 So if we look at our trainer's guide again, and we'll see that we have the program TB treatment card here, and the indicator is hospitalized both at initial and continuation stages. So how do we actually go about doing this. Well, let's look at our program to begin with. And we can see here that at the initial diagnosis stage. That we have the, we actually have two different data elements here. And we look at the, the type of treatment, we go here, can see that the type of treatment for hospitalized. This tells us, this tells us if it's a facility based or community based type of treatment, or if it's self administered treatment. But this patient has been hospitalized and we'd want to know in our indicator, how many patients like entered this program, hospitalized, and we're still hospitalized during the next phase. So if we look at, if we look and open this continuation stage, we can see that the type of treatment is hospitalized still. So we want to sort of see this cascade of care, or this, this linkages between the initial diagnosis stage and the continuation stage. How many people were still hospitalized, right. And so this is a little bit different from the previous indicators, which we're only looking at data which were entered into one discrete event. Now we have to look at data that I've been entered into two different events. And so how does this work exactly. Well, let me go back to our program indicators here. So I want to walk through how to create this program indicator, and then we'll take a short break and you can do this enrollment type indicator on your own. And then we will go through two additional exercises about enrollment indicators after that. So hold off on creating this one right now, just absorb the different steps that are taken, and then you can do this one on your own. Okay. So in the very first phase here. Now we're going to be naming this our initials then hospitalized initial and continuation to different stages continuation one stage. And then we're going to say, you know, all treatment. And then all hospital for code or something. Right. Once again we'll skip over the color and icon options that we have here. In the description, we can say a little bit more about this. So patient must be hospitalized at both the initial case report and continuation stage. Then we'll skip the decimals because this is essentially count. And for the aggregation type. Again, we are still counting up the number of patients right so each enrollment would only get a count of one. And if we are counting this up the organization and hierarchy, then this would still be a account of one, we're counting the total number of patients. So now, when we go to analytics type here. Again, this is where things get a bit different than we've previously seen. Sorry. So we have the events type, but these data come from two different events. So event type won't work for us here. Now what we use is the enrollment analytics type. And you can see that the analytics period boundaries are also different here. So now our boundary targets are the enrollment date after and the after the start of the reporting period and enrollment date for the end of the reporting period. So much more detail again on analytics period boundaries. But I think it's important to note at this at this stage that the enrollment date will will tie your case to the period. So if you had a patient who was enrolled in January, and then their continuation visit was in February, a month later, and then you were looking at the total number of cases in the pivot tables. This patient would count for January, because their date fell in January. This sort of hints at what analytics period boundaries get into and what they allow you to do. So for now, let's move on from the analytics type and the analytics period boundaries and go to our expression. So once again, we have a question of what we're going to be putting into the expression. Right. So what is the final calculation that we want to make on this enrollment. I'm not going to put the event count now right, because this actually isn't counting number of events that meet this criteria. This is counting the number of enrollments that have these data values for hospitalized in the initial case report stage and the continuation stage. I'm actually only going to count once per enrollment. So what I need to put here is not events count. I need to put is enrollment count. And so this is going to output a value of one for the enrollments that meets these criteria. Okay. Now let's go into filter. So this is actually a bit similar to what we had with our previous program indicators, except for one key difference. Now our data are coming from two different program stages. So I select initial diagnosis here. And I say that this is a type of treatment. The initial diagnosis stage right eggs hospitalized. I'm going to check the option set just a second but just to demonstrate this here. Right. I'm just wondering what these encoded values are here. What is this Jewish. This is actually the UID for the stage where this value comes from, and the UID for the data element where this value comes from. So we will see here that when we say initial diagnosis stage type of treatment equals hospitalized, and then we add continuation one stage type of treatment. I will say, and here someone's drawing on the screen. Please stop that. And if I say the continuation stage equals hospitalized. We will see that the UIDs for the data element, right, are the same. So this is the same data element type of treatment, but the UIDs for the stages, these part here. They are different. And this is also reflected in the interpretation that we have of our filter here. That's made for us. So type of treatment initial diagnosis stage equals hospitalized and continuation one type of treatment equals hospitalized. Okay. Anything else to add here I don't think that there's anything else that we need to add for this one right now. So I am going to click save. And then you can make sure this was noted. And now I needed to make sure that the option for this one is actually a hospitalized. So I'm going to go into the options that management and see that the T type treatment type equals hospitalized. Yes. So the name for this option set, or for this option, the option set is the exact same as the code, which is how the data are stored in the DHM ST database. Okay. So again, going to save this. And I'm going to do what we did earlier and check the pivot table. I actually had it here already. So I'm going to refresh and make sure that I can see these hospitalized initial and continuation hospitalized initial and continuation. And we can see that the program indicator value that we just generated together is the same as the correct value that was a pre generated. So we followed all the right steps to calculate this fine. And we can interpret this value to say that 21 patients who were enrolled in 2021. They were hospitalized when they first entered the program. And during the continuation stage, they were still hospitalized. Okay. So now I'm going to pause there. And you can walk through the steps for creating an enrollment type program indicator on your own. We will come back just in in five minutes and see where we are with this and go into a couple more advanced enrollment type program indicators. Examples. We can move on to two additional examples now. And for those of you that are still struggling with any of the first four indicators that we have discussed. Again, please feel free to mention that in the zoom chat or the Slack channel, and we can help you with them. So, before we begin our next program indicator exercise on weight change. I want to review something from the slides that were presented earlier. So if you see that there are some specific scenarios in which we might be able to use program indicators. Right. And I said that we can perform a percent calculation which values are stored in the same enrollment or change in value across an enrollment. We might want to compare values that are entered for the same enrollment. And we can even as we saw in in the maintenance app earlier, show this program indicator in the form so that you can dynamically see an indicator updating while you are creating a enrollment and entering data. So we were discussing we didn't go so much into filters and what they are, but a filter essentially specifies exactly what we want to include in our program indicator, right, which events or which enrollments do we want to have included. And so filters are applied to the set of events within the program. Before the indicator expression is being evaluated. Right. I think this part is, is pretty key to understanding, because we will be doing something a little bit different with the indicator expression in this example. So again the filters are applied to the set of events or enrollments within a program by going through each individual event before the indicator expression is evaluated. If it evaluates to false, then the event or the enrollment is ignored. So when we look again at this program indicator, this is the percent weight change across the TV program and the aggregation type will be the average. So these are a couple of things to keep in mind while we construct this indicator. But just to get started, there are a, actually, I think there's another couple of additional points here as well. This should calculate the difference in weight between the initial diagnosis and end of treatment program stages. When the cultural result is negative and the treatment outcome is either cured or completed. So this means that we also need to include a filter here for when the cultural result is negative. And also in the filter for the treatment outcome is either sure or completed. Okay. So I'm going to go back to our program indicators here. TB treatment card can see all the work that's been done already. So I'm going to create a new program indicator for this changing weight. So maybe the details here might be initials, weight change, and of treatment and initial. And also I can see in here just as one quick example, sorry to jump around, but initial diagnosis you see it has weight here and the initial diagnosis stage. And also at the end of treatment, there's weight as well. Okay. So we're going to be drawing from these two different program stages. So the short name, I'm just just make this weight change. Now the description. Wait. The end of treatment and initial diagnosis, excluding, let's see this from earlier, the cultural result is negative. Sorry, excluding it's only when the cultural result is negative and long patients who are cured or completed. Okay. So again, here. We have a few different components for this program indicator, right, which make it a bit interesting. So here the aggregation type is actually not going to be the count, because I don't want to count the work. I don't want to count the number of cases that had a change of weight during their treatment. I don't want to count actually went to average the average change in weight for each patient from the beginning to the end of their treatment cycles. So I don't just want to account to give me the number of people who added weights. That does not very interesting. I don't want to sum it up to get the summary of changes of weights. I want to get the average patient weight. Again, this, this analytics type is not going to be event, because the data are found in multiple discrete stages. So we're going to use the enrollment analytics type. And again here we can see that our analytics period boundaries are from the enrollment date. After started reporting period enrollment date before the end of reporting period. So, remember earlier we saw that we would skip this display inform checkbox. Right. So we can think about what this actually means is that if we go into our capture app. Right. We can see that there is a number of different boxes here that might show up and looking into show had widgets. There's also a indicators box as well. Here we use this to display the patient's age for example. Right. But we can add a number of other program indicators to this box as well that might dynamically update as we enter more program data. Okay. So I'm going to check off display inform, just to see what happens when we enter in data for this program indicator, while we are entering in, while we're entering data, what happens is program indicator. Okay. Again our expression is going to be unlike the previous program indicators that we've worked with. Right. Because the expression is going to actually be actually going to be the difference of weight from the end of the treatment to the beginning of the treatment. Right. So, again, difference in weight between the initial diagnosis and the end of treatment. She'll diagnosis. Wait. So diagnosis. End of treatment. Now in the filter. So the final output will be this difference between these two. These two data element values bound in two different stages. Now the filter. Remember from this previous discussion on filters is that the filters applied before you do any additional calculations in the expression. So we can think about working from the filter backwards. So, again, we are going to be we're only going to be including those that have a negative. We're going to be sorry. Excluding those, including those when the cultural result is negative. And the treatment outcome is either cured or completed. And then we're going to add equals. So this part is fairly straightforward that the culture results that we see in the tracker capture up, I go back to the tracker caption. Yeah. The result is negative zero colonies. I think it's actually the same here so it's negative zero colonies, right, negative zero colonies, and the treatment outcome is either cured or completed treatment. Right. So now I'm going to add the second condition to this filter. So that's, and these two additional conditions are using an or statement, right. So it's very important that we consider using brackets appropriately to make sure that we need both negative and a second condition, which is either two different data values. So, either treatment treatment outcome. Yep. And the treatment treatment outcome equals cured. And I add an or or and I can just add that again here or treatment outcome equals completed. And I'm going to close the parentheses. So before we do this calculation on the, on the weight change that was made during the progression of treatment. First we're going to eliminate all of those enrollments that do not meet these criteria. So if, if the initial diagnosis. Was it not, was it was negative. And the, it's not negative, and the end of treatment outcome was not completed or cured. So now I'm going to save this and we're going to see what it looks like in the program so I'm going to clear my cash here, just to make sure that this shows up in the program. And then we can see what happens now. MTV treatment card. Register a new case. Test rate change. So treatment type of patient is new. The disease site is extra pulmonary. I'm going to say the cultural results is negative zero colonies. And I'm going to say that the rate is eight. The initial stage. Okay. Is that the initial diagnosis. Maybe I'll make this one day before. Now I'll just make this continuation one stage. And I will change the weight classification. I'll say that this is me I messed up the options for this indicator. So I will just quickly check that. That's probably it. Okay, so now I'm just going to check the, the options again. Option set is for culture results equals negative zero colonies. The end of treatment treatment outcome. I will check that options that was actually treatment completed as well. So this was actually different. I'm just going to check the, the option set codes. I think I also needed to just say what the final way change was here as well. So I'm just going to clear the cash and come back to this last test case. Test, right to change. New rate change. One hundred. But I actually need to go to the end of treatment stage. Save there. And we're going to say that their rate change is now 100. And their treatment outcome is cured. Change in weight for the end of treatment and initial diagnosis. Only when the cold result is negative and patients who are cured or completed. Or treatment completed. Okay. So this is taking us actually. It's a difference between the weight at the end. It's a difference between the weight at the beginning. So if I say that this was actually 120, then you'll see that this changes to a positive number. So now we can go through doing this one on, on our own. And then hopefully if we, if we have time in, in the next 10 minutes or so, then we can get to these super indicators. So I'll give you five minutes to try to catch up to this. And then I'll give you the final section on combining indicators together. Okay. So I see that a, a number of you have managed to create a program indicator for the weight to change. Sure. Log in. And if you have managed to make a permanent care for the weight change. That's great. We can also take the final step here. And then we can also take a look at what we have created in the tracker capture data entry form, but also looking at what it looks like. Once we aggregate up. Okay. So I will just look here again at the different program indicators that we have created in the pivot tables app. And I will see the weight to change. End of treatment and initial diagnosis. I think again, we have a look at this. So the weight of data is going to be around earlier. It's full of. Zero. It should be calculating to something. It's more than that. But it actually. I'm not sure why that is maybe there weren't data for that. In those, those periods maybe I'll say. This year as well. Well, okay. So we can come back to that maybe. what happened here. But if you look at the weight change final mass initial, yeah, that one was calculated correctly. We can see that there is a difference in the weight change for 2021 and 2022. So we can see that this is an average over all of the different cases here. Yeah. So one last program indicator that we have here is the combined indicator. And for this one, we will actually need to be using DHS2 analytics. So we won't be testing this one just now, but it's still helpful for us to go through the steps for for creating this indicator together. So again, the program is the TB treatment card. But now we're going to be building a percentage and that's a TB incidence rate per 100,000. So again, if we look back to down to our, can look into more detail about what this is with a TB incidence rate per 100,000, we want to calculate the TB incidence rate per 100,000 population. This is equivalent to the number of new TB cases divided by the total population times 100,000. So in order to do this, let's walk through some steps here. We will not be using the program indicator section here because we already have a program indicator for new TB cases. Look here, TB new cases. So this one has already been generated and it counts the number of new cases that have been shown. So TB patient type equals new. So we're not going to do that. We're going to skip ahead just a little bit and we're going to create a new indicator. So that means that we're leaving the realm of program indicators and we are now within aggregate indicators. These are generally used for percentages, for example. So here for the name, you might make this one TB incidence per 100,000, TB incidence. Again, we're going to skip the color and icon. But if we look at the description here, we can actually just copy and paste over this longer description that we had, number of new TB cases divided by total population times 100,000. We'll also skip over annualized here. There are some interesting things that we can do with an annualized type indicator. But I refer you to the documentation for how to do that. It's a bit of a specialized statistical term that we don't use as often. That's most in data input. We will leave this as the standard. And now when we come to the indicator type, this one is pretty essential to just double check. So here we have the different types of indicator types. And that's per 100,000. So you can create your own indicator types in your DHS2 system to be percentages or per 1,000 or per a million, for example. And when we click down here, we can see that the demo system already has numerator only per 1,000 per 100,000. And we're going to make this per 100,000. We will skip over the legends, URL, et cetera. But what's really important about the indicators is the numerator and denominator. So we're going to go to the numerator here. And those of you who are familiar with the DHS2 aggregate indicators, this might look a bit familiar because it's essentially the same process. But we will create the description for the numerator as new TB cases. And we'll go to programs. And now we can actually select a program indicator that we've just created or we already have generated for the TB treatment program. And we can use that within our percentage indicator, within our indicator for prevalence of TB. So we have to scroll down here, but we can see all of the program indicators that we've already created. And I go to TB new cases. Right? So I move that over. You can see that this is the UID, the code for the program indicator for TB new cases. And the preview for what this is shows up here. And we can see that this is also a valid expression. There are a number of different things that you can also do when calculating indicators, which again, we'll refer you to the documentation for things like is no values or creating logs or period offsets. These are a bit advanced for what we'll do today. Also things that you can include like ordinate counts for different areas, constants, data element values, and try to identify attributes. But for our purposes, we're mostly interested in comparing a program indicator value of the new TB cases. So I was like done here with a population level denominator. So going back here, we can see that the denominator will be the total population. So this denominator for total population might be entered as part of a census or it's a standard part of your HMS. But this is essentially getting a sense of the catchment area or total population for a service area that is administering to TB patients. So for this, we're going to go to data elements over here once again. Right? And I'm going to select total population to start searching for it. And I'll show up here, total population. So I move that over. And this might be like an annual or a monthly data element or a data set that can be entered that will have the total population for a given organization unit area where that's at the district or national level. Okay. So now I'm going to select done. So this description again tells the whole story here. So it's a number of new TB cases divided by the total population times 100,000. So I've just saved that indicator on the incidence per 100,000. I also need to add my initials to that. I apologize. So because we've just created a indicator and not a program indicator, the indicators are not generated on the fly, which means that we cannot create the same way that we just tested our program indicators. We created one and then we immediately checked the PivotTables app to see the results. This isn't a dynamic query that's generated in the database when we request information on the indicators. We actually need to run the analytics tables in DHIS2 first. Some jobs. So we'll go to administration. So we're going to go into data administration and we will run our analytics tables. This should be quick since there's not a whole lot of data, but depending on the size of your DHIS2 database, this can actually take quite a while. So generally, it's best to produce and test these types of indicators on a development environment and then import them into your production environment. Otherwise, the testing process can take quite a bit of time. So we've just completed updating our analytics tables. It took about 20 seconds. And now we can see what this looks like in the PivotTables app. So here's indicators. I'm not going out of program indicators. Didn't put it into a group. TV incidents per 100,000. And I'm going to select as my period of last year and my organization unit as training land. So let's see what happens. So I have actually configured this program indicator to be TV incidents per 100,000 to match the preconfigured indicator. So the correct value is 15.6 of the population times 100,000. So the incidence is much lower than 15.6 percent since it's per 100,000. As we mentioned, there's other ways that you could use this for not just having a population denominator, but if you were to go into the indicators and see it's available, for example, on a denominator, you could also put something like new TV cases in the denominator. And then the numerator, you could do another type of disaggregation that you might have had earlier. So new pulmonary cases, for example. TV new cases, yeah. New pulmonary TV cases. Just add that over here. So this would give you the percentage of all new cases that were pulmonary in a given month and organization unit, given period in organization unit. You can say percentage there, right? So that's another example just very quickly of how you can use indicators in combination with program indicators to provide an additional level of analysis on your data. We have about 15 minutes left. So I will now hand it over to Shurijit for the final closing. Yeah, thanks, Brian. And what we'll do now before we close, we'll just give you kind of a couple of minutes to have a look at the exercise for this one. We would really appreciate if you don't go into analytics and run that. If you all run it, the system's going to just fall over for sure. Okay, so in the learner's guide, it kind of gives you the steps to follow when you're doing the exercise. And we'll give a couple of minutes to do that, then I'll explain. So we'll give about, you know, five, seven minutes for that. Okay. And then we'll come back and I'll explain what we'll do when we come back together. All right. So this is the final exercise in the learner's guide that you can have a look at here. I'll pull it up.