 And now we're going to get to our data collection tool. Now, this really depends how much money you have, how your unit is set up, how much research has been done there. In most areas in the world, we are severely resource constrained. We do not have research officers, people to do data collection, nothing like that. It's only units that have got a good track record, that's published a lot, or are in big first-world kind of ivory towers that have good data acquisition tools and have data available electronically already, et cetera. In most resource constrained environments, this is not the case. So what I'm going to suggest here is really the bare bones minimum that you can do yourself, that anyone can do irrespective of their resources as long as they obviously have internet connection. And from this base, you can build and build very fancy tools. So what I'm going to suggest here, you see there is just using Google Forms. We're just going to use Google Forms. Now be very careful and careful about what you're going to put on your Google Forms because that is putting patient data on unsecure servers, to some extent at least unsecure servers. So you've got to be absolutely sure that you are allowed to do this, get permission to do this. And it is of so much or utmost importance to protect patient identification if you use a tool like this. And for that matter, any kind of tool that you use has got to protect patient anonymity. And this is how we're going to go about it. On the left hand side, we had our data required. Then you see two other columns that I want to discuss with you in this section. We'll have our anonymity format and you see the data point values. Now the first thing, of course, we identify patients from the admission record, from the clerks notes or from the hospital's electronic record system, whatever the situation is. So we've got to have a patient. Now that can be a patient name and surname. It can be a patient national identification number, anything. But we never want to capture that kind of data in our data collection tool. No matter how fancy it is, it is wrong to do that under most circumstances. So how do we anonymize that, especially if we want to use an open tool like this? Well, we're just going to have a patient ID. And that is just going to be a natural number. Natural numbers we know starts at one, one, two, three, four, five. So every patient that I admit, it's just going to have a number starting from one. That would be the easiest way to do it. Very separate from this data collection tool. Even if it's on a piece of paper, that isn't anywhere near this data collection tool, I might keep a record that says, patient one is John Doe with that hospital number and that national identification number so that I can refer back to it. But that is never captured on this main data collection tool. That is kept absolutely separate. We've got to identify everyone who's involved in this research. We'll identify who's going to keep those records. And perhaps even in the staying age, as a basic set of information is to keep that separate and on paper, somewhere locked away. Of course, build up from that the more sophisticated your system is, you might have better ways of dealing with this. But if you are resource constrained, that's going to go on a piece of paper and that is really going to be hidden away and held quite separately from all of this data analysis so that we know if we wanted to look at something specific, we can go back and see patient number three was so and so and we can go back to that file. But as far as our data collection tool is concerned, we'll only have patient number one, patient number two, no one will know who that is. Next up, remember is our infection category. We need to know if it's major or minor, but we will not put in our data collection tool the word infection and then have minor and major fault in on our data collection tool. No, we'll write something like cat one for category one. You can choose whatever you want as long as no one can realize that cat one refers to infection or type of infection. No one should be able to draw that parallel between the two. So if someone gets hold of your data collection tool, it just says cat one. No one knows what cat one is. And the data point values that are entered into the cells in your data collection tool in our instance here, we're going to use Google Forms. We'll just use something like minor as A and major as B. So every time it's a minor infection, we might write A and every time it's a major infection, you might write B. You can come up with whatever you want. You can call it grass and sky. You can do whatever you want. You can have more than one, more than one things that refer to minor or major and you'll see in the next row there, we're actually going to do that. This is the basics of A for minor B for major, make it more complex than that. So here we get to gender in our form where the data collection goes, we're going to call it cat two, call it something that's even worse than that, just so long as you know what cat two is and that it refers to gender. And here we see on the right-hand side, so we're going to use either CX or R. So if there's someone who's collecting that data, can either write CX or R when it comes to female and they can write FL or B. Now that's a study thumbs up and you can make it much more complex than that. So if someone sees that list of data, they're going to see C, F, X, B, R, it makes no sense, okay? And as I say, this is just a baseline. Please make it more complex than this. This is just for illustrative purposes in this project of ours, Julia project of ours. When it comes to age, we might write var one, variable one, you can write whatever you want as long as you know var one refers to age. Now this can become complex. We can do large studies of course with lots of variables on which we're going to collect data points. So once again, you can keep separate from your data collection tool what these anonymized column headers are going to be called for. Again, for the argument sake, just for this little project of ours, we're going to call it var two. Now what you can do again, baseline, please make it more complex. We can all, everyone who's involved in this project can decide whenever you jot down a patient's age in the data collection tool, mentally add the value of five. So if someone is 33 years old, make them 38. And so you go for everyone. So that's a little mental arithmetic. So that's one way to anonymize the data. There are the age variable at least, there are many more ways you can make it much more complex. But do not jot down the patient's actual date of birth in any form and not age either. Especially at extremes of age, people can be identified because of the extremity of their age. There is not a lot of 99 year olds running around. So come up with some way of scrambling that data age for argument sake, for the sake of this research project, we're just going to add the value of five to all the ages, which means when we do our data analysis in Julia, we've got to subtract five from everyone's age. And that's very easy to do in Julia, no problem. HBA1C, we're going to call it var one. CRP, we're going to call var three. And in this little project, we're just going to collect the actual data points. Very difficult if you get a hold of lab values. A lot of patients are going to have HBA1C values of, say, for argument sake, 4.4%. And a lot of patients are going to have CRPs of 345. Well, I should use a more common, say, let's say, 55. In the real world situation, though, I would also scramble those two data values by some simple equation as well. But for this project, we're going to do that. So in the next section, we're just going to go to Google Forms. Most of you will know how to use Google Forms. We're just going to set up very quickly a part of this data collection tool before we go over to Julia and do the interesting, but do the actual analysis.