 Also, I want to make sure we don't get confused. There has been a problem with Hoonette that I've mentioned previously about debase files, about the old debase files and now the SQL life files. We're trying to get the migration over. I'm hoping, so right now it's working very well but there are still issues. So basically my goal before the end of the August is that debase files will be 100% compatible again. Right now that is not the case. Right now, because of the debase compatibility issues, we use a software, a Microsoft software called the Microsoft Axis Database Engine to use debase files. And on a lot of computers, because of compatibility issues, that software is not running properly. So reinstalling the Microsoft Axis Database Engine usually fixes it, but not always. If it doesn't fix it, you completely manually uninstall all of the Axis engines. This file that I mentioned, you might have two or three versions of it. Uninstall Hoonette, reinstall everything. So far that always fixes it, at least in the short term, that probably happened again. So right now we use debase files through this library from Microsoft called Microsoft Axis Database Engine. In January, Microsoft announced that they were temporarily removing support for debase and they have not brought it back yet. So we have started, especially in this one here, this one called update data files to SQLite. We're using debase files, but not using the Microsoft Axis Database Engine. So we are trying to use our debase files in a way that is not dependent on Microsoft. We did this one because we needed this one to do the upgrade. So eventually, I don't know about my computer right now. I'm using mostly SQLite, but if I go to data analysis, if I go to data files, and if I try to open the older version, this is the, oh no, that's not it either. Where's the Hoonette data? Let me change this to all files. W0, 1, 95. Looking for a debase file. Well, here's many. So let me just copy this one. So you drive Hoonette data and paste. Let me see if I can open that or not. So I'm going to go to, so what I'm exploring is a potentially different issue, which is confusing. If I go to the data folder and there. So this is an old debase file. I'm going to try to select that, okay. And let me just do a simple analysis percent resistance, or is okay. I'm going to click on begin analysis. And it will either work or it will not work. So on my computer, it is now working. But because of the debase compatibility issues, what some people get is an error message. What some people get is no isolates found. So both of those things are the debase compatibility issues. So what you're describing sounds different, but I just want to warn you that sometimes people say the file is gone when the file is not gone, but Hoonette is having trouble reading the file. What you have told me is you're not looking inside of Hoonette where these compatibility issues would be important. You're looking at my computer, you're looking at Windows. So if you're looking at Windows, these compatibility issues are not related. But this might be relevant other times. If you click on this button, begin analysis, and it says no isolates found, that could be the debase compatibility issues because we use the Axe database engine. By the end of August, I think we can do this without the Axe database engine. It may or may not be slower because, you know, they were a professional company doing their Axe engines. We're making our own engine, which will be compatible. It may or may not be slow. We just don't know yet. I do know that Adam really did optimize the upgrade to SQL Lite. So I'm hoping it's going to be as fast. I just cannot say that right now. So again, that's not the virus issue. That's a bit detour on this debase stuff. Go ahead, please. Go ahead, yes? Yeah. There is an interesting issue regarding compatibility. Yes. We are actually installing UNIT version 2020 to all phase two sides. So they are entering the data using not debase actually, SQL, something like that. SQL Lite, SQL Lite, yes. Yeah, exactly. So at the end of the day, we are planning to prepare annual reports. So is it possible to combine all this kind of data without any problem? I will say yes. And then I will give a qualification on that. So when I go to here to data files, I can I say all files, all files. I can choose all UNIT data files from 1989, from 2000, from 2020. So all of the files for the last 30 years should be compatible and you can combine them. So don't worry about a mixture of debase and SQL Lite. The small caveat that I want to mention is because of these debase compatibility issues, right now that will work on some computers, but other computers that debase won't work, maybe. But that's temporarily, by the end of August, I think that no computer will have a debase compatibility issue because we will stop using the Microsoft Access Database Engine. So right now what you're saying, you can combine any UNIT files from any of the last 30 years, everything should be fine unless you have the debase compatibility issue, there are short-term fixes for it, but we should have the long-term fix, excuse me, before the end of August. In fact, I think we'll have most of it by the end of next week. I just don't want to over promise because sometimes other priorities come up. Okay, so don't worry, debase and SQL Lite, you can mix them together and there should be zero issues unless you have this short-term debase compatibility issue. Okay, great. I keep on mentioning that file, I'm going to go to Google Chrome, Microsoft, oh, where is my Microsoft? There is Microsoft Access Database Engine. And, you know, so this is the file, so this file is optimized for Microsoft Access. It is also supposed to do debase, it's more specifically a technology called DAO. It's supposed to be using it, but that's where Windows has changed it to temporarily not include debase files any longer. So this is the file that's causing us trouble. Okay, so I'm going back to the agenda if I can find it. So that's number two, yes, three points. Put in an antivirus software, that's irrespective of anything. Number two, and also policies about, sometimes what you have to do is there'll be one computer with an antivirus software, so you have to go to that computer to do your USB cleaning before you put it in the other computers that helps protect the other computers. And also just centralizes the antivirus to one of the computers. Of course, we want antivirus in all the computers, but sometimes you just start on one computer. Let's see, that's one point. Virus is definitely you want a strategy for that. Number two is that you also want a strategy for backups. And if you're having trouble, just make it backups every day into some secure place that you feel comfortable with. Number three, what you're describing is strange. I'm not exactly sure what's happening there. And I'm not sure that I will have any additional useful input on that. It doesn't seem like a who not specific thing, but let me know what you find. Number three. Yeah, maybe let me interrupt you, sorry. When I put the Hoonit software in D, in Drive D, the problem will not be happened. So I just got this one through troubleshooting. When I put it in the D, not in the C, by default, Hoonit will be placed under C. So when I make it in D, the problem will disappear. So if you get some idea from D, maybe, you may say something. So let me see, I'm gonna show you something that's useful, but I don't think it's directly related. I'm here under Hoonit. I'm going to file. At the bottom, there's an option called configuration. I'm gonna click on the word configuration, and here you see the default locations. So by default, the data will go onto the C drive, but you can close and you can put the data on one of the other network drives, okay? So there are two different questions. Where is the Hoonit software and where are the Hoonit data files, so the Hoonit program files? Well, not the program files, the Hoonit configuration files and data files. So when you install Hoonit, so I'm gonna go to my installation here, this PC, users, downloads. So here, it's not gonna work right now because I've already updated my software, but now it's started the installation software. So when you install Hoonit, it does ask you what folder do you want to install to? Do you want to install to this drive Hoonit or the P drive? Here, it's simply saying set up failed, there's no error here. This version of the software is already present, so there's no need to reinstall it. But normally one of the first questions is where do you want to install it? So that's the location of the software. The second question is where are the data files? The default location will be wherever you put the software, C drive, D drive, F drive, in the data sub folder. If you want to change that, then you can put it someplace else, okay? So that might be useful for you as well. I don't think it would explain the problem, but that could be helpful if you want the data over here, but you want the software over here. Okay, I just received an email from Ethiopia. I think we probably all did. I just want to make sure he's thanking somebody. All right, is that home? Okay, he got it, okay. Good, that's, I mean, it might be a virus, but it might be a defect in the C drive. The C drives can, any drive can be corrupted, lost sectors, invalid memory. So sometimes reformatting the hard drive is enough to fix that. It would also get rid of viruses in the short term. It'd be strange that it'd be only who in that, but if that's what you're seeing, that's what you're seeing. Maybe they're not using the other softwares as much. What did you say? You said you put the data files on the D drive. Did you also put the software on the D drive? No, no, no, the software itself. The software itself, because? Yeah, yeah. Yeah, I mean, there are things about protections, but usually it doesn't delete things. It just tells you, you cannot do this. You cannot do that. You cannot do this. Talk to your IT department. Those are permissions issues. I have seen like in my hospital network, you know, we do have things on our H drive and sometimes I turn the computer on and there are new icons there that the hospital installs automatically. So sometimes there are issues like that, which there's some central control over your computer, but that would be unlikely, but it's something to consider. Okay, great. But thanks for debugging. It's nice. If something works on the D drive, but not on the C drive, that is what you need. You have a working software. It would also be nice to have an explanation for the problem, but sometimes we cannot end up, we cannot find the explanation, which is unfortunate. The two situations, it's working and you know why? And it's working and you don't know why. You'd rather it be the first situation. If it works and you don't know why, oh, it's still okay, but it just makes you a little wondering what's going on here. The worst one is, well, then there's things are not working and you know why, at least you know what you have to do. And it's not working, you don't know why that's hard. Is it all, if it happens always, that's easier. If it happens once in a while and it's unpredictable, that's frustrating because it can be hard to fix and hard to understand. Gonna move on to the, on the agenda. We did number two, I'm saving number one for last, because it's very open-ended. You know, we have a lot of time, we can do a lot of things with number one. Number two, we finished that. Number three, fixing error messages. Are there specific error messages that you have in mind? I already wrote, well, the answer is depends on what the message is. If it's an error because of something you did, then you need to understand what you did. If it's an error because of an error in Hoonet, you cannot fix that. You can tell us, we can say it's already been fixed. Please download the new software. We can tell you that we can explain why it is. We can say, oh, we were not aware of that. We'll go ahead and fix it. Or it might be a compatibility issue. It's, there's nothing exactly wrong with Hoonet and there's nothing exactly wrong with your data, but there's a problem of compatibility that Hoonet cannot do what it's supposed to do. And that's the main reason for that right now or these debase issues that we're trying to, trying to have a complete solution, a mostly partial solution at the end of next week and a complete solution by the end of August. So that's all I'll say right now, but now my question for you is that you have specific errors in mind. Are there any screens or error messages you would like to show to me? Yeah, maybe let me. I'm wondering what they had in mind. Yeah, go ahead. I'm going to open my Hoonet. Then I will tell you, of course. Share your screen. Okay, I have stopped sharing my screen. Zilalem is a presenter. Okay. Can you see it? Yes, I can. Yeah. Some general, well, you do that. Some general comments. One thing I always wonder, is it a reproducible error or is it just infrequently? Reproducible errors are easy to explore. Go ahead, yes. Yeah, you know, while I'm doing some kind of data analysis, not always. Sometimes I will get error number, something like that, 72, or any other error number will come. So what I can do, I will use another computer to do that kind of specific kind of analysis. Is that the way I will do it? Error message, let me know. Because we want to fix it and we want to understand it. So some of the questions I always have, is this a one-time error or is it reproducible easily or is it just infrequently for unexplained reasons? So those are three kinds of issues. One-time only, all the time versus once in a while. The next question, is it just this computer or is it other computers? In which case, you think about the possibility issues, you think about the version of prunet. Okay, this is a debase file, correct? Yeah, yeah. Okay, click on okay. No, no, no, I want to stay on the debase file. Does that debase file have to... Yeah, it is debased, yeah. Click on okay. Click on begin analysis. Now it works. Does it work? Right, good. Yeah, it works. Yes, and that's kind of good when it works, but you want it to work always reliably. You did analysis previously with a different data file and it said no isolates found and I was wondering about that because that would be the debase compatibility issue which does not... The debase compatibility issue usually gives an error message. Sometimes it does not give an error message and it simply says no isolates found. And we're hoping to have that particular issue fixed by the end of next week. Okay, okay, okay. Okay. Maybe I have seen from your computer some error numbers. Right, and it just depends on what the error number is, what the error... So do you... I think we've discussed previously how to do a screen grab, correct? Yeah, yeah. Just as a reminder for people, a lot of times people tell me what they are seeing but I would rather see what they are seeing. And you just do a screen grab. The quickest and easiest way to make a screen grab is to look at your keyboard and on your keyboard, on your computer keyboard, there's a key called print screen. So if you click on print screen, I wish they would change the name of that key because it does not print the screen. It copies the screen. So they should just rename that key copy screen because then you could just copy screen and paste it into an email, paste it into a Word document, paste it into a PowerPoint. That's one way. There are other ways. There's alt print, print screen does the whole screen. Alt print screen just does the active window or windows has this thing called the snipping tool, the slipping or snipping clip tool. So these are different ways to make screen grabs. So that's one thing that I can... Then I will see exactly what you're seeing. Can you now go to your... Well, usually you have not had an error on this particular computer before? Yeah, yeah. I have a disk over time. Right. I'm getting you out of time. Go back into your Hoonet folder. Go back into your Hoonet folder. Yes, whenever Hoonet generates an error message on the screen, it also generates an error log. So instead of sending me the screen, you can just send me the error log. So here you have nothing in alphabetical order called, no, no, it's not in the log folder. It's in the main folder. We use the log folder for a different kind of log. So in this folder, there will be something called error log. So if you send the error log, that tells me the error number, the error message, the error routine, and it tells me the version of Hoonet that you are using. It tells me the laboratory configuration. So it gives us good information for us to start understanding what is happening. If you're doing a data analysis, it also saves a version of the macro to help us reproduce what you were doing. So in a general sense, these are my general questions, but if you have specific errors, we need to see the details of the specific errors. So there are no error logs here. Therefore, we have not had an error on this particular computer. Okay, would you like to show anything else about the error messages using the other computer? Or, I mean, you mentioned error 72. Can you go to Google right now? Can you go to the internet? Yeah. And do an internet search. And just search for error, you said error 72. So just search for error 72, visual studio, error 72, error 72, face visual studio, visual studio, that's it. So, and then just do a search on that. Sometimes the error message are very helpful. Sometimes they're very generic. Are you sure it was error 72? Yeah, yeah, 72, yeah, I had 72. And also I had error number five. Yeah, error number five is common. I don't see error 72 here. Can you go down, can you just change error 72 to error five? Error five is one of the more common ones. Error five. Right, so basically you see that first one is from Microsoft, and then you'll see answers from Microsoft, you'll see errors from other people. Some of these error message are very specific and very helpful. Some of these error messages are kind of vague. So this is what I do. Sometimes I know what the problem is or I have a good idea what the problem is. Other times we really don't know, so we just search Google to see who else has had this problem. So the next time you see an error message, please copy screen and or send me the error log. I will try to reproduce it. If I can reproduce it on my computer, great, then that's all I need. We can fix it using our computers. If I cannot reproduce it on my computers, then what we will often do is set up an online session where I can see your computer so that we can work through it together. That allows us to try this, try that, explore it. So the question was what do we do about error messages? It depends on what the error message is and these are some of the things that help us to understand. So it's two steps. Step one is the diagnosis, what is the problem? And secondly, once you have the diagnosis, what is the resolution? And a good place to start is to just read down the current version of Hoonat because we're always fixing things. So sometimes the errors that you found, fine, have already been fixed, okay? Any other questions about error messages? So please send to me the details of the error messages that you were seeing and then I can tell you whether or not I know what the problem is and what we can try to do about it. Okay, okay. Sometimes it's as simple as the network drive is done. You ask for a file on the T drive and the T drive is not accessible. What we try to do is we try to give the user not an error message, we try to give them a warning. We would like to say, oh, the T drive is not accessible. That is a warning. But sometimes, depending on what they did, it will get an error message. So error does not necessarily mean something's wrong, it means sometimes that there's some simple explanation which is not an error. It's just not possible right now. And sometimes people say there's an error and there isn't an error, it's simply a warning. Sometimes they'll say I wanna see isolates that are female isolates you're an outpatient and they find no isolates and they say, John, Hoonat says no isolates. And then we will look into it. They didn't have any female urine outpatients. That's not an error, that's just a helpful warning. We're a helpful comment on what you were seeing. So I tried to distinguish between normal behavior, warnings and errors, okay? If no more questions on that, I'd like to move on to the anti-biograms. Again, the anti-biograms is very specific. And then after the anti-biograms, we can go back to number one, data cleaning, because we could spend five minutes in that or five hours on that. If nothing else, I'm gonna move on to the anti-biograms. So can you make me the presenter again? Yes, you should be presenter now, John. Right, share screen. So I hope you can see my data analysis screen. I'm going to go to analysis type. Okay, so here under our, so I'm going back to the agenda. The question was, how can we do an anti-biogram? I wanted to distinguish between, so that's percent resistance or percent sense summary. We have two versions of this analysis. We have number one, the detailed version. Percent R, percent I, percent S, the break points, the zone diameters, the MIC distributions. So number one is the detailed report for RIS. Number two is the summary report. The summary report is usually what people call the anti-biogram. It's like once a year, you want to prepare a nice summary report. If you want to tell them all the details, then do the first option, the detailed report. If you just want to give them the high level summary, then do the second report called the summary. If I do the detailed report, I will get percent R, percent I, percent S. The summary report, it will only tell me if I look over to the right side of the screen, it will tell me the percent susceptible alone or it will tell me the percent resistant or the percent non susceptible or the percent non resistant. So that's just a question of personal preference. If you want to see the R, the I, and the S, then do number one, the detailed report. But if you don't want to see everything, do the summary report. So that's our anti-biogram. By tradition, microbiologists, not microbiologists, by tradition, physicians and pharmacy people like to see the percent susceptible because they want to drug for clinical therapy with good efficacy. They want a drug that's 90% susceptible, 95% susceptible, 98% susceptible. So the traditionally pharmacy people and infectious disease clinicians usually want to see percent susceptible. On the other hand, microbiologists and epidemiologists often like to see the percent resistant. They want to see, oh, it's 3% resistant, 10% resistant. They're basically the same thing except for the intermediate range. What are you going to do with the intermediates? Well, with percent susceptible, you ignore the intermediate so you just get the 2% susceptible. The problem is if you look at percent resistant by itself, you're ignoring the intermediates and the intermediates are usually not something, the antibiotics with intermediate activity, usually you don't want to treat a patient with that. So that's why we usually do non-susceptible. So if I'm doing emerging resistance, I want to see my percent R plus my percent I. So that's what we have in this context, percent non-susceptible. So the first option, percent susceptible. That's usually what the clinicians and the pharmacy people want. Percent resistant or percent non-susceptible, depending on personal preference and why you're doing the analysis for whom, they often prefer percent resistant, percent intermediate, non-susceptible by the epidemiologists or the microbiologists interested in emerging resistance. It's easier to talk about resistance going from one to 20% than susceptibility going from 99 down to 80%. So it's just a question of the message that you want to convey. So for traditional, normal, anti-biograms, we normally do percent susceptible. But if your interest is in emerging CRE, component resistance, for that purpose instead, people will often look at emerging non-susceptibility or emerging resistance, personal preference, up to you. More traditional is percent susceptible because a lot of times we do want to use these for treatment recommendations. So you want to drug with a high percent susceptible. Here at the bottom of the screen, I can do all of the antibiotics or just certain antibiotics. I can do, you see lower left, select antibiotics or all antibiotics. The first time I do this, I'm going to do all antibiotics. But in Ethiopia, you know, you might have 40 different antibiotics tested by somebody and you might not be interested in all 40. For example, if one or two of your laboratories tested Daptomycin and Linesalid, but the other laboratories don't, I personally do want to know the Daptomycin-Linesalid results, but I don't want to put them into my annual report for everybody because only one or two facilities selected it. So the first time I do this analysis, I will do all of the antibiotics. But then after that, I'm going to repeat it and I'm going to just put the antibiotics, the most important core, interesting antibiotics that I want to include in my annual report. So maybe I test, maybe my country tests 40 different antibiotics, but in my annual report, I want to include the 12 antibiotics that everybody is testing and everybody agrees on. So in short, we're going to start by doing all the antibiotics and then we will come back and just filter it on just a subset. If no questions. So right now I'm going to do percent susceptible summary, percent susceptible, all antibiotics. Later, we will come back and look at select antibiotics. Later, we will say, well, I want to do this for the whole country, that's what it is doing right now, or I can do this separately by lab, or I can separately do this for male female, or I can do this separately for locations or location types or specimen types. So by using the row variables, I can do a stratified anti-biogram. I will do that, but not yet. I'm doing the simple percent susceptible summary, all antibiotics, all labs and one big anti-biogram. Let me click on okay. Organisms, let me just do a few of the gram negatives. E. coli, clupcella, enterobacter, and pseudomonas. Okay, and esonidobacter. So those are some of my five important antibiotics. I'm going to click on okay. I'm going to click on data files, and I'm going to look for your data files, and there they are. Let me choose all four of them. Let me choose the first two. Let me choose the second two. Okay. Any questions so far? And you don't have to unmute yourself. You know, I'll just wait two, three seconds. It seems like no questions. So let me go back to analysis type, and I said we have the detailed report, and we have the summary report. I will show you the difference by showing you the detailed report first. When I do the detailed report, and I click on begin analysis, first. Oh, no, I should get, so these are my debase SQL light. These are debase files or SQL light files. That's the, those are the annoying things I was mentioning, but I did this earlier today and it worked fine. Um, in fact, I didn't even join this call and it worked fine at the beginning. I'm going to users, downloads, and access database. So I'm just going to install the Microsoft Access database engine to show how easy it is to do. It is annoying. Oh, and now it's got 32-bit 64-bit. I just tried to install. I need the other version of this, the 64-bit version, which I currently do not have installed. I'm looking forward to taking care of this. So Microsoft Access database engine, Microsoft Access database engine, and I go to download. And this time I'm doing the 64-bit version and it is now downloading, download now. It's already downloading. Sorry for this. It's just extremely annoying. I'm hoping that these debase, these critical debase issues resolved by the end of next week and all the debase issues resolved before the end of August. Fortunately, we learned about these issues in February and we took advantage of the time to make the SQLite option. Unfortunately, we had no other huge priorities, so we were able to dedicate the time to that. Well, then, so I am now installing the 64-bit version. The problem is that if you have 32-bit and 64-bit, it just gets confused. So sometimes that's why we have to uninstall things first. Okay, I just have to close my office first and what else do I have to close? I have to close Hoonat. That's fine. Exit, exit, exit. And close my Hoonat. Close my Hoonat. And go as I go. So I guess it's, there it is now responding and task and Hoonat and task. Good. Good. And retry. And Microsoft, okay, I just wanted me to close one more software. So we can always solve this, but as you can see, it's just inconvenient. Microsoft access, I have to close this one as well and task, close this and retry. And let me just close some of these things I don't need, close this folder, close that folder. And good. It's trying to install it, it's just a bit slow and when I think open, it's now almost finished. Okay, and okay, and exit. Okay, let me see whether or not that fixes this issue. And I go back to Ethiopia, all, whoops, I clicked the open one, I clicked the open one, all data files, analysis type, percent resistance, E. coli, Pselecythomonas, Enterbacter and SNETibacter. Okay, data files, star encrypt. Let me change this, select the first two. And I see that Rodney has just arrived, Rodney, welcome and begin analysis. Good, so it does work. We can't fix this issue, but it's just inconvenient as you can see. Great, was there a comment or a question? Maybe what kind of solution? Man, what kind of problem you experienced? What was the solution? And so just to summarize. Right, so basically I hope that within two weeks, the specific issue that we just had will be complete the result. And then by the end of August, all such issues. Like right now we analyze data files here, but we analyze data files and data entry, combine, export and crypt. So at least for data analysis, I hope to have that fixed by the end of next week. So the issue is that I ran the analysis and it said no isolates found. And the reason that it said no isolates found is that it saw the files, but it was not able to read the files appropriately. Therefore, I reinstalled the Microsoft Access Database Engine. After reinstalling the Microsoft Access Database Engine is now working perfectly. But it might work perfectly for five minutes or five weeks. So the fix that I just made is sort of a short-term fix until the problem happens again. These problems continue to happen because Microsoft continues to try to update it. And the updated version does not reliably support DBase. So that's a technical answer, but in short, I know exactly what the problem is. And I'm hoping by the end of next week, we can finally... So in the course of the last several months, we made a SQLite as an alternative option. So SQLite should avoid these issues, but we still have the DBase files. So the DBase, we'd need the Access Engine to work, but next week we will start to remove the Access Engine completely. And that should be a permanent fix for this problem. Okay? Good. Getting back to why... We're talking about the subject of anti-biograms. So let's get back on track to our anti-biograms. So anti-biograms, I have now asked for the detailed report, that number one detailed report for five different organisms, SQLite, Cepciela, I chose four files to begin the analysis and you see the DBase problem is back again. So it just happened one moment to the next. Let me, you know, I think sometimes what happens, let me just close Hoonet and go back again. So sometimes just restarting Hoonet is enough to quote unquote, fix it. It's not a long-term fix, obviously, but let me just go to, I do have to close my Hoonet or is my... Oh, I'm sorry, to the top of the screen. Hoonet close, okay. So if I simply close Hoonet and go back into Hoonet, sometimes that's enough to clean this issue. So data analysis, percent resistance, E. coli, Cepciela, enter back to pseudomonas, that's a needed factor. If I go to data files, I choose my four encrypted ones. I choose those two, I choose these two. I'm ignoring the one in the middle because that's the one from last week. So we're back on track, it is working again. So in this particular case, leaving Hoonet, restarting Hoonet was enough to quote unquote, fix the issue. But you can see how annoying this is. And I hope by the end of next week, these annoying issues will be completely fixed for D-Base for the priority area of data analysis and data entry. And then by the end of August for all the other areas of Hoonet as well. So back on track, this is the detailed report for E. coli. I click on continue. This is the detailed report for Cepciela, continue. The detailed report for enter back to continue. Pseudomonas and asinidobacter. By a detailed report, I see my breakpoints, human, animal, number of isolates, RIS, confidence intervals, six millimeters, seven millimeters. I see my histograms for zone diameters for MIC. So this is the detailed report, it has a lot of information. Continue. I am now going to change this to the summary report. And I hope that it works. Click on begin analysis. So now instead of having five separate pages, I have all the organisms on the same page. So that's the difference is that it's less detail, but there are more organisms on the same page. So instead of having one page for each bacteria, I have one row for each bacteria. And I see at the top alphabetical order, asinidobacter, enterbacter, E. coli, Cepciela, Pseudomonas. So I see the percent sensitive on the left because that's what I requested. And then on the right, I see the denominators. The denominators are very important. I'll talk about that more in a few minutes. But here you see that for Cepciela pneumonia, they tested Ciprofloxacin 257 times. But Doripenem, they only tested eight times. So that's very, Cepoxacin, I believe, the results because it's a lot of data. But Doripenem, they only tested eight times. Cepoxacin, they only tested three times. So these denominators are very important, but I'm going to come back to that in a few moments. So if I look at the graphs, this is the graph for asinidobacter, the graph for enterbacter cloakii, E. coli, which is relatively susceptible. Asinidobacter relatively resistant. So if I compare E. coli and Cepciela, E. coli is usually more sensitive, and that's what I see here. E. coli, we have a lot of antibiotics, like 90% sensitive, 80% sensitive. Cepciela, we have a fewer number. Like if let me look at E. coli, if I look at above 90% sensitive, no, let me look at above 80% sensitive. I have my amicacin, Doripenem, Gentamycin, imipenem, meropenem, nitroferantoin. So we have several drugs that have over 80% sensitive. Cepciela, I only have three. Amicacin, imipenem, and norefloxacin. And urtipenem is close. Urtipenem, well, urtipenem is right around 80%. So this is again what I'm expecting. That Cepciela is usually somewhat more resistant than E. coli. Part of it is it's natural resistant. Let me look at the, okay. Cepciela you see is 0% sensitive, which is normal. Maybe 2% sensitive, 3% sensitive is normal for Cepciela. Pseudomonas, and then I can look at a particular drug. For example, let me look like imipenem. So imipenem, very sensitive for E. coli, Cepciela, and Pseudomonas, but not very sensitive for enterobacter clioachy, which is normal. Enterobacter clioachy is at higher risk of having imipenem resistance. Part of that are specific carbapenemases. Part of them are lactases that they also have. So again, this is matching sort of my expectation. I'm very happy with how sensitive your imipenems are for E. coli and Cepciela and Pseudomonas. Okay. Any questions on this table right now? Yeah, maybe a question. Yes? Yeah. Is it possible to compare drugs or antibiotics with respect to specific microorganisms? For example, imipenem and another maybe Cephoxidine. So imipenem, we know that it is more strong drugs than Cephoxidine. Is there any possibility of getting resistance for imipenem susceptible for Cephoxidine? Yes, there are many ways to answer that question. So let me come back to that. The answer to your question is yes. And I'll come back to that after I talk a bit more about these antibiotics. So what else do I want to do with an antibiotic? Well, one of the first things I want to do is I want to copy the table over to Excel. So now opening Excel, I'm trying to open Excel. My mouse isn't working by the way, so that happens on my computer. So I'm hitting the Excel button. Let me just try Excel. Oh, here it's coming, it was just slow. Okay, blank workbook and I do paste. So here we have everything in Excel. What you definitely should not do is print this out and give this to your doctors immediately. You only give it to the doctors if you trust the results. So, okay. So the typical anti-biogram, so what I would often do right now is I'd often save this as version one. And then I'm gonna start deleting things to make version two, version three, version four until I finally have my final version. For example, I mentioned here that there were 482 E. coli. That's the total number of isolates for E. coli. But how many times did they test E. coli for each drug? I'm gonna go to the right, go to the right, go to the right, go to the right, go to the right, go to the right. And they tested, okay, let me go back to the left. There's the thing called freeze, I'm just gonna freeze it. So if I go to the right, I can still see my total. So there are 482 E. coli. As I move to the right and I look at my denominators, I can see they tested amokation 284 times. That's over half. They had azithromycin once. Probably a typing mistake, sephoxidin six times. So sephoxidin is certainly a valid drug, but it's a valid drug, but they still didn't test it very often. Septazidine, they tested 211 times. So I wanna get rid of drugs that are not tested. I can either delete them or hide them. But for example, azithromycin is 0% sensitive, but they only tested it once. So it's not meaningful. So some of these drugs are just incorrect drugs. Like azithromycin is not really appropriate for an E. coli. I'm gonna delete the azithromycin percent sensitive. I'm gonna delete drugs that are completely inappropriate. For example, genomycin high, genomycin normal, that's good for E. coli. Genomycin high is for enterococcus. That's not a valid test. And again, there's a quality issue. Did they do that on purpose? Was it a typing mistake? Penicillin invalid, I'm gonna delete those. Sephodotaxin. So I deleted those because those are invalid drugs. And there are other drugs that I might not be so interested in. For example, let me see. Like I'm getting rid of the azithromycin because it's not an appropriate drug. I'm gonna delete the genomycin high. It was only one. It's probably just a typing mistake. Look here at the imipenem. Imipenem is a perfectly appropriate drug, but they're not testing it. They tested it once. They tested it twice. They tested it once. What they are testing is meropenem. So I'm gonna go back to the left. And if I look at the imipenem statistics, you see it's either 50% or 100% sensitive. 100% sensitive sounds great, but they only tested it once or twice. So it's not very impressive. On the other hand, the meropenem has real data. So I'm deleting the imipenem because the denominator is so small. The cephalothin, it is valid. There's a foxy, they're valid tests, but they're not testing it very often. So let me delete the cepotaxime. Let me delete the cephalothin. Let me delete the doxycycline. The doripenem, it's valid, but they very rarely test it. And that's why it's so important, like urdipenem, 86% sensitive sounds great. Well, it's not good. It resisted 86% sensitive, it's okay, but it's a tiny number of isolates. Therefore I'm gonna remove this number, the antibiotic, because there's just not enough data there. I hope that makes sense, because I don't want you telling the doctor's results if that antibiotic was only tested once or twice. If that antibiotic was only tested once or twice. So I'm cleaning this list up to only put antibiotics with good numbers. Does that make sense? I hope, I'm just gonna do cut and paste. What I do at the top, I'm gonna put it at the bottom here. Of course, now I have to deal with the fact that I deleted different antibiotics on the top and the bottom. Let me get rid of the, getting rid of cephoxidin on the top. I'm just lining these things up, so let me get rid of the Doripenem on the top. Oh, and the Cipro, no, not the Cipro, that's right. In my Pena, let me get rid of the Emypennem on the bottom. And let me get rid, oh, that's fine, a different, Peptisopenicillin is completely invalid. So good, now my things line up, so here. So my teaching point for right now is like here in cephalothin, so it's nice having them right on top of each other, because it's easy for me to see what's going on. So here you see, the amication was 40% sensitive, but it was only five isolates. The minimum recommended is 30. So I'm gonna delete five from below, 40 from above, eight from below, 100 from above, 16, 15, five. Well, first of all, I'll ask Enterbacter, they're only 28 of them. So I might even delete that organism because also 30 is not a magic number, 28 is close enough. But here I only have five, so there's certain organisms that there might not be sufficient data. This example I'm gonna delete my Enterbacter cloakie because it's under the magic number 30. Of course, in reality, as long as it's above 20, it's still valuable. And also I'm gonna get rid of the Esenetobacter because there are only 15 of those. For these three organisms, I have plenty of data. But look here, they have 74 pseudomonases, but they only tested AMC eight times. So and nine times, let me delete that, delete that below, delete, delete above. They only tested, so I'm just, I'm deleting below anything which is very small. How small is up to you? I'm going to delete, you know, norefloxacin. I'm gonna delete norefloxacin completely because they very rarely test it. Dallodicic acid, it's also going to delete it. It's these are just small numbers. And here you see for the pseudomonas, there are a lot of drugs I'm gonna get rid of because they just do not test them very often. Delete below, delete above 100%, but they only tested it one time. Yeah, they tested it 10 times, 50%. So I would go through further and clean this up more, but this is reasonable. Let me get rid of the 28 and the 23 and the 44, okay. I'm gonna delete anything which is below 30, but you have to use a judgment call because sometimes it's still an interesting drug and I don't want to lose it. So you want to keep certain things and it's just a compromise between what's realistic and what's unreliable. So this is giving a better idea of the final anti-biogram. Also some of these drugs I'm really not that interested in like Sephyroxine, a lot of people are not using it. Chloramphenicol, we do not use that drug very much in the United States. So then I can clean this up further, just put in the drugs that I want, like Pipericillin by itself, is really not used very much anymore at least in the United States. We use the combination, Pipericillin-Tazobactam. So again, I'm just deleting drugs that I do not want to put into my anti-biogram. Here I see Sephoraxone and here I see Sephotaxine. The two drugs are almost equivalent. I'm going to delete these columns. I'm going to hide these columns in the middle, I didn't delete them. These two drugs are almost the same, but as you can see, the equal more people test Sephoraxone than they do Sephoraxine. So I'm going to delete the Sephoraxine. And if you look at the percent resistances, well, they're usually more similar than this, but the denominators are different. So I'm going to delete the Sephoraxine until finally you get the subset of antibiotics that you want to include in your annual report. Okay. Any questions on that? But basically the doctors are not expecting, we started off with a list of about 40 antibiotics. They don't want to see 40, they want to see the drugs of greatest epidemiological interest, the greatest clinical interest. So I have deleted antibiotics that I do not have a lot of interest in. And that should match the drugs that are recommended for testing. Once I have that list, I can then come back to Hunnet, I can go to analysis type and for future reference, I can say, I can put immediately here, I want the amiccation, I want the ampicillin, I want the Sephoraxine, I want the, which one, Sephoraxone? Where is it Sephoraxone? I don't want the imipenem because I want the meropenem. They're not testing imipenem very much. I want the gentamycin, and I want the toberamycin, and, and Sipro. As you can see, I am not asking for all of the drugs, I'm just asking for the drugs that I plan to include in my report. I click okay, I click on begin analysis and I'm gonna copy, paste that to Excel. And now you can see, this is a lot cleaner. It's only the drugs, let me just move these down below. Okay, moving that down below. I'm moving that down below. The doctors do not want to see the denominators. So I delete this before I give it to the doctors. But as the data analyst, I want to see those. So now that I have a short, so this is a lot cleaner. I don't have a, I don't have completely irrelevant drugs. I can just focus on the drugs I have greatest interest in. And then I can repeat, I see the acinetobacter is only 15, not enough, let me get rid of it. Acinetobacter, let me get rid of it. Enterobacter, 28, it's close enough to 30. And I have 26 meropenems, that's great, but I only have seven septoraxones. So start going through and they only have eight amicacins. Pseudomonas, they just have nine and two. So I get rid of these two numbers. So this feature in HUNET for doing select antibiotics is helpful for the national report, just so you don't have to look at a lot of antibiotics that you really don't have any intention to include in the annual report. Any questions on that? Good. I'm now going to say this is a macro. So this is repeating what we've done earlier for macros. I'm going to click on macro. I'm going to click on new. I'm going to say this is my gram, negative anti-biogram. Save. Save. Why did I do that? I did that because if I ever want to come back again, I just go back to my macros, I look for gram, negative anti-biogram, and it remembers which antibiotics and which organisms. Oh, they didn't remember the data files. Oh, that's a bug. But, oh, I think I know why. I think we have that issue again that it doesn't understand my data. Let me try that. Okay. Yeah, that's what I was just expecting. So I just need to read, I just need, well, okay, as I said, if I just leave Hoonette, I think if I just leave Hoonette, it's going to temporarily fix that issue. Fortunately, most people do not have this problem as often as I do. So I'm glad that most issues, but that's why in our side, it's so important for us to resolve these issues quickly. I'm just going to reopen Hoonette and hopefully this time it will work. And data analysis, not quick enough, data analysis, macros, gram, negative, load. This time it worked. I just had to leave Hoonette and go back in the second time it worked. Okay, great. So you see the macros are so valuable because sometimes I don't remember which organisms I selected. I don't remember which antibiotics I selected. Or even if I did remember, it's still a pain in the neck to have to type them and copy them in again in precisely the same order so you get exactly the same results. That's another teaching point. Everything you do that you liked that you plan to repeat in the future, just save it as a macro. It will just save you a lot of time and effort. There's one additional thing very important that I did not do is remove repeat isolates. If you have 10 E. coli from the same person, that is a sick person. But I only want to count that sick person once because otherwise that one person who is often an ICU person does the overall statistics. How do I remove the repeat isolates? I click over here on the right where it says one per patient. Hoonette has many options for defining one per patient. I can do all the isolates. That's the first approach. I can do by patient. That is what we'll do. Or I can do the first isolate per year or the first isolate per three months or the first MRSA and for the first MSSA. So the variety of approaches by room, what do I mean by a repeat isolate? Well, different people mean different things. So Hoonette, we give different definitions. The CLSI has a very specific recommendation. Take the first, do it by patient. Every patient is equal. In terms of annual statistics, count everybody once. Do not count the ICU patients several times. Do not count the treatment failure several times. Count every individual once. I can take the first isolate. I can take the first isolate with antibiotic results. I can take the most susceptible one. I can take the most resistant one. There are other useful options. I'm gonna either do first isolate only or first isolate with antibiotic results. Because of course, just because the patient has a staff or is in the lab, the lab does not always susceptibility test everything. So I'm gonna choose the second option here. First isolate with antibiotic results. Let me click okay. And let me click on begin analysis. And again, it's giving me that error. So very simple. I just need to leave Hoonette. And on my computer, I have to cancel out of it in this way because my computer is funny. And I click on end task and I do that. I go back and do Hoonette. And I think this time it will work. And I go to Ethiopia all hospitals and I go to data analysis and we go to, and I go to my macros. I didn't do that, right? I go to my macros, gram negative load. It works fine. I do one for patient. I do by patient. I do first isolate with antibiotic results. Okay, begin the analysis. Now when I do it, the numbers will be slightly different. The number of E. coli before was, I don't remember. So it was 482 isolates of E. coli. Well, those 482 isolates came from 443 people. So there were about 40 people who had E. coli twice. Of course, I'm assuming that we have some meaningful patient identifier, so that Hoonette can do its counting. So the first time I did this, I found 482 isolates. But the 482 isolates came from 443 different people. So there were about 40 people that had two or three E. coli. This in front of you is more meaningful. Well, first of all, it's gonna be very similar. On this screen, I see it was emication 95.3% sensitive. And previously it was, well, 95.8. So it was 95.8 sensitive. Now it's 95.3. So it didn't make a big difference, but it made a difference. It just was not a big difference. You tend to see big differences for something like Acinetobacter and Pseudomonas because you have a lot of repeats in the hospital setting, at least in my hospital. Let me see if I see an example of that here. I've already deleted the Acinetobacter. So here we had 74 isolates of Pseudomonas. Those 74 isolates came from 67 people. So the certain organisms that have more repeats than other organisms. Like if somebody is Niceramen and Gynaetitis, I hope they just have it once. Gonorrhea, I hope they just have it once. But Pseudomonas, they might have it every day for a couple of weeks. So removing repeat isolates is very important for some pathogens. And we always recommend doing it. For some pathogens, it doesn't make a big difference because there's so few repeats. Okay, good. And now I can make a new macro. I'll go to the new macro. And I can say, you know, I'm just gonna replace this one. New gram, negative anti-biogram. Because in an anti-biogram, you are supposed to remove the repeats. Save, save, repeat, replace, yes. Good, so I'm gonna click and begin analysis. So great. There are two more things I wanna show you with anti-biograms. I see it's almost nine o'clock, so we still have time. So two more things about anti-biograms. One is I should get a copy at the table. Instead of copying the table, I'm now gonna save the table. And let me call this gram, negative anti-biogram, Excel. I click on save. It is now saving the Excel file. I don't know where my mouse went because anywhere my mouse isn't working. So do you wanna, it's now in Excel. Do you wanna open the file now? Yes, I do. And then I go to Excel. There it is, it's just finished. So what you see here are exactly the same data. What I did before is I copied and pasted the data. By saving it as Excel, it's exactly the same data. But there are some advantages. Number one, it's prettier, it looks nicer. It's got the bold and the yellow and the lines. That's one advantage. Then it's number two, it has a heading. It actually has all those details that were not in copy-paste. It tells me which organisms, which data files, one per patient, it tells you the options. Another advantage is it gives you all the graphs. So if you want the graphs great, if you don't want the graphs, you can delete them. So there's a three big advantages. If I'm in internet looking at something I like, copy-paste through Excel is the quickest and easiest way to get the data into Excel. But instead of copy-paste, you can just do save table. It's also fast, because we're simply waiting for Hoonet to make the Excel file. The advantage of Excel, it gives you the heading, it gives you the graph, and these are true Excel graphs. So I can change the font, I can change the title. Here, I didn't show you this before, but Hoonet has this copy graph. If I click on copy graph, Hoonet copies the graph, but you can't change it. I mean, you can make it bigger, you can make it smaller, but you cannot change ABA to asinidobacter. You cannot change any of the content, but if this is Excel, you can do whatever you want with the graph. So those are some of the main advantages of saving this as Excel. Another advantage is that we can use this to make an automated report. I'm gonna come back to that advantage later. So, good. So I said I wanted to show you two more things with anti-biograms. One of them was the saving as Excel. Oh, and I just thought of something else. I'm not gonna click on isolates, and I'm gonna go to specimen type, and I'm just gonna ask for urine. Okay, okay, begin analysis. And again, it's not happy because of that debase issue. Let me exit, exit out of Hoonet, and let me go to task manager, and let me go to this, and let me just then task, and okay. So let me go back into Hoonet, and all hospitals. But I'm using all hospitals, so then I can use all the data from all the different facilities. We discussed that on the last call. Data analysis, macros, bring back my gram negative anti-biograms. This time it should work. First of all, I do first, well, I actually already did that, first I said the patient already did that. Isolates, let me do specimen type, is urine. Okay, okay, begin analysis. So now I have exactly the same data, but all of this is now urine. So I see that I do have enough data for my E. coli and my Klebsiella, 320 isolates, 117. I just don't have enough for anything else. There were only three enter factors, 13 pseudomonas. So a lot of people ask that they wanna make a urine anti-biogram or an inpatient anti-biogram or an ICU anti-biogram. So I can use these features to do that. Okay, the final thing about anti-biograms is let me go to analysis type. So everything I did here was by organism. Every row had one organism. I'm gonna change this now. Let me just get rid of, I just wanna put E. coli. Now I just have E. coli here. And urine, I want everything. So I brought all everything back. So I'm doing an anti-biogram just for E. coli all by itself. This should not be a surprise to you. It is simply the same row that we've been looking at multiple times, but it is only the E. coli. Great. Now what I can do, the new thing to show you, is I can put in the rows a new variable like laboratory. Okay, begin analysis. And here you can now see that the different laboratories, of course we wanna combine the zero one, zero zero one. So now I can see laboratory one, laboratory two, laboratory three, laboratory four. This is also valuable for the national report because then you can see the differences between the different hospitals. So for example here, a hospital laboratory one is 93% sensitive. Laboratory three was only 67% sensitive. So there's a big difference in the percent sensitive. It might be a quality control problem. It might be an outbreak. It might be a lot of resistance. Or it might be because, well, look at the denominator. They only tested amokation three times. So this laboratory tests amokation a lot. This laboratory only tests it, probably as a second line agent. So I don't believe that, I don't believe that, what is this, hospital three? The hospital three. I don't believe hospital three has 67% resistant, I'm sorry, 67% sensitive. They only tested it three times. I'm guessing that it was resistant to the first line agents but sensitive to the second line agents. So these are part of the interpretation. And then you can go back to the lab and say, can you please start testing amokation in the future? If it is something we want to recommend at the national level. So this analysis is basically useful for benchmarking. We already saw the national, how I want to see the average for each facility separately. Let me see another example of this. You have meropenem. Meropenem, 100% sensitive, 98% sensitive, 95% sensitive. But laboratory four was the smallest, 90.6% sensitive. How often did they test it? Amopenem? Meropenem, 85. So look, hospital four has 90 E. coli. They tested meropenem 85 times, which is great. They almost always test meropenem. So this suggests to me that hospital four indeed has more resistance. Maybe they have more ICU patients, maybe they're from an area of the country but it has more resistance, maybe they're older people, maybe there's an outbreak. Or maybe there's a quality control problem. Imipanem is a drug that is unstable in tropical environments. But this is allowing me to see the different problems that the different laboratories may have. I'm going to continue on to something else. Any questions about this? So this is what I would call a national benchmark. Sorry, John, there is a question. Zalalem put a question in chat. Zalalem, do you want to ask your question? Yeah. Can you identify multi-trigger resistant organisms? Yes, I can. I'm asking for that one, yeah. Good, let's come back to that. Let me finish this up. I'm going to click on continue. I do remember that EPHI, it's all laboratory one, but I'm going to put by, I think you called it institution. So here you see ABT, ALC, ALH. So the name of your institution is in a different column. It's in the institution column. So now I have all of the different facilities that EPHI gets. The most common one, let me sort this by, let me go to the top of the list, top of the list. So the number one is hospital TAS followed by AS, followed by ABT. So this is a facility anti-biogram done with the EPHI data. I'm going to do the same thing here for the gram positives. So let me say, for example, organisms, let me put Staph aureus, enterococcus ficalis, enterococcus fissium, strepococcus pneumonia, okay. Let me put all antibiotics, okay. And let me say macros, and let me say gram positive anti-biogram, save, exit, begin analysis. And then I've got the same issue that I keep on having. So I do need to leave Hoonat and Hoonat 2020 and TASC. What was this? Oops, I guess I did not close it, or just need a little more time. Okay, Hoonat, okay, great. So okay, now I go back in, all hospitals, data analysis, and the second time, it should work. So now I have my gram positives, and, oh, actually, that time the issue was different. I just chose the wrong file, encrypted. I just did not, let me go back and choose the encrypted ones. Okay, and I choose these, and I choose these, and I choose begin analysis, begin analysis. So now I have my gram positive anti-biogram, great. Oh, okay, and I chose the wrong one. I think I simply, I think I just chose the wrong micro, that's what it is, I chose the wrong macro. Step four is enterococcipicalis, enterococcephesium, tryptococcemonia, okay, begin. Great, this is my gram positive anti-biogram. What I'd like to show you now is quick analysis report. Again, this is a repeat, I'm gonna make a new report. This new report is my anti-biogram report. In there, I want to put the gram negative anti-biogram followed by the gram positive anti-biogram. Save, save, exit, let me choose my encrypted stuff. We choose those four files. I'm now showing you the final advantage of the Excel, export this to Excel. It's now during the gram negative anti-biogram followed by the gram positive anti-biogram. So the whole idea here is to start automating it. If you are doing the same thing every time, we want this to be as simple as possible. For example, you might want to make an anti-biogram. You may want to make a separate anti-biogram for each of the hospitals. So you don't want to do that manually. You just want to just have it done quickly. And where's my anti-biogram report? That's it, yeah, that's it. So here I have two sheets. This is my gram negative anti-biogram on the first sheet, my gram positive anti-biogram on the second sheet. So this helps you make your reports facility in a simple way. Okay, I'm going to skip over that quickly. It was just a repeat of what we did previously about reports, but everything in a macro can be put into a report to help automation. Now moving on to the next topic, you've asked me two questions about multi-resistance. You asked me earlier about the imipenem and the sephoxidin, and I'll do an example. So if I want to simply choose two drugs, I do a scatter plot. I can do that with measurements, but let me start by interpretations. And let me compare. I'm not going to put imipenem because they don't usually test imipenem. They do test meropenem. And I don't know another drug. I could put sephoxidin, but I don't think they tested it. Let me put Cipro. So this is now a comparison. Where's my Cipro, my Cipro, there's my Cipro. I'm going to compare the meropenem and the Cipro results. Okay, let me put E. coli and Cepciela, okay. Data files, hopefully this will work. Let me choose all four files. Begin analysis, begin analysis. And again, I've got that stupid issue. So I just, I'm going to say this is a macro, scatter plot, EPHI, save. I'm saving it as a macro, simply so that I don't actually have to give the same answers again. When that, good. I go back there, when that, and I choose Ethiopia all, and I choose data analysis. I choose my scatter plot macro, and then I choose my data files again. I move that one and I sit again. Good. Oh, and I think I chose the wrong, I think I chose the wrong, because it's therefore, it shouldn't have been, I meant to put, well, the real is I don't remember what I put. Problem is when I'm, my fingers are disconnected from my brain. So let me put, let me just put what I originally wanted to do. I'm going to put Cipro and I'm going to put meropenem. Usually I put the older drug on the X axis and the newer drug on the Y axis. But you do have whatever you want. So here I see a comparison of the two drugs. At the top of the screen, it says E. coli with the number of isolates tested. I personally cannot see the number of isolates, oh there it is, I just have to move it because right now I'm going to go to meetings in the way. So I think that's 580 or 380 isolates of E. coli. Lower left hand corner, 3% of the bacteria are resistant to both. 22% upper right are sensitive to both. The biggest category is upper left. 66, I'm going to round that off, 67%. So 67% of the bacteria of the E. coli are Ciprofloxin resistant, but meropenem susceptible. So does this answer part of your question from before? This is if I have two drugs I want to compare, I do a scatter plot. This is very valuable to the pharmacy. I'm going to choose two very similar drugs. Let me choose ampicillin with a moxicillin clavulinic acid. I get my denominator at the top of the screen, there were 219. Good. And you see most of the bacteria are resistant to both. 62% meaning neither of the drugs is great. But which drug is better? Well, if you look in the upper left, you see 17%. 17% are ampicillin resistant, but augmentin, a moxicillin clavulinic acid is also known as augmentin, the brand name. So 17% of the bacteria are ampicillin resistant, but augmentin sensitive. That makes sense. Augmentin is a better drug. Of course it's a better drug. It's a moxicillin, ampicillin, plus clavulinic acid, it's something else. I'm very happy to see nothing in the lower right-hand corner of this graph. The lowest right-hand corner of this graph would be ampicillin sensitive, augmentin resistant, and that's not possible. If the bacteria is augmentin resistant, it should also be ampicillin resistant. So this is useful for microbiology to look at quality control. I'm very happy to see that in Ethiopia, augmentin was never better than ampicillin. It shouldn't be better than ampicillin, and your data confirmed that, so that makes me feel better about your data quality. I'm gonna choose a similar example. I'm gonna compare ampicillin with meropenem. Okay, okay. Well, I'm sorry, I accidentally changed the analysis. Okay, an analysis type, and if I go to, I accidentally changed the scatter, changed. Okay, ampicillin to meropenem. Okay, and I say okay, and I say begin. So here, again, I am very happy. I do not, lower right-hand corner is empty. The lower right-hand corner would be meropenem-resistant, ampicillin sensitive. That's not possible. If the bacteria is resistant to meropenem, it should also be resistant to ampicillin. Unfortunately, a lot of people do have, quote, unquote, meropenem-resistant, ampicillin sensitive, but it's simply a laboratory mistake. The common reason is that, the meropenem disc has gone bad. So maybe there is no meropenem. I also see Zalallem has left the call, I don't know if he's busy or if he has a technical issue. So I was gonna ask him if this answers his question, but I hope that's clear for the rest of you. This analysis is useful for quality control checking, looking for combinations that don't make sense. Just for the pharmacy, is they're trying to see which drug is better. Like meropenem and imopenem are more or less the same thing, but usually one of them is a little bit better. Or like gentamycin and amicacin, amicacin is usually a lot better. Let me take a look at that. Let me compare gentamycin and amicacin. And I say, okay, I say begin analysis. So here you see in the upper right-hand corner, 69% are sensitive to both drugs, meaning that both of the drugs have activity, but the gentamycin isn't great. So 69% are sensitive to both. That's upper right. Upper left is 19%. 19% are gentamycin resistant, but they are amicacin susceptible. So amicacin is clearly better than gentamycin. In fact, it is 19% better. So this is useful to the pharmacy. Like if amicacin were 2% better, it's not worth it. I mean, they're too similar. Gentamycin is much cheaper. So let's keep the cheaper drug. But if amicacin is 20% better, you may want to switch to amicacin as a first-line agent. You do have to keep in mind, however, that it's a reserve agent. We don't want to give reserve agents to everybody because you're just wasting it for the future. So sometimes what I will recommend is if amicacin is a little bit better than gentamycin, use gentamycin. If amicacin is much, much, much better than gentamycin, then switch to amicacin. But then a lot of times it's a bit better. It's not a lot better, it's a modestly better. And then I say, well, if it's an ICU patient, use amicacin. If the patient is not risk of dying, then give the gentamycin because it's a cheaper drug. It's usually effective. But if the patient is at risk of dying, they're in the ICU or they're in the emergency room with sepsis, then amicacin is better. So of course amicacin is better than gentamycin, but is it a little bit better or a lot better? And that will answer the question, should I use gentamycin always? Should I use amicacin always? Or should I use amicacin on the sickest people where I'm concerned about the patient's life in the short-term future? If the patient is not immediately risk of dying, then give gentamycin that will probably work. And if it doesn't work, then you can switch to amicacin. So these kinds of, this graph in front of you is valuable for helping the pharmacy make these decisions about first line and second line testing. I'm now going to click on continue. And what I have shown you is a scatter plot using the interpretations. I'm now going to show you the scatter plot using the measurements. Okay, begin. So here you see the zone diameter. Lower left resistant to both. Upper right sensitive to both. Top left is gentamycin resistant amicacin sensitive. But you see a lot of them have gentamycin high level resistance, that's at the six millimeters at the far left. But then you also see gentamycin moderate resistance, you're around 10 millimeters. So some of the bacteria are very resistant to gentamycin. Some of them are moderately resistant to gentamycin. So this analysis I often find useful for the infection control team as they're trying to track down an outbreak. Yeah, which patients have which bacteria? So top left, I see clone A, middle left, I see clone B, top right, I see clone C, the susceptible bacteria. Lower left completely resistant to both. That's clone D. Lower right is amicacin resistant, gentamycin sensitive. That is rare. So you have down at the bottom right-hand corner of the screen is 1% is gentamycin resistant, I'm sorry, is gentamycin sensitive because it's 20 millimeters, but it's amicacin resistant because it's six millimeters. That is rare. It might be a laboratory mistake, a typing mistake, a zone, a measurement mistake, a disk mistake, or it might be true. There are bacteria in the world, especially in South America, that are amicacin resistant but gentamycin sensitive. So that the one in the lower right-hand corner is rare. It might be a mistake or it might be true. I'm gonna click on continue. So this is about looking at cross resistance between two drugs. Those two drugs might be similar, like ampicillin and augmentin, like imipenem and meropenem, or they might be very different like oxacillin and erythromycin. You know, let me do that with staph aureus. Let me compare suffoxitin, which is what we use to find our MRSA, and cipro. So these are two completely different drugs. I'm gonna put measurement, I'm gonna change that to interpretations. They're completely resistant drugs, but resistance is often linked. And in fact, that's exactly what I see here. Let me see what the, oh, I'm sorry, this is E. coli. So it's interesting, but I didn't mean to do that. I'm hence to do staph aureus, the MRSA. Either I'm getting that error again, or I just too, I didn't choose too good drugs. Let me, I just, I wanna do this. Let me just choose two proper drugs. So I'm gonna do, oh, okay, no, I've got that same issue. I just have to leave on that. Apologies for this by the end of next week. I'll aim to have that. If we updated this to SQL lite, because these files are debase files. And that's why you have to keep on doing this. So let's see, okay, all files. Well, in fact, I can even just, I'll even just go ahead and do that to show that. This is a different issue, but you know, since it keeps on happening, encrypted, looking for the word encrypted, good. Which is all four of those. They remove the other one and okay. And I'm gonna just simply combine this into Ethiopia, with the opia all dot SQL lite. Let me just make sure it says SQL lite here. Yes, it does. Oh, let me just type it here. Ethiopia all and save and good to get combined. So now it's combining into one big SQL lite file. This routine does not use the Microsoft access engine. That was important because you need this feature to upgrade. So now I will stop having those, that issue that we've been having all during the session. Good, I should have done that earlier. I just didn't think of it. Okay, analysis type, what was it going to show you here? Electric, I wanted to show, first I want to see for stephorius, what are the drugs that they test? I got my SQL lite, okay, begin analysis. So I remember tested. I see that they test a lot of sephoxidin and they test SIPRO. So those are two reasonable drugs. So now I'm gonna go to my scatter plot. And I'm gonna compare SIPRO on the x-axis and then I'm gonna put SIPRO on the y-axis and begin it. And let me do this, just the interpretations which are easier to discuss. So upper right is sensitive to both. Those are MSSA sensitive to SIPRO. Lower right are our MRSA resistant to SIPRO. So our MRSA are at the left side of the screen. So if I look at the left side of the screen, 10% are MRSA resistant to SIPRO. Top right are 10% sensitive to SIPRO. So SIPRO resistant is often linked. So another way to view this is let's look at the SIPRO sensitive bacteria. No, another way, no. Let's look at the SIPRO resistant bacteria. The SIPRO resistant bacteria are in the bottom row here. The SIPRO resistant bacteria on the left, you have 10% that are opposite and resistant on the lower right, they're SIPRO sensitive, meaning that if the bacteria is resistant to SIPRO, it is usually an MRSA. However, on the top row, SIPRO sensitive, if the bacteria is sensitive to SIPRO, it is usually sensitive, it's usually an MSSA. So in this facility, the SIPRO sensitive ones are mostly MSSA and the SIPRO resistant ones are mostly MRSA. So you can see that even though the antibiotics are completely different, different mechanisms, they are linked. There's a gene, these bacteria have the gene for, they have the MEKA gene for MRSA resistance, but they also have a SIPRO gene. It's a different gene, maybe on the chromosome, maybe on the plasmid. So by looking at cross resistance, it helps to tell you a lot about the cross resistance and the mechanisms of resistance. So Zalali, I hope that answers part of your first question about cross resistance. I'm now gonna click on continue. You can see percentages, 10%, 7%, 2%, gonna continue. There's this feature here called options. I'm gonna go to options. So here under options, scatterplot options, I can do the percentage of isolates, I just showed you that, or I can do the number of isolates. So now if I repeat this, I don't see the percentages, I see the numbers. 12 isolates, 11 isolates, 3 isolates. Some people prefer the percentages, especially when the numbers are big. Some people prefer to see the numbers, especially when the numbers are small. This is one way to look at what you were asking me, is you do a scatterplot. You just choose two different drugs. Let me just go back to an isolate listing now. So I'm on the first option, isolate listing, to get analysis. Here's my listing of all of my stephorias. Here's a listing of all of my stephorias, a summary over time. You see the time here, there's some bacteria way back in. You see here, there's one isolate. If I look at the table, you see the number one here. There's one isolate from November, 2010. That's a typing mistake. That's why this graph, that graph starts in 2010. I'm just explaining why the graph is so strange. The real data is off to the right. The other data to the far left, the date is incorrect and the graph is by month. So the early months of 2010, 2011 are on the far left. The real data with the dates over to the right or over to the right, let me just change that. I don't want to show you that by month. Let me show that by year. Okay, begin analysis and summary. So here you see, there are a tiny number of isolates in 2010 and 2011. Those were just typing mistakes. Most of the data are 2018 and 2019. That's something we've done previously, isolating. What I would like to show you now is I'm going to click on isolates. And I can ask for here, I'm going to go to sephoxitin. I'm going to ask for sephoxitin-resistant. Okay. And I'm going to ask for Siproploxacin, wherever it is, Sipro, sensitive or resistant. Let me choose resistant. Okay, begin analysis. So these bacteria are resistant to Sipro and sephoxitin. So this is another way to answer your question. So for these bacteria, so if I go to the right, if I move to the right, you'll see that sephoxitin, they're always resistant and the Sipro is always resistant. Not oxygen, we're sephoxitin. So the sephoxitin is resistant and the Sipro is resistant. So there's another way to look for cross-resistance. With the scatterplot, I saw everything at the same time, which is sometimes what you want. In this example, I'm only seeing things that are resistant to both and this is the summary. You asked me about multi-drug resistance. We're going to leave that for next time. This analysis called multi-drug-resistant profiles. I'm going to show you the Dovet your test hospital. This is a little teaser for next time. Resistance profiles by day, okay, or is hopefully it will accept my file, choose the SQLite version of it. So this is multi-drug resistance. So let me move a little bit to the right. My mouse is not working, so I've got to use my finger. So these bacteria at the top are resistant to nothing. These four bacteria are resistant to one drug, erythromycin. These are resistant to one drug, penicillin. Resistance to two, resistance to three, resistance to four. Resistance at the bottom, you see there are four isolates resistant to all seven drugs. And then you see they're categorized as multi-drug resistant, possible extensive drug resistance. This particularly interesting example, because two of them were oncology. Two of them were outpatient, so maybe, you know, they were oncology patients who went home and then came back. So this is just a little teaser for next time. This is a graph, resistant to everything. At the top, resistant to nothing, resistant to penicillin, resistant to two. So the answer to your question is yes, HUNET does do multi-drug resistance. This is also called resistance profiles. If you want to spend discuss this on the next call and how to choose a meaningful set of antibiotics, you're gonna exit out of HUNET. Any questions? John, this is Martin Evans. Can you hear me? Yes, I can. Okay, my question really refers to the frequency of providing antibiograms to a particular hospital. Normally, I think the practices every year, but if you're trying to detect emerging resistance, is it not better to do it, for example, quarterly? And what has been your experience globally about how often antibiograms are provided in order to detect emerging resistance? I do not recommend antibiograms as the primary way of finding emerging resistance. There are better ways. By looking at these things over time, some of those graphs of MRSA, multi-drug resistance. So if your interest is emerging drug resistance, antibiograms is one way, but you're just looking for change of percentage. If it goes from 75% to 75% sensitive, to 70% sensitive, the number went down, but these numbers bounce around a lot because of the small numbers or other random things. But if I saw, for example, okay, so basically, yeah, if antibiogram is one of your main strategies for finding emerging resistance, I think quarterly is good for that. It's sort of an early alert that something seems to be changing. But if your interest is emerging resistance, there are more sensitive and specific ways to address that question. Thank you. So in terms of antibiograms to support treatment guidelines, I suggest yearly. If you're using antibiograms as part of your strategy for emerging resistance, then quarterly would make sense for that. But I would do that, I would not do that by itself. I would do that plus other things that are more specific, such as these graphs, month to month, a quarter to quarter, okay? Other questions? I did show you how to do an antibiogram by laboratory. You saw the different rows, laboratory one, laboratory two, laboratory three. I could have done that by year or by quarter as well. So in fact, I could have asked for an antibiogram, not by laboratory. I could have asked for an antibiogram by quarter. The problem with antibiograms or quarters, the numbers get very small. Some people, what they like to do is they do a rolling 12 months. So in January, they'll do the previous four quarters. In April, they'll do the previous four quarters. July, they'll do the previous four quarters, is what we call a rolling antibiogram. So you're always using 12 months, but you just update which 12 months you're talking about.