 day-to-day. We have three great presentations, two use cases, and then the first presentation, as you can see on the screen here, is from Morton Hansen. He is our lead integration and interoperability engineer here at the University of Oslo. After Morton's presentation, we have Laurentias and Dr. Cinsali from Laos presenting on the integration with M-Supply, and then after them, we'll have a short five-minute break, and then we will be joined by Nuno Reinho and Laundrie Medigan from Medexis on their integration interoperability with DHIS2 in Burundi. So today is jam-packed full of really interesting interoperability and integration use cases and lessons learned. So we're excited to be able to present all these to you today. Without any further ado, Morton, I will go ahead and hand it over to you and you can take us away. Sure. Welcome, everyone. So as Scott was saying, I'm Morton. I'm the lead of the DHIS2 integration team. I'm also working on the platform team and all the APIs and everything around there. I was thinking probably, yeah, we'll do about 20 minutes presentation, and then we have about five minutes at the end for some questions. So if you have anything, please wait. In the last five minutes, we can have a bit of question time. Just want to quickly go over the integration team itself. It's been around for a little while now, but there hasn't really been maybe too public. I don't think everybody knows about it. I think some of the people was mentioning this. It's the first time they've seen this email address. We have an official email address now if you have any kind of integration questions, API questions, and so on. The team is currently just me and Bob Joliv, which probably most of you have heard about before. We do have weekly calls every Friday. If you have something you want to kind of discuss with us in more detail, you can join those meetings. Just send an email to us and we can invite you. We are more and more getting part of the open HIE community, which is a community for creating kind of open health standards. And within that, we are part of the HMIS working group, the subgroup. We are also part of the facility research group. From time to time, we also join the COVID group. So if you're interested in that, just please feel free to reach out to us. Anyone can join these open meetings. They are quite open. So if you have anything you want to bring up, feel free to do that. We are also working on different kinds of emerging standards that are linked to this open HIE project, stuff like MCSD, which is for exchange of organization units. And it's basically a profile of what's called FIRE. FIRE is a standard for doing health inter-building. And MCSD is in the profile of that linked to doing org unit exchange, basically. And then we also have SCCM, which is the same kind of profiling of FIRE, but this time it is tailored towards option sets or codists or value sets or whatever you want to call it. And then the last one is EDX, which is for aggregate data exchange. So that's basically a standard to do exchange of aggregate data. So in addition to, out of addition to, and that's already been used in a few places. Right. So again, just very, very quick overview of what we have. So these are two APIs. So basically everything you see inside of this just to today, the way kind of the DHS2 apps work are using the same APIs as you will be using if you are going to do any kind of integration with DHS2. So for example, if you go to the data entry app, the underlying, the app itself is actually using the same APIs as you would do if you want to do, say, a bulk import of data values to DHS2. So we are very much using the same APIs as you would do for integration. We do have kind of a three separate APIs for data. So obviously it's the aggregate one, which probably is the one that most people are maybe already familiar with. There's a tracker one for creating the entity instance, for creating enrollments, for creating events linked to enrollments. And then of course we have the event itself. And the event is, of course, when I talk about event in this case now, I'll just talk about what we call the non-registration event or anonymous event or this kind of just a free-form event without any linkage to a patient or a person. For the metadata, again, everything that's available for you through the maintenance app is also available to you as part of any kind of integration work you want to do. Typically, maybe you want to get, say, a very common scenario is, for example, you want to get, every week you want to get kind of maybe a list of every metadata that has changed that you can do with what's called the object filtering. So for example, maybe you have a job that every week goes into the API saying, okay, give me all your organets that has changed or been created since a week ago. And then it will give you that, right? Then on top of that, you can also do a filtering of the fields. So you can make sure that you're only getting the fields you want. Because by default, in many cases, the API is actually hiding data for you, right? So it doesn't want to just give you everything by default. You actually want to hide as much as possible. Then you can ask for what you want to do. I'm not going to demo that now, but if anyone wants to demo that, that's very quickly to show you. But yeah. And then the formats. So my suggestion is that if you can stick to adjacent, please do that in all cases. It is definitely the one we have tested the most. It is probably the way forward and the one that's getting the most focus. We do support XML, not in every case, but in most cases, we do support XML. So for the aggregate, we have an XML version of the aggregate. We have an XML version of the tracker, although that is going away. I just want to mention that. And we do have an XML version of event, which is also going away. It might be added back, but the next iteration of the tracker APIs will actually be XML support. And then we have CSV support in a very few kind of special cases. So for example, you can import the option sets that is supported through CSV, or if you want to create data elements, that's also supported through CSV. The reason we don't have CSV for everything is that CSV is very much a tabular format. And when you start talking about stuff like the tracker, that payload is very complex. So it would be a very strange looking CSV, basically, if you want to kind of push all that into that CSV. So we do support some cases. And if you just go to the documentation, you can have a look at that. But again, for my point of view, you should focus on the JSON. Obviously, CSV is very nice if you're doing something in Excel and you want to export that CSV. That's just kind of the common use case for doing CSV. All right. So basically, the rest of the time, I'm just going to talk about a couple of ways you can send aggregate data into this S2. So we have two ways of doing that. They are a little bit different, but they kind of accomplish more or less the same. So the first thing I want to talk about is sending what we call a data value set. So what is a data value set? Well, it comprises of all the things that are required for data value. So that would be the organets, the period, and then a set of data elements with a value. So that might be, say, 2020-01, which would be a January, organets, something, and then data values, data element UIDs, basically. But since this is a data value set, so why this is not just called a data value is because it is also linked to a specific data set. So that means that whatever you have in the data values array must be linked through the data set UID. So you can only kind of send in the data elements that is part of the data set. And on top of that, we have something for data set when you're kind of done with entering data. We have something called a complete date, right? When you go into the data entry app, you click complete. That will send a date to the system and say, okay, now this data set is completed. So this API allows you to do that. And then the second one is the API for sending multiple data values. So now you will see it looks a bit different, right? So gone are the completeness, gone are the data set. That is not there anymore. So the reason we have data set here, it was a little bit about validation of the data values, but it's actually mostly about the complete date. But the limit here, as you can see, you can only have one period at a time, one organ at a time, right? So it does not allow you to send a kind of payload that spans multiple organs, for example. So if you're doing, say, an import job, and you have a MySQL database somewhere, or you have some kind of data somewhere that you want to get introduced to, you don't really want to limit yourself to sending one on one, sending data per period, per organ, right? That will take a lot of time. And that's not really how you want to do it. So this, that's the downside of this one. The positive side is, of course, that you can actually complete it. So if you, maybe you have like a running, almost like a real time integration that you just, whenever there's more data in your other system, you immediately want to just send that to DHS2, sending it as completed, then that's very nice. And that doesn't work you fine. But again, if you have a larger job, maybe you have some historical data, you probably want to use the second format. So the second format, you will recognize data element, the period, the organ, and the value as before. There's also comment, if you want to add a comment. This will show up if you go to the data entry and double click the input field, you know, you get this kind of the historical values and everything. And there's an input field there called comments or description or something. And this is where that comment will end up. And of course in here, there's only one single data value, but this array, you can have as many variations here as you want. And there's no limitation when it comes to mixing different organets, different periods and so on. You can just have a mix of whatever you want. And then you can actually create that big file if you want and you just send that to DHS2 and you'll probably fill in the values for all your organets and all your periods. So depending on what you want to do, you will have to kind of choose that right approach. So I think it's already time for demo. So I already have DHS2 running on my local host. Let's log into that one. Should I zoom in a bit? Or can everyone see? Yeah, it'd be helpful if you zoomed in a little bit. Is that better? Let's assume that's better. Okay, so I'm just going to kind of select a dataset now and we're going to see how we can send data for that dataset. So I already, of course, pre-selected that. So let's go into the data entry app. I have selected some, in this case, I've almost sent for one organets in both demos, but obviously you can also change to have different organets. The second demo will also show how to do multiple periods. But let's start with a simple one. So the one I selected was the mortality under five years. And let's go back a few years. So as you see now, there's nothing here at all. There's no data. So the question is, how do we then send some data to this one? So now we're just considering one single data entry. So that would be one organet, one dataset for one period. And we also want to make this completeness happen. Right? So we're not going to send a value maybe for all of the fields, but we're going to send a value for some of the fields. Now I have the demo obviously already prepared. Let me just show that here. But maybe like wondering, okay, so I see that you have the period here. This is January. But where do you get that UD from? Where do you get that UD from? Well, I will show you that. So if you have worked with this API before, probably this is very straightforward for you. But let me do that anyways. So I'm not going into the final details here. Let's go again. So if our APIs or our API organization in its endpoint does have a list of all of the possible organism in the system. Now, I just turned up paging and kind of did a bit of a cheat. I just searched for that directly. You can also, as I said, you have what we call an object filter, an object particular. You can do something like display name, like, angelic. Which is not what I think you have to do. So I don't think I can assume a little bit. But if you just look here, again, this is all documented. So don't work too much about that. But we have a filter here. We're saying we want the display name to contain angelic. Basically, it's a line. You're going to compare that to an SQL like. In this case, it is case sensitive. But we also have an I like, which is not non-consistent. But let's not worry about that right now. Okay, so we want to the CFC. So let's just take that for our APIs. So now we have the full information. We don't really need all this information now, but at least now we have it. So the next step is to find a dataset. And you can kind of use the same kind of approach. Again, big list of many kinds of datasets. In this time, let's not, let's not spend some time on actually doing the filtering. We don't have that many datasets here. So let's just take this one. You see, it's exactly the same. We add that slash endpoint. And again, you get the full information that includes all the dataset elements, all the assigned organets, and so on, so on, so on. Let's not worry too much about that right now, but we got the UAD, which is what we wanted. This is how we got that. The complete dates, well, that's, you can understand what that is. So basically, this is the January 2014. So maybe you completed that in the next month. Obviously, you can change that. So how we really like that. That's kind of specific to what you want to do. So let's not worry too much about that. But you have a set of data elements, right? So let's not dig too much into that right now, but I want to show you one thing. Actually, let me wait for that. Let me just show you how to send this. Let me just get a new terminal. So I'm going to use a tool called Curl. You probably have it on your on your machine already. There are many ways you can do this. You can also use something called Postman, which is a bit more UI, which has a bit nicer UI, but it's really up to you. There are also some plugins for Chrome, if you want to use that. But I think for now, we're just going to use Curl. It's a very straightforward tool. It's basically a small issue to be client that allows you to send some data. I can probably share this stuff after. I assume we have a place where you can share files. So I will also upload this later. So that's not worry too much about these scripts. I just have some scripts here that can do that for you. But let's just do it manually now, just to show how the process is. So obviously, you also have to give a username password. At the end of the time, you're of course talking to Disha Studio, obviously I have to give some username and password, but that makes sense. You need to point to the file you want to send. So in this case, that's data value set JSON. And then you have to use a sign that includes this entire file. You have to tell it what you're sending. So by default, it doesn't guess. So even though you're thinking, oh, I'm sending JSON data, the system should know that. For the system, it just is a set of characters. It doesn't really know anything about it. Just going to say, okay, we have some bytes, but you need to tell me if it's JSON or XML and so on. So you have to set what's called a content type. That means this is what I'm sending. So that would be application JSON. So now you're saying, look at this here, I'm a credentials. This is the file I'm sending. And this is what I'm sending. So now we just have to say, where are we sending it? Well, in this case, I'm just sending it to my local machine. So that would be localhost API data value sets. So that's, that is the endpoint for sending this kind of data value sets. Okay, this looks fine. Hopefully this is a live demo. Hopefully nothing will go wrong. And it did. This is probably, okay, okay. Let me try it out. Yeah. There are some, I'm running a, not, I'm running my own version of this just now. So it might be some issues, but as you see now, you will get this very long, I can let me make it a bit nicer. You will get this kind of input, what you call the import summary back. Now, of course, I'll send it one more time. So if it starts, I'll update it. And now I took these three values and updated the values. The first time, it would actually say imported tree. It's a report process, successful data set complete date. That's now 20, 15, 0.02, 0.05. This is what you had there. And obviously, if you go into the documentation, you will see there's a bunch of options. And many, many things you can do. Everything from skipping the cars, skipping auditing, being more strict in terms of data element validation and so on. But you will actually get the actually used import options back, which is kind of nice. So then you know exactly the options that we used for your import. So let's go back to this just to verify that that data now exists. So I just switched to February and then let's respect to January. And you will see now, you have the one, two, three, as we are hoping for. Okay, so I'm already 11.24. So let me just jump to the next demo. So this is working now. So this is the way you let me just see here. So it is completed, as you can see. This is very much mimics the way I used to enter data, then click complete. And so that is all verified now. So let's go back here. And now we will have a look at the next one, which is the data values. And as I told you, it's a data values. Now you can actually have a mix of different periods, different organets and different data elements. In this case, we don't do that much, but we do have, as you can see, we have different periods. So I'm sending the same three data elements as we did before. But this time, I'm actually sending it all the three for three different periods. So this is a June, July and August 4, 2014. Okay. And then the sending itself is basically exactly the same again. There's the same endpoint and all of that. The only thing that has changed is now you have, of course, at the point to the other file. So that would be now data values. So the system you recognize, it will actually try to see, do we have a data set in the payload? Well, we don't. We just have a list of our data values. Okay, then we go to the data value format in portal. So that's it. It's saying now success. So what actually happened here? Well, imported nine. So imported means created, by the way. Just so it's kind of a bit confusing. I wish we actually used created and not imported, but what really happened now is actually it did create nine new data values in the data value table. So if I send this same again, you will see this time it updated them. Okay. And let's just verify that. So again, we have the same organize and data set here. Now, we don't have to change that. But now we should have data for June, July and August. Go to June, see the data, July, we have data and August, we have data. So the last thing I'm going to show you before I finish is that it is also possible to clear values. If, for example, you sent some values and think, Oh, no, there was something wrong there. There was actually not meant to be sent. You can actually fix that by changing what we call the import strategy. So I didn't, I didn't test it before the demo, but it should be, should be fine, hopefully. So as you see now, if you remember, if I just go up a little bit, you will see that the default, the two import strategy creating updates, which is fine because that's usually what you want to do. But this time I actually want to blank out a set of values. So using the same payload, I was changing import strategy to delete. And if you go in here, I didn't print it this time, but you will see we actually deleted nine data values. So now if you go back here, say to June, the values are gone. So this is how we can orchestrate a data change for the aggregates. So from time to time, you might send something you shouldn't have sent, and then there is a fallback to easily remove that data again also. Obviously, in the slides, I have linked to the appropriate documentation. So for the aggregates, this link to the aggregates documentation, trackers link to the tracker documentation and so on. So when these slides are shared, you can just go here and you can click on that, and you can figure out the documentation. And the documentation is on Docspyton, obviously, this is Docstabiches2.org. That's our landing page for all kind of documentation, and this is under the developer documentation. You can see a web API. This is the full full API. So this is everything. Okay, I think we are all in two minutes before you shoot the show end. So I think if there's any questions, I think that would be the perfect time for that now. If not, we will hand it over to the loud team. Yeah, Morten, thanks so much. That's a great presentation. A couple of questions did come through. The first question is, can you talk a little bit about our connection or relationship with OpenHIM? We saw yesterday that Hiss Malawi and the Ministry of Health in Malawi have used DHIS2, or have used OpenHIM to connect DHIS2 and OpenLMIS. Could you talk a little bit more about that relationship, or if any advice or guidance on using OpenHIM? So we haven't really used OpenHIM much from our side. So I mean, OpenHIM is basically a mediator that allows you to kind of orchestrate kind of, you know, typical thing you would have. Okay, we have the proper container that will get DHIS2, then we have another container that will do some transformation on that, and then basically sending it somewhere else, right? So it's pretty straightforward to use that OpenHIM if you want that. We don't really have any recommendations to use it or not to use it. I know it is linked to the OpenHIE project, so if you like it and it does the job for you, that would be perfect. It does support storing of credentials and these kind of things. But again, that's just an orchestrator, so it can be replaced by other components also. I didn't, I wasn't part of any, well, I didn't watch a demo from yesterday, so I'm not sure about exactly that, and how the potential could be used with OpenLMIS. So yeah. Right. Okay, time for one more question. Thanks for that answer. In the, this is coming from Robert Modi. He's asking, in the first template, does specifying the complete data automatically mark the data entry complete? Yes. You mean the complete date, right? So that's the completed. Specifying the complete date, date. Does that automatically mark the data entry complete? Yes. Yes. Okay. Okay. All right. Well, Morten, we will let you know if there's any other questions that come through in the community practice. Please, everyone, still put your questions here. We can pass those along to Morten. He can answer them here directly as well. Now we are going to hand it over to Lauren and take us through the Laos case study of interoperability with M-supply. So over to you, Lauren. Great. Thank you. I'll just go ahead and share my screen now. We're kind of co-presenting this from two different locations, but it should be good. Good to go with this. So one moment. Can you see that now? Yeah. Looks great. I will just briefly start my video to say hello and put a face to the name while we still have light here in Laos. But basically, my name is Lauren Tice, and I'm a program manager with the Health Systems Strengthening Team at Chai Laos. And so we provide support to the Ministry of Health. So I'm honored to introduce Dr. Chance Lee Pomavong, the deputy director of the Department of Planning and Cooperation at the Ministry of Health, which oversees DHS2 as it's implemented in Laos. And we're excited today to give you a little bit more of an example of how the M-supply DHS2 integration has worked here in Laos so far. So thank you. But for now, Dr. Chance Lee will give a bit of an overview of the context of the HMIS here, and then I'll go into more integration details and examples with our malaria team and applications of the use here. So Dr. Chance Lee, I will hand it to you now and I'll move the slides along. Dr. Chance Lee, I think if you are able to unmute yourself. Martin, I think we have to make Dr. Chance Lee a co-presenter. Okay. Can you hear me now? Yeah, loud and clear. Thank you. Okay, you hear me now, right? Hello? Yes. Okay, Roland, could you please go to the next slide? Another slide. Next. Okay, so I just blip to you on the loud. The population is about 7.2 million, and also we have the GDP per capita is about 242,000 US dollars, and we have life expectancy about 64, 67, based on the census 2017. And we have maternal mortality is about 167 per 100,000, and also under five mortality is about 42 per 100 per thousand kids, and also health insulin code about 80 percent according to data 2019 and number of addixity, so more than 50 addixity. Next, Roland? Okay, so actually we have in the past, our country, we have the program like XPI of some of the medicine, and also we have stock out some of the malaria drug, and also we have on the other hand we have XPI of the HIV treatment drug. So this is our program. Next, Roland? So that's why the country, we are implementing the M-supply system or electronic logistic management information system so that we can get rid of that problem. So M-supply actually is a desktop-based application that uses offline and synchronized to central server access via online client. So we already start implementing M-supply since 2019, and there are more than 188 where how of the country has been starting to apply M-supply system. So M-supply system is good because we can have receive and aggregate supply envoy, also create and aggregate customer envoy, and also capture on calculation of runtime stock on hand so that we can get rid of the two programs of the as I mentioned, the stock out, and also the XPI of the medicine. Next, Roland? Could you go next? Okay. So on the other hand, aside from the M-supply system, in the country, we also you start implementing electronic information system, we apply M-supply system. We can start in form already since 2014. When we start, we just apply this as a tool only for collecting information like OPD, IPD, and also key indicator for MCS indicator. And then later on, we also try to integrate it with other program like Malaria TV and his way into the system. And then until 2020, we start also thinking about to continue to integrate the server and system into the DSS2. And also we also expanded system into the health center level so that health center can enter the data by themselves. And specifically in within the DSS2, we also apply different different level, some aggregate, some even capture, and we are also moving to the tracker now for some of the key, vertical program. So now the system is already running in the country more than five years. Next slide, Roland? So we know that we work similarly between the DSS2 and also M-supply. We didn't know, we work similarly. So many programs trying to use DSS2 for the medicine and also for the Malaria medicine. So we work similarly between M-supply and also the DSS2. So next slide? Next slide, Roland? So now we try to integrate it, we need to put in together between the DSS2 and also M-supply so that we can have one platform can we can use DSS2 for the generate information for both for the M-supply medicine and also for the linking to the to the program, this is program and also service program into the same the same thing. So so Roland, I will hand over to you to continue what is what is technically how you can bring between the between the M-supply and DSS2 what is useful for for this moving from putting integrated two systems together which we call interoperability between DSS2 and M-supply. Please, I hand over to you, Roland, you can continue the presentation. Thank you. So as Dr. Chancelli was saying, you know, really our M-supply system which is used within warehouses for stock management is really effective in the right tool for warehouse management where we look at real-time stock on hand data issuance kind of distribution history as well as very detailed levels of data about, you know, the stock itself and DSS2 is also really fit for purpose for these national programs that are looking at a bigger picture case and testing data and coverage amounts for the population. So each of these systems is incredibly effective for different capacities of their user base, but really we've noticed that each system can become even stronger if the information is brought together for decision-making and one example of where this is essential and really the way that you know there's shared stock responsibilities across national programs in Laos versus the central warehouse team and so there has to be a lot of communication between the two and this comes together with the commodity distribution plan where based on estimated case and testing burden from DSS2 that we can see as well as practical components of how we're able to get stock into warehouses, facilities, hospitals, it's important that we bring this information together from both of our systems to calculate appropriate distribution needs based on each of these pieces. So in order to do that we and the ministry overall supported the development of an API to push key M supply data to an aggregate DSS2 form on a weekly basis. So we work closely, the Sustainable Solutions is the M supply provider and so it was their team that built this API, but I think just going back to what we saw in the last presentation, we looked at each of our org units, so each warehouse in the country is an org unit in DSS2 and so for each of these warehouses we've identified the highest priority commodities starting with about 200 out of over 4,000 that are managed within M supply and selected these to be pushed to DSS2. So this happens every seven days right now and they'll talk a little bit more about the configuration but right now it comes in into a daily data entry form but once every seven days we look at the stock on hand stock that will expire in the next 90 days as well as stock that will expire in 180 days whereas once a month on the last day of the month we're pushing each item's average monthly issuance over the last three months as well as the opening balance. So the reason that we right now are doing this every seven days is because there's significant change that that does happen at this kind of interval as opposed to something like each month but we also wanted a little bit of flexibility to increase and I think that's become much more obvious and important in the last few months with you know pandemic response and actually needing to look at stock on hand and stock data every day because it's changing so quickly or it's essential to monitor daily. So the way that it's set up is really that a lot of the organization unit and data elements are you know configured using standardized codes so that it is quite flexible and sustainable to be updated um you know we didn't mention but M-supply is managed the system is managed by the medical product center not the same team that's doing DHIS too and so we've tried to build this flexibly so that these teams can work together in the systems that they know best. So within M-supply which we have a couple screenshots of here for each warehouse we've inserted the DHIS2 organization unit code to to know which organization to be pushing this information to as well as for each item as I mentioned that that we selected we enter a universal item code so that if the medical product center needs to adjust something about their codes or their item names it's still the integration is not breaking and then as you can see there's this tick box where we can toggle items for the push off and on depending on if they're needed and all that's really required for it to be pushed is that the appropriate data element is set up within DHIS2. So we use a similar structure for each of these data elements but this is a stock on hand example where we we insert that code to have this linkage between the the warehouse and the organization unit it's coming in as aggregate data and generally we started out by doing positive or zero integer values but we've also learned that for certain items and this is coming up with you know vaccines or certain family products you might actually be issuing portions of portions of an item so looking at numbers and then also importantly we've started using the last value aggregation type so even if we're pushing at infrequent intervals where it's sometimes once every seven days or sometimes more often we can look at the most recent value so this is really enabled you know quite a flexible review of data elements for maybe one item you want to look at in detail or look at you know one location in more detail so depending on the program sometimes you know it's actually enough and a huge difference now that we're able actually able to actually see all of this data in one place so on the left right you might on the left you might want to look at one item across multiple locations where on the right we're actually looking at you know multiple items at a single location over time and on the right you can see we're looking at different time periods where maybe you want to see historically the the value at the end of a month but also this week so quite flexible to look at what I would consider a kind of raw data elements but what we'll we've also noticed and where we're hoping to go is to make sure that we can adjust these and leverage them into more actionable indicators so it sounds like this is maybe similar to to what some other countries that have presented so far in this academy have demonstrated but to share a little bit about our National Malaria program which has been a key user of logistics data here in Laos the team has used testing and case data to develop thresholds for what their stock should be so that we can compare it to what it actually is so they have calculated a monthly need for each of their eight priority commodities using historical test and case trends as I said that would come from DHS too but then we've also this is currently happening outside of DHS too so that we can also factor in some of the seasonality and national stock minimums policy into this I think a long-term vision is to figure out whether we can calculate this using data in a more automated fashion within DHS too but for now it takes you know a few factors and so that's happening outside and we're uploading the estimated monthly need into the system and then using this to really normalize the data right we have stock data that might have incredibly different scales depending on the the facility that we're talking about but we can calculate months of stock available by looking at our stock on hand over our monthly need to to normalize this and make it a bit more comparable and so one way that the program is now visualizing this is comparing this month's of stock available to maximum and minimum thresholds I think this is you know between three and five months for example for a health facility level to guide decision makers to to be able to address stock shortages or expected expire expires so we can see just from this one chart on the left it's very easy to notice when stock is too low and an urgent resupply is needed also you can quickly move over to a different item to to see it's within range and and that is going well but then as you can see in with this particular example we have quite a few items that have excess stock and this is important to look at um to to reallocate and so we can start to leverage expiry data to know whether we need to move these items around or if we can actually just pause or adjust distributions to get this closer to the actual stock that's expected to be used within that facility so we're wasting less but also making sure that there's sufficient stock so one other way that the malaria team is also utilizing this logistics data is really just um applying a a legend to this stock months of stock available indicator um to have kind of quick color coding as well for stock that's within range and and green or kind of warning or our stock outs with yellow and red as well as potential overstocks with this blue and so this has been we've adjusted and are using different visualizations based on the users roles within the country whether they're based within a province or a given district or up centrally we need to look at different scales of this so we've adapted a few different approaches depending on that user one other kind of exciting opportunity that we were also exploring in allow is actually leveraging this to not only look at what is our current stock situation but how can we actually use this data to define the corrective action and figure out what exactly should be resupplied or distributed and so now there is this distribution tool developed with with these particular commodities where we can based on the maximum stock level which is our ideal stock we want to make sure we have enough for the the suggested months of stock on hand um to calculate the quantity to distribute we take our maximum stock threshold so we'll take an example here of sepon this first health center on the list we have 48 48 of this commodity on hand we know that we need or ideally there's 70 within the the facility none of the stock are expiring and so therefore the team needs to distribute 22 to this facility so it's become much more actionable to have to know not only do we need to address some kind of shortage but how specifically can the team build those distribution plans make sure to communicate that to the appropriate party and really leverage DHS2 for making decisions and not only understanding the current situation another area where we're seeing a lot of opportunity for the logistics data and these predictors that I'm about to look at are not currently live but we're currently figuring out you know what's the best way to deploy them here is to you know leverage predictors to ensure outbreak preparedness for a set of items so this is all demonstrative data but if we're looking at our our COVID example um we have our case data that's coming into this system as events and we can see that around the country and similarly similarly we can look at our logistics preparedness so in the middle here if we want to quickly see at all the 188 warehouses within the country just to simply answer the question do we have stock we can use a predictor to look at the necessary set of items and if all of them are in stock above zero we can flag is green we know that we urgently need to correct the the facilities that have stock outs and once we're able to start answering this question and make sure that at least some stock is available at all warehouses and could get to the facilities that we need we can start to ask more nuanced questions such as do we have enough stock and this is where we start to compare our stock on hand data to the minimum thresholds that we've defined to say you know how can we make sure that this is sufficient so we're definitely we're currently looking at this kind of indicator outside of DHS too but really think that there's a a lot to be gained to have this right next to our case data and have the same stakeholders being able to view this information in the same place rather than than keeping it separate so you know there are many ways and I mean this can be a significant amount of data and so it allows this data to be used more efficiently and effectively as we have it available in DHS too alongside the the programmatic data but it's really important to flag for you know any system but I guess in particular this M supply implementation if they consider how the implementation is going and what the status is to be able to factor that into interpretation of the information so so here in Laos that there is a national mandate for all programs to be using M supply and have their stock integrated into the national system rather than their own logistics flows but that takes time and that's of a more recent mandate so we're working on full integration into the system and overall M supply is still being adopted across each of these warehouses it's been a five-year scale up process for us to implement this tool but as that continues to expand and become more the reach is stronger right now we can look at that alongside data quality metrics including stock integration progress as well as M supply site activity which you can see on the right of this screen we know that you know most sites are active and actively using M supply but there are some that maybe haven't been using it within the last month and so we need to take that into consideration when we are evaluating the logistics data that is coming from that integration additionally because the the Ministry of Health does not collect individual or patient level consumption data the M supply data that we have coming in here and the stock movement that it represents can be used as a proxy for consumption but really it's important to take into consideration the assumption that stock is actually reaching and informed by patient needs so with our average monthly issuance for example sometimes we use that as average monthly consumption but it does come with a few caveats but in the absence of individual level patient data this is an incredible wealth of data to understand what the stock situation here is in Laos so going forward these are just a couple examples in the malaria data is currently fully in use right now and this emergency operations center data is also increasingly in use there are quite a few ways that the Ministry of Health looks to leverage integrated logistics and health management information to put this data into context and ultimately make sure that the health services in the country are reaching who they need when they need where they need so a few ways just to to close of where we hope to go from here are you know using this information to encourage and monitor this stock integration that we had mentioned because until all the stock in the country is really flowing into the centralized system it won't be as powerful and once more of this information is in here I think we can also increase the utilization of this more regular and supply data than older systems that were in place such as this monthly aggregate data that programs are directly entering additionally we'll explore the creation of some indicators to aggregate individual products into generic items so I'm not sure if this is something that other teams have faced but you know for latex gloves for example I think there's about 28 different versions or different combinations that we have here in Laos but sometimes we just need to answer the question do we have any so we'll create some summary indicators to aggregate those products we also look to create maximum and minimum thresholds for stock you know across different programs beyond a malaria team right now and as I mentioned there may be some potential to automate some of this and and finally we also will plan to look at stock available availability dashboards in our emergency operations center in this year more than any has highlighted how important it is to be able to have a rapid view of what's going on in the country and and what we need to do to be prepared for not only the 19 national notifiable diseases that we have here in Laos but other emergent ones including COVID so you know I think there is really an opportunity here and so also curious to hear from you guys in terms of questions or other ideas that this may raise for how ELMIS data coming into DHIS too might be helpful for your your teams and your ministries but for now we'll pause there but we'd be happy to connect further if needed so it's a quick dive for your time and we'll kick that back to the group in case there are any questions that have come up great thank you so much Lauren and Dr. Chiseli there are quite a few questions that have actually come up Morton has been frantically typing here on Slack to try to answer some of them and we've also added John Lewis he's from Hisp Vietnam who's been working on this project quite a lot as well to Slack to be able to answer some of the questions but a few that I can point at you Lauren is the interoperability of both systems the MSupply and DHIS too is it able to handle national qualification exercises for supply planning for instance like with malaria HIV or TB have you experienced any kind yeah so have you been able to do any kind of national qualification exercise and supply planning using the interoperability systems yeah so I mean I think right now the way that we're leveraging MSupply right now is not for quantification itself it's kind of an external process I'm actually here with one of my colleagues who works closely and in some of the supply planning for the ministry of in Laos so I'll just speak kind of quickly to that yeah just to point out I think we we know that the DHIS2 data are essential in the MSupply data to informing those exercises but I think what we're finding is that the you know there are definitely more complex analyses that are needed to do a good detailed forecasting using the data but then applying different assumptions that you know definitely need to be done in a more flexible model such as Excel or Quantimed or whatever the program is using but the distribution tool that Lauren presented we're using for distribution planning within the country because that requires it's a bit more predictable and it requires less detailed analysis so that's the difference between those two okay great thanks I guess that was Rose is that right sorry I should have introduced myself I'm here with Lauren just again we're we're working in the same space so great well thanks for being around and answering that question a couple more questions how do you handle or how have you been handling the five year rollout of MSupply across the entire country does Laos have sites that are not using MSupply yet so that our MSupply implementation began in the ministry of health I often say are we work so closely with the ministry of health that we are a collaborative team and Chai has been lead support for MSupply implementation in Laos but it began in 2015 after an initial scoping to look at which LMS we wanted to use a big part of our decision went into being able to use the system offline and and synchronize up to a central location when connectivity was available and so it was a very gradual scale up over the course of five years so it is complete and is in every province um you know it's central at the warehouse it's in every province and in every district in Laos and so as of about a year ago we're fully complete and we're transitioning into that that means that it's installed within these sites and is used within the primary warehouse so there's a few elements of you know starting to integrate additional products into the system that requires you know follow-up training and agreements across teams to make sure that there's clear protocol for what needs to happen but a large part of our kind of continued monitoring is to look at that activity which we did show so whenever a site is not active we I think the the primary reasons that this happens are because some device has broken and we need to repair or replace the computer or there has been staff turnover and so the person who used MSupply has maybe left and they need a refresher training and I would say that's the vast majority of reasons that sites would not be active and so we the the medical product supply center has teams that that go out and provide that immediate assistance so that they're able to to use it and so now that the tool is fully implemented across the country we are shifting kind of our next focus on how do we best use this information because it becomes so much more powerful now that we are able to have it a national view then then just some of the site so there's some monitoring that's ongoing to make sure there's not like kind of continued use of the the tool but you know it's fully fully adopted within the country otherwise okay great thanks there was a question from Robert Modi asking if there is any middleware required to achieving interoperability and John Lewis again from his Vietnam said that no there is actually there's no middle well middleware involved in pushing data from MSupply to DHIS too which is which is pretty cool I think a lot most folks here are probably less familiar with MSupply Laura could you talk a little bit about why you chose MSupply and maybe tell us a little bit more about the platform does it how is the costing structure is it open source maybe just a little bit more background in supply sure so as I said MSupply is provided by an accounting called sustainable solutions based in New Zealand um but they were kind of have done a few implementations in Southeast Asia um and so I think as I briefly mentioned one of the the main considerations at the time the scoping was in 2013 and 2014 and so some of the the situation was a bit different of what was available at the time but we because of connectivity certainly the ability to work offline was a key component of that selection but the functionalities that that are used within the warehouse are largely for distributing stock down to lower warehouses or to facilities itself as well as receiving them within those facilities and so we focused on being able to do that stock management there are capacities within MSupply as a tool to do a patient level distribution and there's also now an MSupply mobile application for tablets we're not using those functionalities right now as well as some other ones that are within there it right now is a licensed software and you know depending on the site and whether it's offline or online capacity it's between five hundred and about eight hundred dollars per per site to use per year um but the MSupply team is also doing a bit of an overhaul to their data um the database structure and is rebuilding it as an open source tool and the MSupply mobile is also open source so it's there's a few different options and iterations at this point um I think now because the the tool has grown a bit since we first started implementing 2014-2015 I don't know if there's any other kind of particular questions about that one but the the ability we found that a lot of the LMS tools um other ones that we looked at were really strong for using in a single site but maybe not this kind of national network of both stores themselves as well as just general customers and or recipients of the commodities and so this was the best you know really fit all of our our primary needs um and has been as I said fully adopted and so we'll continue to to use this one um one thing I will flag as well in the the benefit of some of our integration as well as because MSupply is limited in being able to view data from multiple stores at one time and so that's also been a big opportunity for DHIS too to help us bring this information together and have a broader picture of what's going on in the country. That's a that's a great last thing to say because Martin in is asking a question um is is it correct that the integration with DHIS too is to tap into the track the aggregate and tracker data to cross analyze data or is it just because MSupply does not give you an adequate um uh report generation for commodities needed I guess I'm kind of interpreting Martin's question a little bit more maybe you guys can see it as well it's a second to last question but yeah and I think that there's elements of both of that I think the very immediate fix that that now that it's implemented and we've launched this implementation earlier this year now that it's there we can have much more visualization um across the country um but the long term well three main things one we're able to see this information one place to beat some of the limitations of the the tool that we're using um or MSupply two different stakeholders and and Dr. Tinsley you can potentially speak to this later on if there's additional time or questions but different stakeholders are using um different systems most as we said all national programs are using DHIS too within the country and so that is where they're going to and we want to be able to pull this information together um because it's the more appropriate tool for a wider audience and then third really and where we'll move into now is being able to do more analytics actually with the epidemiological data that's within DHIS too so our you know as we were looking at earlier there's potential to figure out what are the actual stock requirements based on cases and testing and then some of that more nuanced information we can start looking at population data that's also captured in there um but that's you know as I said very immediate uses are really just being able to see this breadth but we want to be able to do this and these analytics together that's a really good point Lauren I think that you know so many countries are just struggling to be able to just show where this the current stock status is at each facility for each commodity and that's the most pressing concern but I think that you all have shown that that you're able to get that data and once you get that data you can start to build out these more triangulated analytics and indicators looking at yeah the epidemiological data the case data the population data um and that's what Malawi showed a little bit yesterday too if folks remember is looking at that like consumption to issuance ratio and case load to issuance ratio and using those indicators to and putting those on dashboards and putting in lines and guidance on those dashboards as well to inform folks about how they should respond to those and it's one of the interesting things is those kinds of triangulated indicators are a little bit better seemingly for projecting where you will be for example if you know that your issuance your case load is higher this month than your issuance then next month you can think that you might have an understock or a stock out situation um and and it gives you a little bit more insight and I think those are the kinds of indicators and analytics that here within DHI is too cool we really want to be able to support those we want to make sure that we produce the analytics and that we are able to have the ability to calculate those kinds of indicators um Lauren and team I think that we've taken up enough of your time thank you again so much it was a fantastic presentation it's been recorded we're going to post it up on youtube as well as on in our google drive for everyone to be able to watch later is there any last thing that you wanted to say or anyone from john or anyone from the team that wanted to say any last words before we go into break so I have something to say a little bit on the how the country used the dss2 and i'm surprised because in the country we are running the the the world bank project one of the indicators that we need to report to the to the project is this the ascension uh commodity and and drug at the health center level so we use the dss2 to to to report to the bank but the data is come from the m supply meaning that these two systems should work together closely so that we can have the information from ascension medicine and logistic report to the bank so that we have the condition that if we report the number of health center that having the medicine enough for 30 days so in certain in based on that report the country or is a health facility will receive a amount of money so that they can improve the quality of service at the health center level that that's all going to add thank you great thank you doctor uh to say I think that's going to be the last word uh uh from the team in lao again thank you all so much fantastic presentation uh it has been recorded so we'll be able to post that up in a couple of hours um all right so now I have the responsibility to give us the word of the day so let me just put that up on the screen okay so the you can all see it hopefully I don't need to go into presentation mode the word of the day is ain't no mountain high enough so ain't no mountain high enough we're going with the thing here and if you could we go into the attendance on the google drive fill in the words of the day ain't no mountain high enough and make sure that you get credit for being here today we are going to take now a short it's going to be a six minute break all right we are now ready I think to pass the mic over to Landry and Nuno they are joining us from um Medexus which is kind of a relatively new LMIS platform but one that we are um working with more and more these days so they're going to take us through their work that they're doing uh in Burundi hopefully and uh connecting to dhi's too as well all right so Landry Nuno you can unmute yourself share your screen and over to you hello Nuno here Landry will uh is preparing the data screen okay okay we can see it looks great I need to unmute yourself can you hear me we can hear you and we can see your screen okay so I am I am now actually in Burundi just drive actually here and uh hope that the connection stays stable there so my name is Landry Medega I'm working with Viper Solutions I've been with Viper Solutions for uh for 10 years and uh as uh as then consultants with logistics backgrounds so today um we are really delighted to uh to have that opportunity to to talk about our use case actually and interoperability with dhi's too then in Burundi so we will talk about Medexus as Scott already uh introduced and then this is actually uh end-to-end uh visibility platform and here as you know I will be presenting today with my colleague Nuno Renault then who is our head developer Medexus at Viper Solutions so the plan today is that we'll talk about a bit of Medexus story and and then the needs and how we came to the design also and talk about the interoperability have a quick demo of it actually with Viper Solutions and then a way forward and my question and answer so how does the idea actually for Medexus uh yeah come so Medexus was born based on actually two main issues we have identified based on our 15 years of experience as Viper Solutions uh a need a long space uh organization NGO so then the first issue is about the data and information visibility and accuracy uh what's then you know what you know actually so in that component actually we see uh quickly a lack of uh visibility throughout the entire supply chain lack of reliable data some are collected actually based on the uh on the paper based LMIS but still inaccurate you know sometimes a tree uh can quickly be innate actually on that paper based LMIS and then we saw some data also missing or sometimes not reported properly actually and also uh then a complete view of the situation down there at the last mile then the second components actually is a group component of issue is actually the processes and the tools uh here we saw actually a lack of robust processes in in place and also sometimes is those processes do not actually manage storage and distribution we see also then a poor you know resulting yeah in a poor forecasting and quantification because of the uh the inaccuracy of the data and also uh some paper based systems a lot actually but still use widely in countries and also where um some electronic solutions exist then we see also that they are still not standardized and sometimes is really uh on the loan systems and this is the reason as to why then we uh came to a new idea of uh creating from scratch then an LMIS system you know and that is uh and complimenting it with our experience our 15 years experience in supply chain management you know and so combining actually the technology and logistics so here we see the reason one of the reason a few reasons as you see here is that we experienced some systems uh before coming to our new LMIS we have team up with uh great organizations and where we saw actually at the end of the day that uh some of the systems we use or country use actually uh are not user friendly sometimes you know and and some are expensive as well at the end of the day uh to sustain actually in a country where resources are really limited and then we see also that there are some uh commercial and open source or open source uh platforms and then which uh actually at the end of the day some of them are not owned by the country at the end of the day and and then not having also all the key futures actually set system needs in country to manage proper uh LMIS data and then uh as you see and and and one of the colleagues already mentioned it we see now that the HS2 uh now is in every almost uh uh in 100 countries correct me if i'm wrong and and then so we need actually as a new system or as a LMIS system to integrate and then so we see in country that there is no correlation between then the clinical data uh collected greatly by the HS2 and logistical data actually and then we see also some poor customer satisfaction weights mainly at the central medical store and here i'm talking about the one is around 30% you know so imagine you come with a product and then uh you had you have to to to to find alternative for the 70% of the rest of the product on your list actually because they're simply not available at central medical store and then we see also a long procurement time actually at that central medical store this is for various factors you know and and then the lengthy is a lengthy process and uh and there is about a tender in a public sector you know and and then resulting in an in an average of 367 days actually to have the process completed actually and and and and good delivered and so we print uh stock outs uh over stocks and a lot of expires so then we came to design uh experiments uh Medexis in Gondi so we first then created actually go globally we decided to create then Medexis to design it in 2018 and then based on our experience as I already said and then also from a bottom up uh approach actually yeah then uh we placed then the uh the one who is manipulated the product actually in the center of our interventions in the center of that you know and to make the system really friendly so that we can then gain also time actually to dedicate to other duties and then we then team up with our technological partner XNR in Portugal and came to design Medexis so in Medexis after implementation uh you can see that Medexis provides visibility in the unfair supply can you know this is for example how we show stock outs in right for example you know so that you can uh you as a off work and then take an action and and make sure that you get the product available and also we piloted then Medexis one year after in 2019 in Burundi covering actually in total 13 products then is uh is more a family planning product one abortion post abortion care product and and some nutrition product as well where then we we then uh at the center of the health facility level we covered then 133 health facilities and uh and each of them actually have has a 9000 average their health facility covered then 9000 uh patients facility and average and then and and then we cover also 13 districts in the country out of 47 and about the health facilities the 133 is out of 100 is out of 1336 in total in country and then also the central medical store and here we then uh besides providing the system and testing the system up we also implemented then our approach you know the building and coaching on job as well so the where we developed then and user guides and training modules and then we trained first the then six master trainers and then in total 89 users and developed together with them then a monitoring and supervision and also where then by we could then measure performance as we rest and then we developed also inventory and stock management reporting tools and different dashboards then to make sure that at the end of the day it goes actually inform decision to be taken and also coached then and supervised the district execution and in total 240 healthcare providers you know we have increased the number as the the number as this because we wanted to make sure because of because of the turnaround time in staffing actually in public sector in the health public sector then we had to to train at least two people then at health facilities and make sure that somebody remains and and to continue doing the work so so far what we have seen actually is that the availability has increased here then you know by 14 percent and coming from a baseline of 80 percent 85 percent because because at that baseline coincidates with the replenishment of the facilities so the level was high already from there so we had to take on the challenge and then bring it then to 99 percent page 99 percent at the end of December of last year and then a stock out the stock out here is that we measure it by health facilities experiencing stock out and here you can see that that you know they've been reduced substantially while using Medexis then from 94 percent actually to eight percent at the end of the the the experimentation and and here you know about the expires where we see them reduced significantly as well then by 85 percent reduction of course a lot and so and you can see that the three percent at the end of December you see here is is because there were a lot of products actually which and that was me so postal the collate actually a post office in here a post product then that was about to to expand in any way and then we had so many volume actually at that time at those health facilities and then the overstock has reduced here slightly we wanted to put it here because it's also a good lesson learned for us because here what we see is that we came to know about the for example a female condom which is really in low consumption because of that low consumption that we had then over that that then informed the decision to to make sure that then they change strategy in demand creation for that specific order you know so now I will share with you the lesson learned during the implementation we categorized that they're actually in two groups the technical group and the technical group and behavioral and skills so about the technical group you can see that the use of another computer is really dedicated because here in Burundi luckily all the most of the health facilities all of them actually have at least one computer they do have one computer sometimes that computer is not functioning well or is second hand computer but somewhere but they do have a computer and then is the chief nursing then using that computer and sometimes he attends meeting with that computer you know going around and then the responsible pharmacist then experienced the delays actually in keen the records actually for the LMIS so we have so here we think that a second computer dedicated only for pharmacy is near and then now we see also that the data where then recorded accurately and that was good and also we haven't seen actually the I recall data validation at that time but then we managed to create them sort of units unit at the district level then to review and then analyze and and and make sure that they validate together then data collected actually from health facilities of the area of information so that was also a good lesson learned for us so that when data is validated from there it can be sent to the the central level and then the internet connectivity is really essential here we see that is important to have a well internet good con internet connectivity and then to make sure that you report then you see your data actually in real time where internet connectivity is not available actually the system is able then later to then synchronize and then make sure that data actually key then can be viewed then in real time when connected back and about the behavior here we see some resistance to change we were expecting this actually but not a lot was really not significant actually because of the fact that people were really high motivated actually to use Medexis and because they have seen the benefit from it and and for example where they were spending actually two weeks to to to reorder for replenishment we managed to bring it down because of the use of Medexis down to three hours in total you know at the end of the day so that was they could dedicate to then the gain in time actually clinical aspect and then the we have seen then as I already mentioned here the data see if you can't reject any time we have also experienced some health you know the level of computer literacy of the health of some of the health workers was really problematic sometimes you know for example we had to work with a really high motivated 55 years old responsible pharmacist who has never touched a you know a computer before and but he's he's really good in maintaining records actually on paper-based LMIS so we have then reached an agreement with the chief nursing who can then manipulate a computer then to after the responsible pharmacist has finished then the paper-based LMIS then to help him assist here in keying actually the data in the system and which was really done then every time then they actually discussed and the so here I will then give the word to my colleague Nuno about the enteroperability here okay let me share my screen and go with it okay I believe you can see my screen right now yes I can see your login screen okay perfect so what we are presenting today is a bit of a proof of concept that we are developing together with the MOH in Burundi as well as with DHS2 so we can have full interoperability between Medexys and and DHS2 just login here okay I won't go through the the normal process of explaining Medexys so just cut to the chase so we have this configuration page where we define a number of of information that is coming from from the the current structural data in in Burundi so this is a one-time configuration and we do have the the facility so the idea here within our our development was to be able to use the pre-existing DHS2 structure with facilities or and commodities as well so there is no need to configure both the the facilities in both platforms as we can just seek all of the data from DHS2 and then import it in in Medexys automatically by simply by clicking of a of a button the same thing with the commodities where in here we can have the option to import the commodities that are created in DHS2 as well as export the the commodities that are created in Medexys into DHS2 so in this case if we want to import let's say this product we just import the data and then fill all the all the necessary information to import and create this commodity in in Medexys in a really easy way to to reach the interoperability between the two platforms okay so now we have in DHS2 one data entry page that let's say for January or for March where we have a number of products then stock on hand monthly consumption and Medexys received so all the commodities that have entered the stock of the of the facility and this is DHS2 so if we go to November we don't have any data yet and let's think the month receipts from Medexys into DHS2 so we have a number of items that have arrived in in this facility inside Medexys and we can just export all of them into DHS2 and if we just refresh the page go here we have the Medexys receipts here and we have the same for all of the three options so in this proof of concept we are doing this manually to to have full control over the the information but in a at a later stage the idea is to do this automatically and without any human intervention every at the end of the day at the end of the week something like that depending on the on the needs here if we go to November we have the stock on hand and with all this data we can then process the the the data on the DHS2 side and go from there also we do have the the ability to not only send data to to DHS2 but also to retrieve data that is saved in DHS2 so let's go with April and what the system does is reaches DHS2 grabs the data and shows it here I can see April so the consumption is the the same on on the two sides and we just validate the consumption and import this data into Medexys to then use it for the replenishment process inside the Medexys so same thing with the end balance same same process we select the facility the month that we want to import and go from there and obviously this is as as I said in the beginning just the proof of concept and we can work with any variables that we want to pass from Medexys into DHS2 as well as receive any data that is available in DHS2 and imported in Medexys to be presented so at this point there is no mountain high enough for for us to to really achieve the interoperability between the two platforms okay so now I will stop sharing and pass the the discussion to Landry if you have any questions please feel free to to make them thank you thank you Scott do you want to do the questions yeah sure so seems like quite a lot of folks also interested in this as well a few questions is it time for questions guys Landry do you have anything else to add or are you ready for questions yeah we have one more so point four with some some texts on the conclusion and the way forward for us sure so no go ahead yeah we've got some time so go ahead okay so thank you very much then we see Medexys actually as as an LMS as a key tool actually to ensure continuous availability of products here to the last mile and and thereby then ensure actually sufficient and permanent accessibility to the most vulnerable segment of the population here to the product really needs actually and and then the way forward here is that we extend then Medexys as as I said then to nutrition products and now extending it to to immunizations as well as as of end of this month and then we're funding from UNICEF and also then it will be then it will be seven additional health districts covering then the 90 health facilities in total and then we are also actually engage in discussions actually with the government and waiting for this their decision then to adopt one single national integrated ELMIS for all commodities actually in country you know so and also then we now that we have achieved that great milestone with interoperability with the HS2 we would like to continue and have a full and proven interoperability with sage system here in country as well why sage because sage is actually the main used system by the central medical star so right now that sage that system of the central medical star doesn't talk to you know so this is something we want to see if Medexys can play a role there and then you know and then sort of interface and then talk to both systems so thank you very much and so this is now time for yeah I will I will give the word actually to the moderator here yeah thank you laundry Nuno thank you so much for the for the great presentation quite a lot of interest and questions coming in on the community or on the slack channel so let me just throw some of these at you here the first one is is the connection between Medexys and DHS to achieve directly or do you have to use any kind of middle wear it's directly so it's programmed let's say within Medexys and it's configurable yeah that's great and you know for those of you who have probably attended several DHS two economies in the past you've never had we've never had another platform present during an academy before and in this academy you've seen that we've had a little bit from open LMIS last presentation on M supply and then we're actually having the lead developers and implementation team of Medexys actually present to you and that's because we appreciate with DHS two that we have to work with these platforms to be able to cover the use cases for supply chain DHS two can't work alone here we need to we need to collaborate and that's why we're having this bilateral communication and working directly with these platforms specifically Medexys quite a lot right now okay some more questions does Medexys have an offline version we have the offline version in the mobile app and we are working on version 2.0 of the web portal to be able to work completely offline yeah so soon we will be able to to have offline capabilities both on the on the app as well as on the on the web platform yeah right okay great another question is can you go over a little bit you go over a little bit more about how Medexys handles receipts yeah sure and we have probably best if I screen share again or something like that so we have that the normal process the health facility places a requisition that it's validated on the on the higher level and then once the the requisition for certain commodities are validated then you know you know you can you can show your screen now if you'd like to actually just show it yeah okay seeing my screen right now yeah yeah okay so what we do and obviously these data entry are all the data entry that are possible in the system depending on the role of each user you will be seeing one or other data entry but since I'm logged in as admin I have the the option to see all of the available data entry so and we have organized these in a very sequential way let's say so the requisite the the health facility creates a requisition that passes through validation from a higher level as I was saying and once this requisition is validated it then turns into an order with allocation and allocation here is automatically suggested by the system based on FIFO and FAFO algorithms then goes through the picking and packing process and once it's it's packed it creates automatically a shipment then you can arrange those shipments into consignments and those consignments are then distributed and validated at the the health facility when they arrive so starting from there you can do your consumption write-off transfer stock and this is the transfer stock within the same facility you also have the inter-facility transfer we can look at this one so we have different options for the inter-facility transfer that could be a normal transfer let's say you are redistributing some products that you are that you have in overstock and you are on distributing them to another facility nearby or you have the option to do the stock return so if the items are more cold or if they need to go back to Central Medical Store then you can have a stock return that only follows or you can only send those products into the facilities that are defined in the facility supply path so we have every facility with the supply path that is organized by program if necessary and then we can the the stock return works inside that supply path only then we have the normal inventory checkup and also the system based on on the consumption the system can suggest the the inventory settings which are the the inventory levels defined for that for that commodity so the reorder level the emergency order level and the maximum level or optimal level so then depending on the consumption over a period of time the system can say okay you may need to readjust the the inventory settings that you have defined for this product also you have the option when there is no no inventory on the supply path to accept or to register some products that you have bought outside of the normal supply path so we have two ways to to answer the question we have two ways of to of receiving items one is through the whole process here with the requisition and then go through all of the steps and the other one is to directly add items that were purchased outside of the of the supply path great okay answers the the question yeah very thoroughly thank you so much um a couple of other questions would you guys be able to say a little bit about how much it costs to use midexis yeah I will pass to to laundry so if you have any technical questions for for myself I am happy to to answer all of those and then we pass to to to laundry if that's okay with you yes thank you no no uh actually um yeah midexis as you know we developed the midexis from on our from scratch then from our resources so we had to use uh from start then licensing fee but now we work in actually hardly you know on in in changing actually that model so that we make it affordable you know meet also our vision and mission you know the organization and and then we will then be offering this product one because it's almost finalized we will offer it then you know based on on a case by case of course but then then we'll be on open sort of um affordable like using the premium then to see sorry laundry you're going sorry laundry you're going in and out you said that you're going to use a freemium model yes maybe maybe I can I can answer since laundry has some connectivity issues uh we are on the upcoming version we will be applying the freemium model so midexis will be free with some basic capabilities and then the countries can buy uh specific modules to enhance the the the the free capabilities so that's the idea in the in the upcoming months but uh but right and that that's a that's a fairly standard model I would say so the freemium model so but i plus solutions is a non-profit correct you're not a for-profit yep company correct we are a non-profit organization correct great okay so sorry go ahead sorry in squad yeah so then the the most important thing for us is actually to just cover our costs if we can cover our costs then it's fine absolutely yeah cover our costs and then meet the mission and vision of the organization and that's it yeah wonderful wonderful great a few more questions here in the last several minutes that we have how does the midexa system calculate the need for each site do you have different methods to calculate the different consumption data or caseload data to calculate the need for each site in in this case we have this option to to define the the needs for for the based on what I've told about about the emergency level the reorder level and the maximum slash optimal level and depending on that the system um when when the so let me show you here on the requisition if you select the facility and then let's see this is a bit out of the so the system automatically based on your stock on hand and the maximum levels that are defined by by program the system can suggest that you should order in this case these two items you know and the system suggests these the the stock on harder and the quantity needed to reach the the the optimal stock level great okay and then I think we have time for one last question are the picking and packing creating shipments managing shipments consignment modules are those working and okay maybe you could yeah I think you actually I can go through all of them I didn't do did it because of of the time for for the answer so in this case so let me start in the beginning so here you have a couple of requisitions that are still waiting to be validated let me grab one and in here you have all the information about the the requisition so the destination facility who from the the defined supply path for this facility who will be the facility that will provide these these items also the information about the available budget and the requisition cost and you have here the information on the the items that are still in the pipeline so they have been validated but they have not yet reach the the facility so you can have a clear idea on the amount of of or the value of the products that are still in the pipeline to be to be deliberate and based on that you can reject the the requisition or validate it and submit for for allocation once you pass this this stage the order now can be allocated and the system automatically as I said earlier based on FAFO and FIFA algorithms the system automatically calculates or or defines the the best batch that should be sent in this case so the batch and the expiration date and also the location where it is in the origin facility okay if there is need for it the stock so in this case the system since we have stock on on origin facility the system automatically sends all that that was validated before all all items but the user that is allocating can ration some of this so if we click here what we change is we can change the quantity we send and we then create or this order stays open until we fulfill the the the quantity that we need once we have this allocated it goes to the picking and packing let me see if we can okay so this one is for um we have this we just need to mark it as picked and packed assuming that we have packed the the whole quantity and then we we need to add the packing information so the number of boxes the weight the the cubic meters the volume in cubic meters or you can just short close it because it was not found in that location once the the picking and packing is is done we can then create the shipment so based on the origin facilities you then have okay I have two consignments that I need to send in this case both are going to the same place let me try to find one with multiple destinations here yeah this one so we have multiple consignments and multiple destination facilities and you just mark them okay I want to send all of these and you have your shipment information that you can then use to to program the couriers and and things like that so the managed shipments please once the shipment is done you have these different status that is waiting for transportation data where you define the let me see if I can find one with some data okay where you have the information on the consignments that are allocated to this shipment and you can add the transportation details so the plan departure date the courier name the license plate and the loading capabilities and after that once it's the transportation details are saved you just need to go back and confirm that this information is is correct and the the shipment is now loaded and in transit okay you know I'm going to have to cut you off there but I think it seems like a that you have a very clearly defined and comprehensive process built into the workflow just to reiterate that these you know the reason that we're we're letting Medexas go through these functionalities is that these are functionalities that DHS2 does not have and DHS2 will not have at least in the near future in the foreseeable future and so if you if these are functionalities that you need to manage your supply chain you need to look to using tools like Medexas to to complete these to be able to cover this this use case there is one question from last question from Sophie which I will answer they are asking if we could use DHS2 data like caseload data to inform supply and ordering and supply chain data yes we want to be able to do that in the future we're going to be working closer with Medexas hopefully in the barundi use case maybe some other countries to be able to push and pull data between the systems so that potentially in the future we could be using caseload data instead of consumption data for some of the resupply and and calculations going on within Medexas with that laundry Nuno thank you again so much it was a really wonderful presentation it have has been recorded we'll post it up on YouTube we will also put it on the Google Drive Nuno special thanks to you for incorporating the word of the day into your presentation and again that word of the day is ain't no mountain high enough please go and mark your attendance type in ain't no mountain high enough high enough just like the song and and that will get your credit for being here today yeah no thank you Martin Alice any additional things that I'm forgetting maybe we can remind that the experts lounge for Asia will be held tomorrow instead of today right yeah so that was scheduled for today we're gonna be able to do it tomorrow so if you're joining us from Asia or you're just interested to find out a little bit more about what Laos has done in their use case John Lewis from his Vietnam will be available to answer those kind of questions for you with that I would say that all the presentations will also be uploaded so that you can get the contact details from any of the presenters that you've seen here today if you'd like to reach out to them and talk about maybe using Medexus or understanding a little bit more about the M supply use case in Laos they have made those contact information available through their presentations with that we I know we're a bit over time thank you for your patience thanks for staying on the line we will be sorry just one thing I'm going to copy in the chat the link to the feedback form oh right yeah I'm always forgetting something the feedback form is very important to us we do take a look at that every day so please do provide us some feedback and we appreciate that sincerely yes we do and then anything else Alice are we okay I think we're okay now okay all right thanks for keeping me on track um then we will call it a day again thank you all for joining day four tomorrow we will be right back here at the same time 11 o'clock Oslo time and we will be going through how to actually configure DHIS to to be able to produce some of these indicators that you've seen Malawi and um and Laos show so that'll be exciting for you to be a bit more technical so if you're not a DHIS too technical person you might get lost quite quickly unfortunately but if you know how to configure indicators already you know a little bit about the DHIS to architecture you should hopefully be able to hang with me