The Problem with Interoperability

The one topic on which thousands of experts working on electronic health record (EHR) are focused is interoperability. This is no surprise. Since the 1960s, visionaries have declared that in the era of medical specialization, an information system is needed that replaces today's information silos. Indeed, the patient information you have in your office is such an information silo. Let's face it: You and other clinicians sometimes have to provide patient care without knowing what has been done previously and by whom, resulting both in wasteful duplication and in clinical decisions that do not take into account critical data related to patient health. We know that is a fact of life. Physicians were trained and have adapted to "practicing medicine blindly."

Then it was proposed that there is need for physicians to know what their colleagues are finding, prescribing, ordering, or considering regarding a specific patient. The idea of electronic medical records (EMRs) was born. The vision included an EHR system that allows full interoperability between all practitioners involved in a person’s health care. The problem is that most people didn't understand what would be involved in full interoperability. Many focused on system interoperability, hoping that continuity of care could be achieved. If one computer system could talk to and understand another, much would be achieved.

Reality is different. Even if in the foreseeable future (for billions of dollars) we can see to it that all computer systems in health care are able to communicate with each other, we will be a long way from the goal of interoperability. Just think of good old paper charts. In a rudimentary way, they are interoperable. If you can send or fax a complete chart to a colleague, you have basic system interoperability. However, have you looked recently at, let's say, a 212-page chart created in another clinic? It may take you hours to find the information you need. Even if you are just looking for allergies; they could be practically anywhere in the record.

It is obvious that, besides system interoperability, other issues need to be considered. The arrangement of information, for instance, is important. In your practice does every physician insist on his own template or format? What we really need is a universal "table of contents" for patient charts that enables you to find specific information within seconds. The key here is semantic interoperability. This includes the meaning of language and the need for universal nomenclatures and vocabularies. Add to this the problems with the overwhelming volume of data in a patient chart. Clearly, a navigation system is needed within EMR systems that allows easy access to the most timely and relevant information.

For anyone who is watching the developments in this field, it has become clear that full interoperability will not be achieved for some time to come. The problem is so huge that the best experts cannot tell whether this will take 5 or 25 years. However, there are some shortcuts that are under serious consideration by the medical informatics community.

First, priority should be given to the separation of relevant and irrelevant information. When you look at a chronic patient's chart, how can you identify the relevant information? The continuity of care record (CCR) provides all the clinical information and can be used as a summary, not of just an encounter or episode of care, but of a person's most relevant health status and health care treatment information, current and historical. Second, there should be a separation of management data from patient health data. In other words, it is not necessary to share everything about the patient's care and related management issues, but it is important to focus on the relevant medications, findings, or other care decision- making issues. The move toward interoperability is a journey that will affect everyone in health care. It will change your particular ways of documentation and of practicing medicine.