Versus the director of the center group. Research. Operations of the law as well as for research research from evidence based and design your facilities. And you to receive a message. You just told me your. You love your. Work life change. To return to your. Horrible college work are there for years for years it was instrumental in starting. To agree. With. Writing and you. Heard from objects including a rigorous literature review. Disruption spittles use Office of project funded by the US Army as you know designed to ensure here spaces for. The most recent research project was a no no nurses perceptions and. Why don't you. Thank you off for coming today so I'm going to talk about behavior mapping in the work that we've done and I'm going to tell that through a couple of different stories so this is all around the research faculty development grant that I received last year and it's going to give you some background about the work that we did before that that led up to that Grant and then also what I think is even more exciting some follow on work that we're able to do now as a result of that grant so as Kathy said I'm the associate director of the some great design lab I also do some work with i Pad the Institute for people in technology working with their center for transforming pediatric health care delivery so we wear lots of hats to. But we have to do. The same to great design lab I'd be remiss if I didn't tell you about that but a large portion of our our audience today is from there so they know all about it but we're over in Texas where on the top floor of the building where parking is and we invite you to come over if you've never seen it there will be a reception on December eighth that all good invites you to with professors leanin Zimring and Cowan their class on health care design of the future so that's a great chance to come maybe get some free snacks and see what we do over in the lab and these are the kind of the the four main things that we do in our lab but I'm not going to belabor that I also want to give special thinks to first the research faculty development grant from the College of Architecture so very much appreciate that it was just that enough to kind of spur us to do a little bit more initial here later bring in another project. To really further this work but I also want to point fingers at a couple of people that help me out that do the technical pieces of this Lisa Leemhuis here Ph D. student working with us and she's been involved in all parts of this project including the slides and will be very instrumental in taking this project to the next stage with developing our new behavior mapping tool that we're going to look at and then also match what's Who is research faculty extraordinaire and who's someone who never says no we can't do that he always says yes we can do that even better than we can do ten different things at once and I'll make it happen in two weeks so we really appreciated some of the work that he did in kind of setting the vision for this I also put them out there with their pictures to go in morning that I'm going to deflect all hard questions to them except that Matt was rising up not not to be here so but then you can find him on campus later. So behavior mapping just what is that so we can all start from the same same footing so it's really it's just a spatial record of information so taking information about what people are doing and overlaying that onto a map so we understand where it happens in space to help us start understanding the how and. The whys of people's behavior. So we have employed three different kinds of behavior mapping and then I'll talk about in the course this project there's basically it has to do with the the perspective of the viewer the person who's recording the information so you can have a fixed perspective where so I could stand here and record everything that happens in this room at regular intervals and it's always from that one perspective where you can have a mobile perspective which we did in one of the projects where if you're trying to look at so you want to see everything that was happening in the library you might create a route walk that standardize route at a set set of ears and then record everything that happens along that so that you capture everything outside of not just in one room and then there's another way of doing it where you actually follow a person around so shadowing so you can see that particular role what they're doing so we did that in one case so we could understand specifically what the provider was doing so we follow them and then record everything that they do as they move through the space. So what is the value behavior map and why would we do it because we have learned at least it is time consuming and can be a little difficult or challenging to but why do it what we think that it gives more it gives richer more nuanced data and it helps us understand as said earlier the Howells and the wise so you can compare it to we can have an excel file which I love I appreciate very much but this data is kind of flat and it doesn't tell us a whole lot about what's happening and we might miss something about about interactions and what's going in the space so we really feel that you get a fuller sense of the activities when you can map that onto a space and then you'll actually see so a lot of the examples that you can layer this with other information that will give you a much deeper understanding of what's happening so this is why we think it's valuable going to give you two examples of other projects that I wasn't involved in but they were done by our students here from Georgia Tech just to give you a flavor of what you can do with behavior mapping So this is a study that was done by. I got her Ph D. here several years ago. And she was looking at the two D. I see you know I see you at Emory that Craig helped help influence and we've got to do a lot of research there so she went around and each of us had black dots is a nurse and she was able to overlay their interactions where she saw them in the one map is just where they were the one on the raft and then the the map on the right actually shows river when they were interacting with each other and so layering that with space and tax which I'm not going to address because John Boehner says here and I wouldn't dare try to do that. But let me get on to that we could actually start understanding as I said the why so where is it that the nurses were choosing to be when they had interactions so she was able to conclude from this that the nurses were positioning themselves in places where they had better visibility of the unit and specifically of the patients in their rooms so now we can start in the standing why they choose different locations and so that could inform future designs where we could build in more of those kinds of spaces. Another example is being done has been done by a current Ph D. student Michelle Osman who has been looking at waiting rooms and how people arrange themselves in waiting rooms so might be similar to how people arrange themselves at a presentation nobody's in the front row. But so she's done behavior mapping they're recording on paper where people sit and been able to see. What's not maybe obvious you know it might be obvious that people leave spaces between between seats or put stuff on chairs and in pairs but what she's going to see in a more recent study that she's done that unfortunately I don't have any images from but she's been able to see that people actually position themselves so they have a view of both the check in desk the reception desk and of the window so you might think that the prime spot is going to be the spot with a window view but because people want access to information they're sitting where they can see both so again this gives us something that might not be intuitive people might not even be able to tell you that's what they're doing and how they're choosing to business in themselves but by doing the behavior mapping she could see that over time. I'm so those are two examples of other people inside the College of Architecture doing. What we were looking at specifically is trying to understand collaboration so I want put this in a little bit of context of why we were doing that so our topic I think it still applies to lots of other non health care clinic spaces the topic that we've been interested in is patients in our medical homes which is it's basically it's an outpatient clinic but it's a new model of care and for our purposes today the most important thing about that really is that it's shifting from being provider centric so it's not a doctor that provides your care but really a team of the doctor the know the medical techs care coordinator So it's that whole team and specifically what's happened is to be accredited as a patient in a medical home you have to have team meetings so you have to have so the spatial implications or you have to have a team space where the team can get together to meet Well the hypothesis from all these organizations is that by having this team space there we increase collaboration and communication and of course the hypothesis is that the hope is that that will also lead to better care but what we wanted to do is look at we realize that there really wasn't any guidance about how to design the team spaces the best we could come up with is from white collar offices from Open Office plans but their goal is different than what we're trying to accomplish in a clinic they don't have a shared goal of fixing one patient right then so we knew that that while that might have things to tell us that it didn't tell us the whole story so we were interested in looking at that and we had an opportunity with an organization called the Cherokee Indian hospital they. Operate highly highly qualified patients in a medical home and they were getting ready to do a new clinic space and so they actually came to us and asked us to study their existing spaces and give them feedback on how they could design best design their new space so we did a project with them I'm going to go into that a little bit just to tell you how we did it what we learned and then that actually led to a couple. Other projects where we've been able to convert find our method so I want to show you our starting point of where we we did this but the challenge here you know our goal was to understand how people were interacting with who they were how was that care team working together so that was our goal we don't have good metrics for collaboration but we can look at these proxies of people talking to each other. So the Cherokee Indian hospital that I talked about that was our first experience doing this back in January twenty fourth all the way back then so what we did for that we actually had a it was a fabulous natural experiment that was just handed to us they had two team rooms that had very different layouts and not only they had these you know within the same organization but they also teams would move from one room to the other throughout the course of the day so we had the opportunity to observe the same team with the same care process in two different settings so I mean this was just we couldn't we couldn't have designed it any better. The only design flaw we could have fixed was that not having snow cop snow pocalypse happen at the same time which interrupted our data collection we had to come home with what we did is we developed the simplified for their plan so you see here plenty of those out and print out one hundred sheets of this we develop a coding schema we pilot tested it we practiced it we trained everybody so would have. Into rate or reliability and we went on site and we positioned ourselves tried to be as an accurate as possible position ourself in the team room where we could do this fixed fixed up Servatius method that I talked about earlier so we did a sweep of the room every five minutes recorded where everybody was we were able to actually record the type of person it is so doctor or nurse the direction they were facing their posture where they standing or sitting where they interacting with any devices or are they using a laptop or talking on a phone and then if they were talking to anybody and that was actually really important right that's what we wanted to see so we were able to take. A look at layer would be able to take to. Data points and say these people were talking to each other so we could really kind of code in whatever information we thought was relevant. We came back to the lab rushing back to the lab to not get stuck in the snow and after we'd all got back to the office we realized that this we had this great data but we didn't really have an easy way to process the data and we hadn't had only thought through what we're going to do with this right so we have all these sheets of Danas like this is great and I can you know we can sit together as a team and say this is what I think I saw and what I think was happening but we couldn't share that with the client we couldn't come up with anything conclusive so we realized we needed to be out to put this data on the map and show people this information so we came up with a system we put you know put a grid overlay the grid on top of those data collection sheets we recorded everything in Excel we were fortunate that at the time we had some undergraduates working with us because this was a very time intensive process but we we figured out a way to convert this this paper data into something that we could use digitally. But then of course we had this great excel file and now what do we do so this is where Matt Schwartz comes in and we were extremely fortunate as I indicated earlier mache works as a yes man and he says Sure I can do the impossible and we'll create this so over the course of two weeks he created our software we called it snow because of the snow was very much on our bodies mind if you were January twenty fourth and plus it's you know these patterns of snow on our on our maps so he was able to create his visualisation to where he could take all of that data with all the rich variability the rich data associated with each data point write about who it is and what they were doing and overlay that onto these flare plans and visualize it and filter that data so we came up with these great graphics the works really helpful for us understanding what was happening but also to communicate that to people so they were there on the top shows with different people all of the different the different colors or different red also we can kind of start seeing how people. Distributed themselves in the space and you could see that without having to use any language that having to convince people there on the bottom we used to share where communication was happening so those patterns of communication and one caller I can't repeat want to think the blue is providers in the red is medical assistance you could start seeing and that's what's going to be on the team right so if we want to measure team communication we want to see communication across those different roles so we can actually visualize that and see where was that happening where were people when they were talking to each other. And then the next opportunity came we finished that project we were very happy with it we thought we'd love to do more of this well we happened to get a grant from the Academy of architecture for Health Foundation in conjunction with Herman Miller to study patients our medical homes to do case studies of patients or medical homes so in the process of doing that I called up mercy care and said Would you you know could we do a case study we understand you're really good patients are medical home and are just another great you know just meant to be kind of thing that fell in our lap she said well you know we do run a great patients are medical but not a team room so really a bad case to have you know you're a great case study as a researcher I'd love to see that difference she said but we also we have a grant in place we're going to get to renovate and develop a team room so that. This is can be great not only could we see you know we could use you as a wood is how do you do patients that are medical home without the space but gosh if we could find funding to do observations we could do a before and after so that kind of all our labs and then in the research faculty development brain opportunity came out I said aha this is the one we can do this and we found out we learned about a tool for behavior around being a digital tool which would simplify and answer a lot of the problems we had so we thought great this is what we're going to do we're going to we're going to before and after more secure we're going to see even more what does what matters about space so they've been a great partner of ours and we're going to be presenting actually. With Lisa Ling at the health care Design conference next month about the work that we're doing with Mercy care. So. So now this is the work this is the actual body of the research faculty development grant work what we did is we use of a called the DOT tool for detailed observation to him which is. It's a behavior mapping troll that they created but B B H architects in Raleigh created and so we license that from them they took the floor plans of the clinic that we had they digitized that put it on this ma'am this was on it this is the screen of an i Pad and then we had these data fields at the top we were able to customize what we want to collect based on who was there in the you know who the rooms were what we wanted to collect the devices they were using interactions so we actually created two we created one to do this mobile observation where we walk through the clinic and another to a slight variant on it to do shadowing so we could see what the providers were doing because again they didn't have a team room so we couldn't do like we had done it Cherokee where we stood in the room and watched everything that happened because this was all happening out in space. So this is their for their plan of Cherokee Indian hospital and is just a show this is the route that we took so we would start from the mapping we would start at the front entrance of the clinic and we would rock the quarters and we had it you know we had a very systemized systematized would stop at certain places with a certain perspective scan the environment record who was there we had coding scheme. This involves lots of prework testing this tool while it also involved just looking at lots of pictures of the provider so we could memorize who was who because we didn't have an easy way so we could get all this right so a lot of pretty work went into these days of a shout of data collection and we were able to collect over twelve hundred data points on behavior mapping nine hundred points foreshadowings we followed them providers as I indicated so we looked at this but what you can see just right. We have the home I up appears that the provided offices up at the top in the back and it's actually behind a couple of doorways to get in there and the nurses were up in the front so intuitively you can say there might not be a lot of collaboration because there's not colocated. And in fact that's what we found is that there wasn't a lot of collaboration but. What else I was here is that we were able to demonstrate this in a way that's a lot more compelling than just saying you know we're not located together supine not talking a whole lot we can actually quantify it and show them what we observed over the course of those days so what we saw is that doctors are talking to doctors and medical systems or talking to medical assistance and then we didn't record anything about what the content of that conversation is so it could be very useful information but we know that a care team is a doctor or a nurse practitioner and a medical assistant so if the care teams are collaborating they need to be talking in fact after being there for a couple of days I finally said how do you know what's I don't even know how you don't know what's going on or who's in a room and they said all religion the medical records we're seeing in the medical records but so there really wasn't a lot of opportunities for them to interact and to speak face to face that we could we could actually see that through the day that we collected and then we could even map that put that on the floor plan to hear we can see we know who they're talking to but now we know where that's happening so not surprisingly as you might guess the doctors conversations the providers are having the conversations back in the provider office and the nurses are having their say up in the nurse so you know it's it's not surprising that this helps reinforce that the architecture was driving a lot of this this behavior and that that's why these conversations were kind of isolated into these different roles. So the high point of doing this was to figure out Well one of the main points the point for the the Grand. Right was to figure out whether or not this doctor was going to work for us and went on This was the key to solving all of our behavior mapping challenges so it was very helpful it did vastly improve the ability of taking taking observation data and putting it in spatial context right so you know the moment we you put your finger on that screen you get these X. Y. coordinates that locates it in space so that was already there we no longer had the challenge of the hundred pages of paper that we have to lay the grid on we already have this information so that was really good they were able to customize it for us they were able to put our floor plan in it they were able to change the. The rows we put we were able to do a few other customers ations to get the kind of information that we wanted so that was great. Because there were some challenges in being Georgia Tech we think we can do a better. So run so they modified it for us which was really good that they were able to do that but that process was us communicating with folks at B.B. a truck attacks waiting for them to make a modification it coming back to us just trying it out so Roland had flexibility it doesn't have the kind of rapid flexibility that we sometimes need for our projects so had we got on site and started testing and found we needed some other kind of modification if we weren't able to do that. They not only did that scale. We and to get the best value out of out of this tool in the data that we collect we need to be able to customize it for every project and for every research question so we want to that we have more control over and that we can change to suit each need and not having to go back and kind of go through the process that has all these risks of of data failures so that was one challenge it also the data collection itself wasn't flexible. Of course that's a trade off it was easier to do and faster because we could just record it on on the i Pad for that we couldn't do as we couldn't make notes like you get down. On the screen here so if something odd was going on or if we saw somebody in the space that wasn't part of our coding scheme. We couldn't record it or we couldn't you know say a group of school kids taking a tour that's why we see different behavior right so there was no way to put that would have to write it on his paper and know that make that mental note later in the in the records and modified the biggest limitation though for our use for looking at collaboration was that we couldn't associate data points so when we did our just hand scratch records write we could say these two people are talking with each other that's who they were talking to in the dot tool and we could do is we could say this that this this point is a doctor and this point is talking to a nurse we can't tell you which nurse that is then we could put another point which is this is a nurse this is a nurse this is a nurse and we could take one of those be talking to the doctor but we wouldn't know rich one was the one talking to that doctor so we lose some of that that spatial information and particularly around around collaboration so that was a real challenge for us. You know you really have to make some some suppositions at that point about who's doing what kind of collaboration so we really missed that point and wanted to be able to just visually connect those dots. It also wasn't able to make the kind we weren't able to filter the data and make the kind of. Associations and representations that we want and we want to be able to explore the data a little bit more freely without necessarily having all of our hypotheses fleshed out from the beginning so these are the outputs that Lisa was able to generate from the dot tool directly right so we have on the left is the flare plan with the location of each person and they are coded by role but that's for the whole data collection period so now we know OK that's where people were located when we happened to walk by and record them the one on the right is from the behavior of the shadowing where we followed a provider around so we can kind of see some power. And what's happening there and we have a nice pie chart which tells us the devices or activities what they were doing. But I don't know who was doing which who who was on the computer and where were they on the computer you know there's all these pieces of information that we might want to cross in different ways that we couldn't do with the interface that they provided us so it was. It was nice to be instantly create these but we actually ended up creating our own going back and an expert in the data I'll show you that a minute so these are the custom graphics that Lisa created by exploiting the data into Excel and then using G.I.S. to put this on here so here is where she was able to say these are the grass we looked at earlier I think about where people are having the conversations so now you can pull out just the providers and say ma'am just the providers rather talking and so we can see that on one map or map just the sort of five medical systems while they're talking on the other map and we did lots of different ways right to explore the data and see what are these patterns we can sort of recognize what's happening and what's meaningful to show to the client to the user so that they can understand these patterns. Well there are lots of tools out there we are aware of that are probably not aware of all of them but we know that there's a lot of opportunities with sensors and bedded sensors using our if ID in the floor there are putting cameras up lots of ways that we can collect behavior mapping we anticipate being able to integrate some of these tools and what we're doing but we don't think they really solve our problem or the questions that we're trying to answer because the kind of research we do will be going around a lot of different places will constantly be changing we're not going to be studying one place over the course of a year so we want to be out to go in and you can't embed sensors in our if I do if you're just going to be there for a week. We want to understand the context of that day so we really we feel like we as the observer is still really important but we also. To make that easier easier to record that information quickly and also then create graphics and visualisations out of it but so that's not so we won't use these tools but it's because I do want to hear from other people how they've done this and what they suggest but. We don't see these as answering the questions this isn't going to give us the level of data in the kind of data that we need at this point. So here's the product that I think is the exciting what we get to do next so that we leverage the ten thousand dollars research faculty development into a half a million dollars in military health system so not directly but we do we did just get a project in from the military health system from the Army to study their outpatient clinics their patients are in medical homes and one part of that project is going to be doing case studies of six facilities so we're going to go into site visits of six of these outpatient clinics and we're going to be looking at their space will be recording a lot about the space and doing other kinds of analyses but we also want to do this behavior mapping so that we can understand we can kind of abstract from that what is there care process and how does it vary across these different spaces and make recommendations for them so we saw this is a great opportunity while we're still deeply immersed in analyzing the data from Mercy care and preparing to go into our next round of observations of more secure the after observations we thought we can collect we can create a better tool so we've actually been able to put funding in our project to collaborate with the Interactive Media Technology Center here at Tech and we are going to be creating the own we don't know what's called yet but the Georgia Tech sim to great behavior mapping tool and I really want this to be useful beyond just patients in our medical home clinics we want this to be useful for other kinds of questions so that's why I really want to through this form invite everybody if anyone's really interested in this topic you know if you can help us frame what those functionalities are we want to create a tool. It's going to serve multiple projects and that we can then create databases that we can use for years to come because right now what we do is we get this information from Cherokee Indian hospital and it's in this you know bespoke SAP where we can't ever use that again now we've got this other stuff in well that was a student and it's gone away on those paper files that got recycled so we can start creating structure databases where we have this this data collected in similar formats that we can use across across case studies. That would be really valuable but we also want to do something integrate that with the visualization and managing of that data right so to address the question So when you know Lisa got this output from the dot to lament I can't create the graphics that I want to create these graphics don't really mean much to us we want to dissipate that up front and create create tools that help us visualize it but also allow integration with other software packages such as space syntax because that's a really important one for us to be able to to layer that information and add in other data as well that we might collect noise levels at different spots lighting views things that would help us describe the experience of a person following that path so we're going to be doing this over we're actually going to be doing the tool over the next three months. Bringing out what we want to tool to be so this is just our first cut what the desired functionality is for the tool that we want to create in this where I would really sit solicit feedback on what are some you know if there's a question we haven't thought about that if we could just build that functionality and would like to do that now versus find out about it later so we're kind of envisioning this as being to serving two purposes the data collection so giving us that tool in the field that we can collect that data but then thinking about the what we do afterwards to the managing and visualization of the data so you know this is to have it in a consistent and give us flexibility so that we can modify what we collect on individual project basis and then to be. An intensive paid activity that happens there so one to even out in the field how do we record the context but then providing us the tools to to to more easily create the visualizations that we want to filter and organize the data in ways that are meaningful for that particular research question and then to generate actually generate these visualizations which are in a day they really are much more powerful than trying to describe to someone Well here's what's happening in your clinic be able to show that representation of what's happening. And that's it I would love to hear from him he's got any ideas or thoughts. Quickly to up there with me because she's actually going to be leading up the project for developing this tool it's her her idea that we get a new tool so if you have any ideas later about what what you'd like to see in that. Tell her because I wouldn't understand anyway. Just just. Where your hands are likely to. Be not just hearsay. Says. I think there's. Yeah I think this is very. Fine what. Is. Here's. Your. Sign. Well. We want to cover all of that of course well so Michelle Osman who's who's a Ph D. candidate here is working works for STILL case now and she actually used the DOT tool as well so we will be pressing upon her I know she had some. Feedback about the tool and in ways in which should want to see it work so we're definitely recruit her to be on the team but maybe we can expand that to thinking larger great. Yes definitely and that and that's been the challenge right is that I mean we really want to be able to say we want to have a say that the design of the clinic improves patient outcomes I mean that's the big right but we can't really say that so we're doing is going to milling that field and saying there's already an assumption in the medical field that communication and collaboration. Is going to improve patient outcomes right so they're saying by betting on patients or medical homes and by saying that they need to improve the team communication that that is going to in and of itself result in better outcomes so we can get a mirror and say when can we show that we've improved communication and collaboration and yet ultimately we will have to then connect that to and I'll probably be a big data question right it is one of those outcomes that we're looking at can we look at improved Population Health for that clinic and can we relate that back to communication and collaboration but we're also looking at are there other proxies of communication collaboration that we could measure that don't necessarily you know because some of us happening electronically so how can we measure how well the team how cohesive the team feels or how much of. Mation they share. Yeah. Right I mean certainly it's I mean it's definitely not but it's not the sum total of what collaboration is so some of the kind of add ons that we envision being able to do to get a little bit closer at getting and measuring collaboration can we start to measure and Lisa's look to some tools this Sociometric badge that can kind of measure conversation can we measure the length of the conversations could we measure who initiate the conversation right so we can start recording that information so maybe there's something in there about if hierarchically if the person the notice or the medtech initiate conversation more often in one space that are not or can that start to be kind of a an indicator for increased collaboration obviously if we knew what the content of the communication was that might help but that it's just too much privacy data but but so those are some of the kinds of things like how much back and forth is there what what's the tone. But are there ways that we could just start even measuring you know how people feel we did a survey we did a simulation in our lab to look at two different teams spaces and we developed a survey where we asked them about the space but we didn't say you know do you like this place or that space but we ask things like does this space allow you to find your teammates when you need them with this space allow you to observe the patient rooms to know where people were so trying again to get at that it's not directly addressing collaboration but trying to to make them start visual thinking of the spaces how does it help me connect with my. With the rest my team. And obviously that's not just driven by space it's OK. So we did not we did not do that so we only recorded rolls. Yeah. Right. So. You know we're right and of course we always we approached it with like the minimum amount of information we needed to get agreement to go in there and we we did mostly care so we always stopped at the patient door we didn't go into patient rooms into things you know I think they would be OK with it after we had been there for a while I mean obviously there were skeptical about us being in the clinic and were going to be disruptive So imagine that might be something that we could work with a team to get. And maybe it was a revert you know some kind of coding scheme where it's doctor one doctor to doctor three and then you know that Dr Ronn and and C M A five are on a care team together but it would be useful to have kind of you had to to at least code them by CARE teams. Right. I think so what we have and. I can't write Yeah I think the yeah they might not be happy about it but. I think we could do it. From. Just. You know. But it would certainly give us you know another layer to look at. Because I think a lot of the differences that we saw work were definitely personality driven right so some of you can create a collaborative space but if if a doctor has got up hierarchical mindset and doesn't want to collaborate that are not going to do it regardless what space but it means you can start seeing kind of you how much how much influence can the space have how much can the space node people to collaborate versus over there and you know there's no he's. Right and. This is. You know what I meant was. You know you. Yeah. You. Know. What's. Right. And that's almost up to some of the work I've seen done with videography with and fixed spaces where they're able to kind of see you know you can see how fast people are walking and maybe start a little bit understanding light of what slows people down as it when they get this close to somebody or is it at these openings in the hallway that people tend to be more receptive to having communications. Yeah. Yes we're doing it yeah. Right. Right. You know what I'm thinking what can we do with this police body cam data that data set exists right to me there with that idea map all the Krispy Kreme in Atlanta or something. To a Dunkin Donuts map those own. Eyes. Yeah you know I would say and I don't know that this was all from behavior mapping as much is doing so we had floor plans from five different clinics and did analysis of those as well so the things that we saw in this might not be collaboration and office right so this may be only applies in this but having visibility seem to be a real big big driver of. Being able to see where people are it in a given time we also found that having the ability to be able to balance that that openness and his ability to balance it with focus and heads down to kind of be invisible for a moment to do their concentrated work so that was a real delicate balance. Having so other things so to particular in mercy care what we saw always having that open space was great that they could collaborate with each other but they also had to create some kind of delineate a boundary or barrier between them and the patients so they had to keep that in mind that they wanted to keep patients out of their space so a way to say you know this is this is. My Space but to create that that I contact with with their coworkers so in the Cherokee Indian hospital we saw they had one room where people were sitting up against the wall and facing the wall so they basically had their backs to each other all the time around the U. shaped thing and in the other team room they had peninsulas where they were across from each other it wasn't us or the team the team was still maybe kind of had their backs to each other in this little you but they could see other people and we noticed a lot more interaction in that where you could see people and it just gives you that opportunity to say wait I have a question that you maybe can answer and so yeah so that's I would say that anything else that I I'm sure we had other recommendations. If you want to add anything Lisa. That. Was. One of the. Ones or. So. And you might even get to use it on your own project right we will share with our Georgia Tech friends. Yeah. I mean I think it applies it absolutely applies in the health clinic to there was a lot going on with the electronic health record so that's that's another layer of information that we could add to it that I think would be very useful to see you know because it more secular we were just seeing you know we were seeing a lot of not communicating but we know that there is something happening right on the back in so if we could going to see how about in even looking at seeing what kinds of interactions are happening electronically and what kinds of interactions are happening face to face I mean again the hypothesis was that these informal interactions these happy haphazard interactions were actually creating higher quality. Yeah. Yeah I mean it would be great to see how it how it changes that too right so when they have that information because a lot of places were tell you we're talking these different clinics they would say well how do you know where the doctor is well I can look on the child or I know somebody's been room and then in another clinic at Emory they were just I can see everything so even though they had an electronic medical record and they could see it they had higher quality data available to them more immediately just by looking up so but you know I think it would be important to look at both of those interests to do an analysis of what kind of. If that change we're not doing it currently but I mean with electronic medical records that data could be there and could be added in. So yeah I think it's a great point.