With the increasing interest in Health Information Exchanges (HIEs), the debate on interoperability is becoming very complicated. Interoperability merely states that computer systems are able to exchange information and use that information in a meaningful way. That already sounds like the meaningful use discussion going on in the US around funding of EMR implementation by healthcare providers. And yes, it is just as complicated. In a very worthwhile contribution David McCallie takes a valiant first stab at simplifying interoperability. However, the ensuing discussion seems to focus on issues around patient-provider communication, where most countries are still struggling with the provider-provider communication. Here’s a first go at simplifying provider access to patient health information.

Three main categories of interoperability are distinguished by David McCallie, based on US legislation:

1. Regulated uses for Treatment, Payment and Operations (TPO)
2. Permitted uses for (mostly anonymous) regulated reporting
3. Meaningful uses which are not (yet) regulated

For each category David identifies types of communication (and hence interoperability requirements) based on the types of communicating parties, e.g. provider to provider within a single healthcare organization, provider to external provider, provider to payer, provider to quality reporting, etc.
In his discussion he advocates a simple solution for provider to external provider communication, i.e. secure messaging to replace the current fax-based communication. However, experience in different countries with this solution shows that it seems to reinforces the current situation: hardly more information is being shared than previously, whereas studies indicate that substantial benefits can be realized by preventing the loss of lives and adverse events caused by lack of information. Therefore, I would suggest a further subdivision of the provider to provider category in the proposed framework.
Provider to provider communication, in many of the plans discussed in regions around the world, aims to make relevant personal health information available to providers when and where they need it in a form that is useful at the point of care. This loosely defined goal leads to the following categories of provider to provider interoperability:

1. Responses to direct requests for information and services
2. Background information for referred or delegated care activities
3. Pertinent patient health information for first contact in emergency care settings
4. Pertinent patient health information for first contact in other care settings
5. Care plans and progress reports for coordination of comorbidity and multidisciplinary care

The HIT community strives to alleviate the burden of communication from already overburdened care professionals. To this end we try to maximize the re-use of data originally intended for other purposes and hope to come up with standard solutions. Focusing on standardized re-use sparks debates about privacy, information protection, trust, etc. As these debates are carried out by the same care professionals we are hoping to serve, we need to take a different road.
For instance, responses to requests (such as lab tests or diagnostic imaging) are likely to be incorporated into the EMR of the doctor. This EMR might incorporate decision support rules that trigger alerts based on the (numeric or logical) outcomes of the requested tests. Therefore, structured and coded data is necessary in this case. With respect to referrals for non-emergency patients, doctors may indicate that a letter providing a complete picture of the background and the reason for referral is preferrable to a coded and structured document within an EMR. Even though lab tests and diagnostic imaging results may be part of the referral letter, their form and function in this category are quite different.
We propose to carefully look into the definition of data sets for the different interoperability categories and striving for professional consensus on the minimum data set needed for that category. Enhancing this definition with clear benefits for the parties involved helps to focus on the need to change rather than the technology to be implemented. This involves looking at the actual relevance of the data in daily practice and considering the ways in which this data could best be made available at the point of care.