So far, we've been talking about communicating data from one to another point, but as you go up the stack, remember that we have workflow and we have multiple institutions working together. So, the task now becomes how do you orchestrate the interactions that use these other standards of information exchange in data? The leading example of such a workflow management standard is called IHE, Integrating the Healthcare Enterprise. It's no surprise that it started out in radiology. Once they had the success of DICOM being managed in one place, now they needed to deal with an entire institution and they have multiple modalities sharing information multiple ways. So now, it's a many-to-many relationship. There have been many projects using IHE. These are some that are posted on the IHE website. It's kind of fun to think about what are the many-to-many relationships underneath these projects. So, the second one you see, telemedicine in 34 communities. You can just imagine, you have all these patients connected to a whole bunch of doctors, who are connected to a whole bunch of hospitals. So, it's a many-to-many-to-many relationship when there's a conversation between one doctor and one patient over one of those telemedicine lines. Where does that data go? Remember HL7 was point-to-point. Now, we got it to be, that sounds like broadcasting. How do you ensure completeness in all that? If you go further down, public health reporting of immunization. Well, you know when a child gets vaccinated at a public health clinic, it could be that the physician who manages a patient needs to be updated. It could be that the supply chain, people who are supplying the vaccinations, need to be updated. It could be that the immunization information system of the public health agency needs to be updated. Again, a lot of coordination from one place to another. I'm just giving you the data update. What if that immunization required a permission or maybe it needs to check in first before giving an immunization, has the child had that already? So now, it has to go to sites that it knows about. Check whether it has to be authorized to ask a question. It asks the question. There had to be rules about whether you believe the answers or how complete are the answers, which schedule to use. So, there's a lot of data information and knowledge that need to be coordinated in any one of these scenarios. I went through the slide before quickly but I want to point out that the heart of the IHE standard are these integration profiles. So, the rest of the diagram is how one of these profiles gets tested, updated, and implemented. But the heart of it is the profile which says what is the use case like immunizations and then what are the steps for that dance. There's obviously a language within the profile for how do you articulate the dance step in this choreography, which we're not going to go into. But clearly, if you have an interest, you'd want to look in there. No surprise, there's going to be a lot of implicit knowledge about clinical practice in these integration profiles. The knowledge is something that you might say is obvious, like before I ask this clinic whether or not they have a vaccination information on this child, I should be permitted to do so. You know that, I know that, but remember the machine's an idiot and you need to tell that explicitly. So, many of these dance steps are not rocket science, but very crucial to get the whole process of moving along. So, as you go even higher up in the stack of interoperability, we return to ONC and policy. I showed you this picture before. I'm hoping that now that you've gone through the whole process of going down the stack and back up, that you have a new appreciation for the different pieces of this diagram that you appreciate, that the outer circles are functions, that the different stakeholders want out of this thing, that the inner circle where you see sending, receiving, finding, and using data. Those are really kind of low-level functions that any interoperating system needs to support. On the bottom, you see things that we've discussed like data formats. As you now look at this list of policy and technical components to the interoperability plan, I think you'll recognize a mix of above-the-line and below-the-line issues. I bold faced the data issues since we are in the section of data information and knowledge, but it's actually issues across the stack that you are now sensitive to. So with that, we've gone, as I said, down and up the stack. You now know the components of each level of the stack. I think you can identify them. I think you know a bit about how to exchange data at each level. So, I think you'll leave this series of lectures with the basic tools to understand not only what you'll hear in the other courses, but in what people talk about informatics outside of this course altogether.
