 And just a quick run through for the agenda today. In the first session, I will see how we can configure validation rules related to internal consistency, which is comparing two data items in the same form or in different forms, but the ones which we are reporting in DHRs too. We'll follow that with an exercise. How can you create the same validation rules yourself? And after a short break, then we will discuss creating validation rules using threshold values. For that session, we'll quickly run through the predictors as well, and then focus on adding a validation rule related to the predictor values, again, followed by an exercise for today. So let's go to the DHRs to demo now. And I'll take the same examples as yesterday. So maybe I'll just recap the example for HIV cases that we took yesterday so that we follow the same. As we had seen yesterday, there was a validation rule which said that HIV test positive cases should be less than or equal to total test performed at the facility. So if I have by mistake maybe added an extra 0 here, and I tried to run validation. So it shows me the error that the positive cases should be less than or equal to total test performed. So let's try to add this validation rule. But adding the validation rule, we'll go to the maintenance app. There are many options that you will be able to see. You might have access to some of these. You might not have access to a few. But we've given you access to the validation feature. If I go to validation, I can create a validation rule or a validation group or a validation notification. So to add a new validation rule, maybe if I want to see the existing validation rules, I can just click on this list icon to see all the rules which are existing. As you know, this HIV rule is already there in the application. So for this demo, I will create the same rule, but I'll prefix it with my initials so that there is no duplication in the names which are there. To add a new rule, I can click on plus here. Or on the previous page, I could click on plus over here. Now these are the details required to add a new validation rule, name, short name, code, description, instruction, importance, period type, operator, org unit, and skip this rule. So we'll go to all of these one by one. I'll give a name to the validation rule. My rule says that HIV tests should be less than or equal to HIV tests. I can just copy this as the short name. I can give a code if I want to, or I can skip this. Description, now this description, this is what is visible when we click on run validation. So I will give the same descriptions. And these are the instructions. So when we look at the details of the validation rules, these are visible to us. So we can give an instruction which the person who runs the validation, he could further assess what happened and why is this validation error coming. So maybe just an instruction to them, we can say that make a visit to the facility or call the facility or follow up with the facility where HIV test positive is more than HIV test performed. Where HIV test positive cases are greater than the HIV tests so that they can cross check or follow up with the facility. Now we can give an importance to the validation. So we have three options, high, medium, low. If we think that validation alert is of high importance, maybe if we think it is a possible outbreak or something, we can give it a high priority here or we can just say medium or low for them. So maybe for this, I'll just say this is of medium priority. The period type here needs to be same as the data entry or the data set. So that it validates for the same duration. So if I'm reporting my data for HIV test done and HIV test positive monthly, my validation also needs to be configured monthly so that it could accordingly compare the data between the two data items. So I will say monthly. Now yesterday as we have seen, there is a left side expression and there is a right side expression for the validation rule. And in our case, our left side expression is HIV test positive. So I click on left side expression and I assign the HIV test positive. Now, how do we do that? On the top, you see this option of missing value strategy and there are three options. Skip if any value is missing. So if I want my validation rule to be configured in a way that if one of the values between the left side or the right side, if any one value is missing, I have value for HIV test done but I do not have value for the positive cases, I can skip this test. So skip if any value is missing. So this can be done when we know that our validation check can only be done if both the values are present, we can use this. Skip if all values are missing. If both the size left and right side, both of them, the values are missing, then we can select this option to skip the validation rule because not to necessarily show a validation error when you have not reported the data yet. So I can say skip if all values are missing or I can say never skip. So always run this validation. So for this rule, I can just say never skip because even if both the values are missing, it will not show an error because both will be considered zero and the error will not appear. I need not say that skip if any value is missing because if the tests are performed only, then the positive cases will be reported. So in my case, I can just say never skip for this validation rule. And then I have to give a description for my left side. So my left side is HIV test positive cases. Now here I have to assign or select my left side which is positive cases male plus female. I see data elements, programs, or unit counts, constants, reporting dates, here in the selection menu. Since these are my data elements, I will search for them from here. So I can search for HIV tests positive. Now we see three options for positive, positive male female. So this positive is actually the total of all the positive cases. And these are the disaggregations. So it's always suggested to select the disaggregation so that my validation rule adds up both male female and then does the calculation. So I will select both of these data elements, HIV tests positive. I will double click on this data element. Then you see these operators here. I'll just click on plus and HIV test positive female. Now you, what you see here are UIDs for data elements and this might not be very clear on what we are selecting but we need not worry about that. If you scroll down a little, here you will be able to see what is your selection. So I have selected HIV positive male plus HIV positive female on my left side. Let's quickly look at the other options here. You can also select data items, program indicators or number, value type data elements from the program. So you can select the programs and then accordingly make the selection. Our instance does not have any tracker programs right now but in case you have to make a correlation of validation rule, say my total cases or total child population under one year is an aggregate data element while number of children given doses for a particular vaccine is calculated from the tracker capture, from the aggregation of case-based trial cases. I can aggregate that data from tracker, use that program indicator from the programs and then use it for my validation rule. So I can create validation rules through the event data elements or the tracker data elements using the same app. Then we have our unit counts. If we want to compare the count of a particular type of facilities in our validation rule, we can do the selection from here. There's an option for constants. So if we have some constant already available, which is to be used in our calculations, we can select the constants from here. And if you have to use reporting rates, that can be done from here. These are little complicated and not used on routine configurations but data elements and programs are used more frequently. And for our session during this academy, we are focusing mostly on the aggregate data elements. So that's what my rule is using right now. HIV test positive, male plus female and safe. So I have saved my left side here. What I had added in the description is available here. Now this is my left side. Now I have to select my operator. So my operator here is HIV test positive should be less than or equal to HIV test performed. So I will select less than or equal to. Then we have the right side expression. I will select right side. Now my right side is HIV test performed. At any point, if I have confusion on what needs to be selected on the left and the right side, I can look at my data entry form and look at the specific data elements which are there. So as I know in my data entry form, also I'm using male and female segregation for the test done. So I will search for these data items, HIV test performed male plus HIV test performed female. And I can cross check by scrolling down that I've done the right selection. And I say, if I want my validation rules to run only at particular levels, I can just select the all unit levels here. Otherwise I can give this as usually not needed. And if not required, I can just skip selecting anything there. And if I don't want this rule to run during the form validation, I want to run it separately when I do the run validation. Through the data quality app, I can just say that skip during the form validation. But I want to see it during the form run validation. And then I say, I will just check. Let's say when I'll just cross check because it was still buffering on my screen. So here is my validation rule. A quick review again. So I gave a name to my validation rule. HIV test positive is then equal to test performed. I added a description, an instruction, importance, period type. It should be same as the data set type period. Left side expression, operator, right side expression. I'm safe. Now let's just quickly test if this rule is working. So to run this rule, I will go to data quality app. In the data quality app, validation rule analysis, run validation and select one region. Okay, before that, I can still run this rule from here by validating, but it'll take a lot of time because there are so many validation rules in the application. So to run these rules specifically or make them specific to the program so that the HIV program person only has access to the HIV related validation rules, it is a good practice to also add them to a validation rule group, which I forgot to do earlier. So we have this validation rule groups and based on my validation, so if you see there are like 70 validation rules, I don't want to run so many of them and they might not be relevant to me. So I will just go to validation rule group and add them to a particular validation group. I could either add a new group by clicking on this plus icon. HIV validations. I can give a code or description. Validation rules for HIV. And I can select my validation rule. I will just add this one and save. Alternatively, I could just select any of the existing groups and update. So if I already had a group for HIV and I'm adding a new rule, I can just edit the earlier group and add my rule over there. Okay, so I added a validation rule and I added my validation rule to a new validation rule group. Now let's go to the data quality app to run this validation. Run validation. Here I'll select first March to May and only the rules in GA HIV validation rule group. It shows similar, that means my rule is working as it should. If I go to the details, I see the description over here and the left side and the right side. Any questions, anything you want me to repeat? Or maybe my co-facilitators could let me know. If not, then I think we have... Let me just check. Okay, we have some time. So let's try adding another validation rule. Or maybe we could give you more time for the exercises because the next rule that I could demonstrate to you is the one which we discussed yesterday as well for immunization, where for DPT doses, I want to see the doses given and compare it to the stock values for DPT. So the logic for creating the rule is same but it has use of lot of logical operators, the calculations. So you need to like consider one, two, three, four, five and six, seven and eight data elements while you just calculate the values for the HIV DPT doses used. We also have a similar example in the exercise. Maybe we could... I think there's a question we could take that and then we perform these exercise. And if there are issues in adding this validation rule for stock, I could add one for you and demonstrate after you try it yourself. Okay, there was a question. Yes, hello. Yeah. The validation rule, the importance part where we had... I think we chose medium. So I wanted to know what's the implication? Implication as well. So if I have like 40, 50 validation rule alerts coming to me while I do this analysis, I will see the importance of them here. So maybe if I have a lot of validation alerts which are of high importance, I will look at them first, then go to the medium ones and then go to the low ones. So this is just for prioritizing what alerts need immediate intervention and for what you can wait a little. So just to prioritize. Okay, thank you. Any other questions? No questions. Go to the EDX page for today's session, Validation Rules Configuration, Validation Rules Config. You could download the exercise from there and once you download the exercise, we need to perform the first two exercises right now. So the first one is to create a rule similar to what I just created for HIV data set. And to review it there. And the second rule is related to stock for the DPT doses. You could try doing them and we could add another rule for stock if you guys are facing issues there. So go ahead, log into your exercises link and use the same usernames as yesterday. And you can raise your questions on this lab. Thank you. Hello. Yes, please. Hello, I have a question. Yeah. I see there are two operators. I cannot understand it. One is compulsory peer or another is exclusive peer. Could you please explain those two operators? Sure. I'll just go to the Validation Rule. Look at the operators which are there. We have equal to not equal to greater than, greater than or equal to less than, less than or equal to which are self-explanatory. Then there are two more operators, compulsory peer and exclusive peer. So compulsory peer means if you are entering value into one of the value, even one of the data element you must have and have to enter value into the right side data element. So if left side expression has a value, right side expression should also have the value. So that is the compulsory peer. Exclusive peer is only one of them could have values. It could either be the left side or the right side which could have values. I do not remember any examples for this, but maybe I could just look into my notes and get back to you with some examples as well once you do your exercises after that.