Diana Ortiz is a clinical documentation integrity (CDI) subject matter expert at 3M HIS. She is helping to lead healthcare systems across the country to use Z codes accurately to enhance patient care, improve care coordination and referrals, as well as support quality measurements. The Journal of AHIMA caught up with Ortiz to discuss her role and the evolution of Z codes.
Tell me about your role at 3M HIS.
Currently, I’m the leader of the clinical and coding content team, which is made up of all the people who make computer-assisted coding happen. With 3M’s acquisition of M*Modal, our group is a vital component to the revenue cycle equation, as we supply all of the content to clinicians. We have to make sure we give them what they need at the point that makes the most sense in their workflows.
What led you to being a CDI specialist?
I have a background in health information management (HIM), as well as nursing, and have done quite a bit in the CDI space. I’ve been in the CDI space for about 10 years now, which is crazy because it feels like just yesterday that I started. Prior to 3M, I worked at Atlantic Health in New Jersey and Aster Health System in New Orleans, along with various clinical CDI roles and coding roles. This role at 3M combines all of my experience into one.
When did you start getting involved with Z codes?
We’re always looking for things my team can do to help make coding teams more efficient. What’s the next big thing? There is a transformational change around Z codes with the coding guidelines in October 2021 that we thought could open the door in terms of the way we approach computer-assisted coding. In the past, it was really only auto-suggesting codes off of physician documentation. That’s always been the limitation. There are a few caveats to that where they have opened the doors on documentation before, which is kind of what they did with Z codes.
With the new regulations, we saw it as a good time for us to work with a small group of customers to see where they want to see this content. We approached seven to 10 customers to start with and asked, “Do you want this kind of information in a physician type alert? Do you want it to go to a physician? Do you want it as a coding added? Do you see it being auto-suggested? Do you see intelligence served up to your CDI teams?”
We didn’t receive a uniform answer, but the majority were most interested in what we could do from a computer-assisted coding perspective. How can we make it easier for the coder to capture in these ways? We’re building content across workflows because we want to solve this problem holistically, but it also depends on where customers are at in terms of what they’re using today. We have a huge client base with computer-assisted coding, so that’s been our main emphasis.
Why do you think there are such differences in what clients are looking for?
That’s a good question. I think it’s somewhat where they’re at in terms of what solutions they have in place today from a software standpoint. But another part of it is really the culture. Different organizations think differently around what types of things they would put in front of a physician versus a CDI versus a coder, and who is going to take the brunt of the work and how they’re going to adapt that.
And do you, or does 3M, have an ideal or what they think is the best practice they wish all health systems would implement?
Holistically, we would like to see this problem solved across the continuum, across every workflow, and have everybody own a piece of it wherever they can—and then maybe scope out what makes sense to each of those audiences. If you’re truly trying to make the biggest impact, that holistic approach is really going to get you the complete picture. A coder can only code what is actually documented by somebody somewhere in the record. A clinician might be addressing something but not getting it down on paper, but if CDI has the opportunity from an awareness standpoint of where it can come in, in the documentation, they can have broader conversations with nursing, social work, or even the teams that lead patient health questionnaires so it can be pulled from that. I think it could potentially elevate the CDI role to really help own an initiative within an organization.
Why haven’t all organizations implemented these practices? What do you see as the biggest roadblocks?
To some degree in the current state that organizations are in, the question is really what’s in it for them today. For some, it might be where the organization is geographically, whether it’s urban, or rural, for example, that plays a factor. They may have programs that are a bit more dedicated to this in terms of resources for a hospital based on Z code data. That was one of the main questions we asked each pilot group that we were working with: What’s in it for you? Who in your organization is driving this? Because it is all over the place, whether it be population health, quality, finance—and sometimes it moved in the middle of the project as far as who was really owning it. I don’t think I have found an organization that isn’t interested in it, though. It’s just that some have had more at stake for longer, so they’re probably at more of the operationalizing phase versus just awareness and interest.
Can you talk about why it is important to collect social determinants of health (SDOH) data? What kind of things do we expect to see outcomes-wise once we have this data coded in?
I don’t know what the future holds as far as a crystal ball, but the proposed rule for October 2023 will see the expansion of these codes. We don’t see the collection slowing down, which means even if there isn’t direct the Centers for Medicare & Medicaid (CMS) dollars tied to it today, there’s likely a movement in that direction. That’s all the more reason to get it right. Some states have grants in place where there’s money, so that’s important for an organization to operationalize, whether it be food banks or transportation.
Fundamentally, it’s really about how to get patients better care. If we don’t have that social aspect of what it is, that might create barriers to them. We can’t move the needle with an individual patient or within certain populations. Certain events can translate to higher risk, such as constantly being in and out of the ER. How do we break that cycle? To break it, we have to know specific information about the patient in order to refer them to the correct program. It is about transforming care and aligning resources behind it. Have we moved all the way into that, so much so that it looks like an MSD drug reimbursement? No, but there appears to be movement toward that.
Are there particular Z codes that are more important to collect than others?
This question is interesting, because I haven’t seen anything saying that one is more important, but I did ask that of every pilot group that I worked with because I thought it would be easier to focus in on a certain area. Even within where they did have programs in place that were specific to certain Z codes, they still said how important it was to capture all of them because I think they see the long-term picture of it. They didn’t want to sell themselves short with a smaller scope. I will say, within the work that we’ve done where we’ve taken in additional documents that are now able to be coded from (e.g., nursing assessments, social workers), we’ve taken all those notes and we’ve seen improvement in some over others. Maybe it’s easier to get certain clinicians to document certain information over another, but it doesn’t mean any are more important. It just means there are probably some that are easier to move the needle on.
Are there any other challenges you see with these different health systems and in your role specifically of educating them?
The software can take you only so far. Organizations that we worked with where we’ve seen really good success had to have an initiative internally with organizational leaders driving this work. We can start to auto-suggest a code for a coder, but if there isn’t the importance within the organization of looking at the data and an expectation that the coder is going to grab it when the software suggests it, you can’t really impact change. It has to have those two things: people and technology. They go hand in hand, and even more so as you move it up. There has to be some ownership and direction from leadership. The only other thing that comes up a lot is the current limitation around the 25 codes that can be on a claim. These populations often have a lot of issues going on, which lends themselves to being longer than the 25 max. There are initiatives even on the 3M side, working within our government relations, to carve out possibly a separate section for Z codes. Can we find a way to get more of them onto the claim without taking away from the medical conditions that need to be there?
What steps can health care systems take to improve their ability to collect and use this data?
Number one is to get a baseline of where they are today. Then it’s about gaining a better understanding from people who are in the records every day or familiar with the note types, where the biggest opportunities are. What can we do today to help move the needle and then measure that? That’s what we’ve been doing with customers. Maybe you’re in a good spot and you feel like it’s reflective of your population, but chances are there’s probably opportunity for improvement from documentation, coding or both.
Is there anything else that you would want to share?
For clinicians within this space, it’s a pretty passionate area for them because there’s such an opportunity for change within the way health care is delivered. On my team, I have physicians, pharmacists, social workers, etc., who are ecstatic about this movement and the change around it and really want to work on it and work with customers to help identify any way that they can to help. There’s excitement and awareness around this initiative, so hopefully that translates within the industry as a whole.
For more information on Z codes and SDOH from the Journal of AHIMA, check out the following articles: