)
System B has defined patient gender lists to include (M=Male, F=Female, A=Ambiguous, O=Other, U=Unknown, N=Not Applicable)
If your system is System A, how will you support all the possible values for System B if that is the standard set forth by the industry? How will you convert every possible value in your EHR system to one of the officially sanctioned values?
“Code set” is a term you will hear often and will need to understand. These are basically the groupings of standardized lists of terms. There are hundreds of these to govern all manors of data from gender, ethnicity, race, and marriage status to medications, problems, allergies, and many more.
Starting to understand what standards you need to follow and beginning to make changes to your EHR software should be one of your most important priorities. You should start this conversion effort immediately in your EHR if it doesn’t already conform to these standards. Retrofitting your EHR software could take an immense amount of time and resources. Other possible options include using middle tier software to do the conversions for you “on the fly” so you don’t have to change your underlying EHR data.