Health Transformation

In the last week of May, international standards organizations in healthcare held their meetings in Gothenburg, Sweden. Several times a year, hundreds and hundreds of people working in healthcare, or at least having a keen interest in healthcare, gather together to discuss "semantic interoperability". There are quite a number of doctors and nurses involved, who are fervent contributors to and supporters of semantic interoperability. However, if you ask a random doctor how semantic interoperability is helping his or her patients get better, you are very likely to get startled or puzzled looks. Semantic what?

If you ask the same doctor on possible treatments for influenza or how to handle anaphylaxis in case of peanut allergy, the answers will come directly. This is true wherever you are in the world. If a doctor wouldn't know, many patients would probably suffer or die needlessly. A number of reports and testimonials have been published on the number of patients suffering or dying from a lack of semantic interoperability. So, if semantic interoperability is equally important to the healthcare professional, why is it still unknown territory to the majority?

The key to saving lives with semantic interoperability is to take into account any patient information that is already known and available. This availability of patient information is often the main driver behind intiatives to introduce an Electronic Health Record (EHR) in healthcare organizations, regions, or even countries. However, research shows that availability of information is not a key motivator for healthcare professionals to adopt an EHR. Their motivation lies in saving time and improving medical decision making and that is where semantic interoperability of healthcare information comes into play. If we can clearly show how saving lives translates into making money, many others might embrace the concept of semantic interoperability as well. For now, I'll focus on saving lives and leave the translation to making money for future discussion.

In order to be able to take more information into account in medical decision making, while at the same time reducing the effort to process this information, we need computers to assist the healthcare professional. The vision is to combine the vast amounts of patient information and to present them to the healthcare professionals in a form that is actually helping them to care for the patient. Some of the possible forms are medical alerts, probability distributions for specific diagnoses, suggested treatment options (including personalized medication dosage) according to guidelines, or full 3D rendering of affected parts of the body.

If we are to provide a meaningful presentation of data for the healthcare professional, the available patient information has to be interpreted by the computer system. It has to "understand" the meaning of the information (i.e. the semantics of the data). This is of even more importance when the information is coming from different healthcare organizations (and usually different computer systems). Many lives are lost because vital prior information from previous visits or other professionals was not taken into account when deciding on the proper treatment for the patient. Hence, information from different computer systems need to be shared, and therefore these systems need to work together (i.e. they need to be able to interoperate).

Given the breadth and depth of healthcare information, many years have already gone into specifying the ways in which to meaningfully share this information between computer systems. A lot has been achieved, but much more work still needs to be done. That is why meetings like the one in Gothenburg are so important. However, if we keep talking about semantic interoperability, no doctor will ever spend an extra euro or dollar on systems that support these features. Rather we should talk about meaningful ways to let information help saving lives.