A guide to Fast Healthcare Interoperability Resources (FHIR)

Nov. 28, 2016 / By Tosh Kartchner, RN

This fall’s HL7 conference was full of very informative learning sessions about Fast Healthcare Interoperability Resources (FHIR).  Presentations covered what FHIR is, how it can be implemented, whom it can benefit and details on the specification documentation. Learning these objectives can help individuals in the industry get a leg up on interoperability, analytics, decision support and other goals/needs for FHIR implementation.

One high level goal of FHIR is to meet the needs of the top 80 percent of content that is being used in the industry. Also high on the agenda is to make sure that processes make use of today’s forefront technology to provide a paradigm- and architecture-agnostic reference that is publicly available and easily understood. If these resources are not created in a way that makes them easy to put in place and clearly understood, the adoption rate will suffer. This is where the simple structure of resources becomes valuable. A resource is a small piece of the medical record that gives just enough information to provide value, but not so much that the main point of the resource is lost. Examples of resources include a procedure, patient condition or even a medication allergy. The building blocks of a resource are referred to as elements. These elements can range from “Address” to “Patient” to “Date.” In addition, resources can be contained in other resources. When a resource is finalized, it provides all the possible elements for a profile (more on that soon) that an end user implements in their electronic health system.

Resources are the starting point for covering the most used content in the industry. To get the benefit of FHIR, users create profiles. This is done by taking a resource and its elements, then further extending or constraining them for specific needs. For example, when implementing the AllergyIntolerance resource, one might choose to disallow the ability to have a value for the comment element. In this example, the profile is not the same as the base resource. The profile and its elements are clearly defined making it easily consumable and understood by the sending and receiving systems. Another example is to bind an element to a value set to define what values are applicable for the element. The value set that the profile indicates does not need to match that of the resource, as long as the value set is clearly indicated and retrievable. This again provides flexibility to meet the need of specific users while leaving the defined structure for clear understanding.

Extensions are an important feature to understand in order to help entities capture the desired information. Extensions provide a standardized mechanism to place additional data in a resource. Using profiles and extensions, the sender and recipient of FHIR resources can exchange data in an unambiguous format.  

Value sets are used to capture the set of encoded concepts allowed to populate certain elements in the profile. Value sets provide a distinct list of what is expected for the given element. With this defined list, the recipient knows that the information being sent is valid. An example of a FHIR value set is “substance-code” which contains a list of encoded substances a patient may have an allergic reaction to. You can find this value set used in the AllergyIntolerance resource, and within this resource, it indicates the substance that the patient is intolerant to. Since value sets group coded concepts, it is important to have the standard terminologies that are used in the industry so interpretation and validation of the concept from the value set is achievable. The 3M Healthcare Data dictionary and its subject matter experts continue their efforts to provide these terminologies along with the structured value sets and APIs to support the adoption of FHIR.

In short, FHIR is meant to allow flexibility for ease of implementation, while at the same time defining the structure in a way that is easily understood and consumed. We are implementing various FHIR value sets and the FHIR terminology API, along with implementation and education services for FHIR adoption. Learn more about FHIR here.

Tosh Kartchner is a registered nurse working in medical informatics at 3M Health Information Systems.