 Okay. Yes. Hello, everybody. Thank you for joining the call. So we thought it would be good to focus on, I have shown you, we have spoken a lot using my, not my hospital, but my WHO test data set. So we're going to repeat some of those and go further using the Ethiopian data set. I always like to analyze the local or national data because it always shows, it's more interesting for you. It shows up issues that I had not thought about or issues that I don't have in my database. If my database is perfect, it's hard to show it for improving capacity when the files are fine. So I hope this will be useful for you to come up with a general strategy for understanding the different kinds of analyses that you could do, the different kind of issues and problems that arise. So before starting with who that I sent out this email in front of you last night, my time, I'm going to start with the Word document. So sometimes people ask me, John, what, how should I analyze my who net data? And it really comes down to what are the needs? What are the priorities? What is your objective? So I'm going to start off in this Word document by focusing on the section one called report audiences and reporting needs. There are different stakeholders, different collaborators with different needs. At the facility level, you have the data submitter in the laboratory often, or maybe the IT department, that gets the data ready, sends the data to the national level. This individual is the individual primarily responsible for checking that the data are complete and accurate and a good representation of the laboratory work. So the data submitter, usually microbiology or IT staff, is a key person for a feedback report to say, yes, thank you for that submission. And it looks okay. Or there are some problems. You also have other collaborators, you have pharmacy, infectious control, infectious disease, IT, hospital administrators who are also interested in the laboratory data and the resistance. So they are not going to be managing the data, but they would be very interested in the results, especially from an infectious control perspective of a pharmacy perspective. So there are the local reports that we can discuss. What would be useful for them to have? But at the network level, they're the project coordinators at the national level, what kind of reports would be useful internally at EPHI and the network coordinators at presenting and analyzing and summarizing the data. That's one thing at the network level. Also at the network level would be nice if the network organizers at EPHI prepared a report to be disseminated to the network participants. You would have to decide the content of the report. Sometimes the name of the hospital is confidential and you do not want to share that around. Or within the network, depends on the country, they're happy to share the names. So this kind of network report inside of the network, you can decide what would be appropriate for that. Sometimes what people do is the quality measures that keep confidential, because they don't want to say this is the bad hospital. But the epidemiology measures maybe are not confidential, as long as the data look reasonable. If the data don't look reasonable, then maybe just remove those. So that's facility level, I'm sorry, that's network level reports. They're national stakeholders, which often have less detail, the government authorities, ministry of health, and then professional societies, health care providers. What would you like to disseminate to your governmental partners, the general media, if there's an annual World Antibiotic Awareness Week, it's in November and you want to summarize some elements. So often these stakeholder reports are often less detailed and focus more at the highest level on advocacy and awareness and trends. And then finally the international stakeholders, project collaborators and including funders. And oftentimes what they want is often high level. They often want to project monitoring. How many samples have been received on each month, each laboratory, or some of the laboratories behind with their submissions, what are the quality measures, are the quality measures improving over time. There's the international semi-private stakeholders and then there's the international public health community. For example, reporting to Glass or putting a report onto a public website that anyone in the world can read and download. So there are different stakeholders with different needs. So there's, I'm not going to tell you there is one way to analyze the data. It sort of depends on who's going to see it, the level of detail they want, the level of detail they are permitted to see. And reporting needs and frequency. There's basically the short-term immediate reports, data submission and feedback. For example, we are now in the month of June, so I can submit my May data. If I submit my May data to EPHI immediately, either before I submit or after I submit, I want to just get a quick feedback. Thank you for the submission. That was 500 isolates, 200 E. coli. Everything looks good or things are incomplete. So I'll call that a data submission and immediate feedback report. And secondly, there's priority organism findings, you know, MRSA, VRE. So if I submit my May data, I want a detailed analysis of the May data. But then we also have long-term reports. How does the May data, compared to the rest of the year or to the previous 12 months, epidemiological trends in completeness, epidemiological trends in MRSA, CRE, treatment guidelines, advocacy awareness, this often becomes your annual reports. And then the project monitoring, are people submitting on time or the data of good quality? So I am working with the Fleming Fund Regional Grant on historical data, so it's basically just a one-time thing. We get the historical data, we analyze it and we finish it and we have a report. And it would be useful to do that with Ethiopia with your historical data. Let's spend some time analyzing and summarizing whatever historical data that you have. But for all of the Fleming Fund Country Grants and the IDDS projects, in addition to the historical view, you would also like the perspective view. Daily and weekly, immediate alerts, monthly, quarterly, yearly monitoring and trend alerts. So that's the end of what I would like to talk about right now on the Word document. The Word document then goes into greater detail about these different kinds of reports and analyses. So we will go through these, but not here in Word right now. We'll go through these in some of Excel. So, okay. I then sent you these four Excel files. I'm going to open the first one here called the Hoonet Standard Report. So right now, my focus is the content of these reports. And then after I show you this first Excel, I'm going to jump over to do a live demonstration of the software using the Ethiopian data. So what I have now opened is called the Hoonet Standard Report. I have used one month of my WHO test hospital data. And it's just the highlights of some of the most important findings. I think you will find it very helpful and very useful. But I did write this over 20 years ago. So we are now in the process of expanding this, improving the content, expanding the content, improving the formatting. So what you see is a work in progress. So the content is useful, but in the next several weeks, you will also see improvements in the content and in the formatting. So how did I make the Hoonet Standard Report? I went and I will show you after I go through the Excel. I will go to Hoonet, open my laboratory, go to analysis, and then I will choose the option called quick analysis. When you go to quick analysis, well, I may as well just do that right now. I'm going to go to Hoonet. Hoonet, I choose my Ethiopian EPHI laboratory. I go to data analysis. I go to quick analysis. And the first option here is called the standard report. So I'm going to go through the Excel output, but then I will show you here also how I did that using the Ethiopian data. You see here on the screen, there are four reports. So we have already started expanding the content. So we're going to be improving the content and the formatting. So here on the screen, you see four standard reports. And that's why I sent you four Excel files. Report number one, report two, report three, report four. Okay, great. So now I'm going to show you what that report looks like. So this is the old Hoonet Standard Report that I wrote 20 years ago. And I'll show you the things I like about it and the things I don't like about it. Okay, first of all, you see at the bottom there is sheet A, B, C, D, E. So we have different sheets for different kinds of feedback, different kinds of analysis. It goes up to I, I believe, F, G, H, I, I think that's the last one. Yes. I'm going to go back to the left. So the idea here, so basically if somebody sends you a data file, how are you going to analyze that file? Well, you could do this and you could do that and then you could do that and then you could do that. But you're spending a lot of time manually reviewing hospital one data. And then you're doing the same thing for hospital two data and then hospital three data. So the idea of the standard reports is to do all those analyses that you would like to do, but automatically. Somebody sends you a file. We are also working on the ability to do a web submission. Basically, if you submit your data on a secure website, the website should be able to prepare the standard report automatically and then send it back to the individual. So, so we're trying to make everything automatic to make the concept of feedback easier and faster. That way you can focus on the interesting things and the actions, not simply on the data preparation and data analysis. So what is in the standard report? Here, section A is a summary. There was one file from one laboratory with 622 isolates. Today's date, well, it was in fact yesterday. In fact, it was my birthday, June 10th was my birthday. So I'm 21 plus again, or 50 plus again. And all 622 isolates came from the year 1995. So just basically a little bit of a summary to make sure it matches your expectations. This is the WHO test database. So you will be able to, this file I created that we are looking at, you can create this yourself using the same data because you have access to the same one month of sample data. So section A is simply a very high-level overview. It's the kind of thing that the, you know, the stakeholders may want to see. They don't want to see all details, but this is one of the simple details they may want to see. That is section A. Section B, data fields. So as you can see, one thing I don't like is the formatting. I did this formatting 20 years ago to be printed on paper. The Excel option was added later. So that's one thing we will be fixing quickly, is making sure the formatting looks better so that the columns line up. The column should line up in a couple of weeks. So by the end of June, this report will look better. So section B is called data fields. Again, there were 622 isolates. You already knew that from sheet A. You see here two columns. Well, you see the name of the data field lab, identification number, date of birth, percent complete, percent invalid. So in my simple database, it is zero percent. Oh, we have a small mistake here, date of birth. Well, it's missing. Okay, but here you can see that specimen data 100%, organism 100%, laboratory 100%. But something like age and gender, you can see here at the bottom, the following fields have no data. No age, no sex, no first end, no last end. Part of that is because it made a confidential data set. And then you see these antibiotics have no data as well. Then here at the bottom, you see there is no sex gender, so male and female is missing. Then I can see that 8% of the database is cardiology. That's card. 7% is C-surge. That stands for cardiac surgery. 14% is emergency room. 8% is ICU one. So you can see this can be very valuable to infection control people, so they can sort of see does it match their expectations? So that's a data quality check. How complete is it? And where are most of the data? If I go down further, I see not just the specific location like surgical intensive care unit, I see the general location type. 14% are emergency room, 13% are ICU, 53% are other inpatients, 1% is other location, 20% are outpatients. And then I can see this specimen data and I see a bit of lactamase and ESPL are empty. So the idea here is basically to give you an immediate feedback about which columns are empty, which columns are complete, how complete are they, and are there some invalid data that somebody should try to fix? That's section B about the data fields. Section C is organisms. And again, the formatting we will improve. So you see here at the top, the top 10 most common organisms. I'm just going to change the font here. When I change the font, you'll see something magical, well not magical, but courier. Now that I changed the font, the columns line up correctly. So that's one thing we'll do is we'll change the font so that the columns line up. So these are the 10 most common organisms who regulates negative staff, 105 isolates, 105 in January, 86 E. Coli, 86 in January. The reason for that is that I've only analyzed the January data. These are some other important organisms that are not common, but they are important. Streptococcus pneumonia, and I see a meningitis. So here at the top, you see the top 10 results. And then here at the bottom, you see other interesting, like Vibrio cholera, I hope is not at the top of the list as one of your common organisms. But if you have it, you still want to see it. So the idea of the summary report is I'm not showing you every single detail, because you might have 80 different species, you know, Proteus and Serati and Providencia and Bacteroides. And yes, if you want all those details, you can do that. But in the summary, the idea is to just try to focus it on the highest level interest, the section C. Section D, antibiotic results are about the most important things. What is the percent MRSA, VRE, CRE? So we'll come back to this using the Ethiopian data. We're in a section F called flag isolates. These are high priority at the top and medium priority results at the bottom. For example, these, this isolate, it says CFR that is such a vector Freudii, and it is non susceptible to a carbon pentam, such as imipentam. So we're trying to highlight for you in this data file some of the most interesting, most important results. High priority would be something like CRE. Medium priority would like ESBL or MRSA. They're also important, but they're also relatively common. That's section E. Section F is about invalid data, but there are known, well, I'm sorry, it's about low frequency results and there were not any laboratory configuration. It's looking about, is there any problem in your configuration? We'll come back to that again when we look at the Ethiopian data. Section H, comments about the data file, no significant comments here. And finally, invalid data, no invalid records were found. So I'm going to close the Excel. Do you want to save the changes now? So in short, it was a lot of automatic information quickly and easily. I'm going to show you now, I'm going to start for the beginning again. I'm going to go to Hunnet and you see the laboratory here, Ethiopia Laboratory Code 01, name EPHI, National Microbiology Laboratory, Open Laboratory. I will go to data analysis. We've already looked a bit at how to do normal interactive data analysis and we will do that, especially if any questions come up. So you can do a normal interactive data analysis, but now I would like to show you this other feature called quick analysis. And I'm now on quick analysis. I've been talking a lot. Before I continue, are there any questions? It doesn't seem like it, so I will continue. No, there's a couple of comments, but they're nothing important. They're not questions. So, yep, please go ahead. Great. Thank you. Because yes, I do not see the chat messages. I don't want to interrupt. Yeah. Can we track them in case someone has a question for you or wants to ask something, but we're good. That's perfect. Thank you. So here on the screen, on your version of Hunnet, you will see one standard report called the Hunnet Standard Report. On my version of Hunnet, I see these three additional ones and I sent you these three additional ones last night by email. So I will show you how you can also utilize those. So, but on your version of Hunnet, you would have the Hunnet Standard Report. You might even have this European one we decided to suppress. There's one called Earsnet. It's really only meant for the European Earsnet project. So we started removing it from other peoples. So this is the predefined standard report, which is the one I just showed you in Excel. These other reports, Adam and I created last week, yesterday, we have updated them. So there is a difference in Hunnet between predefined reports, the first one, and the user defined reports. The things I like, what do I like in the predefined report? The predefined report is available to everybody automatically. It has a lot of valuable content. It's concise. It is fast. The things I don't like, the formatting isn't great. You cannot configure it. You cannot add things. You cannot remove things. It is what it is for all of its strengths and weaknesses. You cannot change it. These other ones are called user defined reports, and those have the advantages that they look better because it's nicely formatted for Excel. They are customizable. So you can change them. You can add them. You can revise them. You can combine them. The only disadvantage is that you have to do it yourself. So the advantage of the standard report is that it's easy, automatic for everybody. The advantage of these other ones is you can do whatever you want that Hunnet allows you to do, but you have to do the work yourself. So we're now in the process of merging these two approaches together to give you a configurable, a predefined standard report that you are able with better formatting, but also that allows you to format it. So over the next several weeks, you will see gradual improvements in this area called quick analysis. And also we could make or we could start customizing it for Ethiopia. EPHI network report, laboratory feedback report. So I want to raise the ideas on this phone call, and then you can think about how you would like to customize them. So here I have clicked on the Hunnet standard report. Here on the right, it says it does say edit, but you can't edit it a lot. I already showed you that the Excel has section A, section B, section C, section D, section E, F, G, H, I. So you can turn the sections on and off according to the sections that you want, but you can't change them. You can simply just show them or hide them. So this allows a little bit of editing, but more or less it's basically I can show or hide some of the sheets that I want. There are two sheets here. Yes, there are three sheets here. I'm going to highlight them. I have left here, check, microbiology alerts, low frequency alerts, and invalid data alerts. Those three sheets, I'm going to show you. I already did show you. So I'm going to go to the file open. I'll try to open it from my email. Let me go back to my email. Hunnet standard report Excel, and I just opened the wrong document. Hunnet standard report Excel, waiting for it to open. So I showed you this section E and section F. Section E here you see it has the patient identification number, that 45735, as well as the specimen number. So if you have a carbopenem resistance E. coli, I can immediately know which patient it is and which sample it is, which is very valuable. However, it's also a bit confidential, the patient identification number. So the reason I highlight these three is these potentially maybe have a confidential patient identification number. So if I want to share this with somebody outside of the network, I might want to do a standard report excluding those three, because those three are going to have a patient ID. If you are okay for people to see the patient ID, fine, share it with them. For example, if you want to give this back to the hospital, the hospital, sure, of course you can give them the patient IDs. So these three that you see here, in fact, in the future, we're going to rename this a little bit, just to remind people that most of these outputs that you saw are simply aggregate, non-confidential statistics without patient data. But these three, the microbiology alerts, the low frequency alerts, and the invalid data alerts, those would have the patient identification number. So in fact, you might have two versions of the report, one version without the patient details and a different version with the patient details. I'm going to leave this, everything selected. So step one, I choose who next standard report. Step two is I choose my data files, and here is that one here of data that I received, 773 kilobytes, and I click on okay. And then I have the choice of exporting to the screen or to Excel or to access or to Microsoft Word, that's a new feature. So right now in Word, it has the same formatting that we use for the Excel. So it still doesn't look nice, but the idea is a very powerful idea. For example, for people interested in all of the details, I would like to give them the Excel output. But for an international stakeholder, a lot of times the Word or PDF summary would be sufficient for their need. So right now it's moving in a good direction, and you will see improvements in these areas. For the first example, I'm just going to simply export this to the screen. So I chose, I just did three things. When that standard report, I selected my one year of data, and I said please output the data to the screen. I now click on begin analysis. I could time this to see how long it takes, but okay, let me click on begin analysis. It is now doing about 20, it's doing a lot of analysis. Okay, but you see how fast it was. It did all those analyses and it finished. So here you can see in the summary, the data file, there was only one data file. So what I'm showing you can be run on one file or on three files or on 20 file. It can be data from one year or five years. It can be from multiple facilities. So I'm simply seeing here that I ran this analysis on a single data file. The number of laboratories is two. That's a little bit unexpected. I thought there was only one laboratory, hospital zero one. So that's a little bit unexpected, but I will show you, I do know why it says two. The number of isolates in one year was approximately 1,800. The very first date was January 1st, 2019. The last was December 30th, 2019. I am happy about that because a lot of time they're typing mistakes. You know, sometimes people will say, you know, sometimes people will, they don't put the year 2020, they put the year 2002. So the year 2002 is a valid date, but it's not the correct date. So in the data file that I received, all of the dates are in the year 2019. Sometimes I see data from the year 2027. Again, because somebody mistyped. They're sort of going into the future, they put a future date. So I'm very happy that the data file I received has correct dates, January 2019 until December 2019. All of the isolates came from the year 2019. So the data file that I received does not have any big mistakes in the dates. Yeah, there might be small mistakes in the dates, but you know, at least the year is correct. That is section A. Section B, data fields. So I can see that percent invalid, nothing. There were no invalid results. For example, the sex should be male or female. If you put the letter G or P or T, those are invalid genders. So I can see the laboratory code is 100% complete. The sex, I'm happy to say 99.9% complete, age 98% complete. The name of the department is 29% complete. The location type, inpatient, outpatient is 23% complete. And that's something we would like to improve to have the best quality data. The specimen number, the specimen date, specimen type organism 100% complete. ESBL is 1% complete. And that's okay, because you don't do ESBL on everything. You don't do ESBL on staff. You don't do ESBL on no growth. You only do ESBL on E. coli and Klebsiella. And it is not a required test. On a previous call, I mentioned this idea of scores. Like, for example, if the organism column is missing, I want it to be there. Specimen date, I want it to be there. Age and gender, that's kind of up to you. It depends on how realistic it is. At least for EPHI, it is very realistic. So you can decide which of these fields, I'm going to distinguish between primary fields, secondary fields, and other fields. For example, the primary fields would be like organism, specimen date, specimen type. Secondary fields, maybe say sex and age, they are interesting. But they may or may not be required. And that's something that you can decide in Ethiopia. What do we want to call required? And then ESBL is other. It's interesting, but, you know, it's, I'm not going to, I don't, you know, that's up to them. And the idea of what is required, what is recommended can also evolve over time. You know, for example, something like age and gender might be very difficult at the beginning. So maybe this year it will not be required. But after two or three years, maybe it will be required in the future. Or similarly, date of admission. Date of admission is very difficult for a lot of labs in the short term, but it's something that maybe in the long term, maybe could be added later. The following fields have no data. There's no first name, no last name. There's no location. So at EPHI, they are entering the department 29% of the time. They are entering the location type inpatient, outpatient, 23% of the time, but they're not entering a precise location. You know, which lab, which district, et cetera. There's no institution. Beta-lactamase is empty. Carbopenamase is empty. Inducible clundamycin is empty. MRC screening is empty. And then you can see, is this what I was expecting? Because you might say, why is the location missing? You talk to the data entry people and you can see if they have the information, if they're putting it into the correct place. Then I'm going to jump down to the sex, and then I'm going to go back up to laboratory. Let me move this up a little bit. So the top section called part one are the summary statistics. Over all percent complete, over all percent invalid. Part two are the detailed statistics. Female 45%, male 55%. No value was one. So one time out of 1800, the gender was missing, either because they did not know the gender or they forgot to enter it, or maybe it's a quality control strain. If it's a quality control strain, male or female doesn't make sense. So just because it's missing doesn't mean it's invalid. So 6% are emergency, 6% are ICU, 1% is medicine, 11% is neonatology, 71% is empty, that it was simply was not entered. So if we're looking at data completeness, this is something if possible, that we would like to improve. Similarly, the location type, you see it's a shorter list, the department is a bit more specific, emergency medicine surgery, location type is at a higher level. In patient 5%, ICU 14%, missing 77%. The laboratory, you see we have two labs, one of the labs is laboratory 01, and the other laboratory is 001. So I'm kind of guessing they maybe changed the code in the middle of the year. So we would like to fix this, we would like to claim this, because right now who not thinks this is two different laboratories. One laboratory is laboratory 01, the other, which is most of the data 73%, but 27% is a different lab called laboratory 001. So this is something we'd like to clean up so they all get the same code. I'm kind of guessing that they changed the code in the middle of the year. Here at the bottom, you know, I can see that 8% of the data were from January, 11% from February. I see there's a lot of data February, March, April, 11%, 13%, 10%, 11%. But then the last four months of the year, it's dropping down 6%, 3%, 4%, 5%. So either the laboratory was less busy and they were doing fewer samples, or maybe they just didn't do the data entry, maybe the results are there, but they're in a notebook. So this allows me to see, and in the future version, you will see this as a graph, because it's always easier to look at a graph. Sometimes there are too many graphs, so that sometimes you want graphs, sometimes you don't want graphs. So even by looking at this, I can see overall, every month, so let's see, the year has 365 days, so one out of 12 months is about 8%. So 8.3%. So approximately every month is about 8% of the year, but you see some months are more than 8% and some months are less than 8%. So we're doing less testing or less data entry for the last four months of the year. ESBL 1% was positive, 0.1% was negative, 99% was no, there was no value and no proper pandemic results. Before I continue, any questions or comments on these data? These are your data. Okay, I don't hear anything, so I will continue. That is section B. So some of my comments here is, and also EPHI is different from a hospital. A hospital often would want to have the patient's name. Of course, also as a reminder, EPHI sent me a confident and encrypted version. So there's a very good chance that their version of this file has the patient name. My version does not because that is part of the encryption process. So some things are missing simply because of the encryption process. I would usually recommend the location. So HUNA does have these fields called department and location type, but that is really meant for international standardization, whereas location is usually where we put the local location code. And the other comment I mentioned is, I can actually show you later how to clean up the data. So most of the data, 1,321, it is called laboratory 01. 477 times, it's called hospital 001 or laboratory 001. It would be good if we could merge those together into simply 01 or 001. So no comments on section B. I'm going to go to section C called organisms. As we saw at the end of the Excel, and here it's formatted a little bit more nicely because of the font, you see that the most common organism is XXX, which is the code for no growth. So out of the, I don't remember, out of the 1,800 isolates, a bit under half are no growth. Followed by XXX is no growth, XPA is no pathogens found. So that's basically also no growth. Followed by E. coli, followed by Tuftsiella, followed by Staph aureus, assinated vector species, XSA, I forget what that stands for, no significant, or no strapped group A, I think is what that is. Staph coagulates negative, PCE is Birkholderia Sipatia, makes it just makes species. In a future version, in addition to the three letter code, we will also write the full name. I know the codes, but I don't expect everybody else to know the codes, especially the stakeholders. So you can also see January for every March, let's look at E. coli, a nice organism. We had six E. coli in January, 11, 15, nine, eight, six, 18. So the number, so you know the thing I really notice here, I'm going to go to the Staph aureus. What has really changed is I think that they may be stopped entering the no growth. So the number of E. coli really does not change at the end of the year. The number of Staph aureus really does not change a lot at the end of the year. It changes a little bit, but not a lot. Assinated vector doesn't change. So a lot of these organisms, the big drop is in these first two. So my guess is that EPHI was entering all of the results for January to August. And then in the middle of September, they made the decision to stop entering the negatives. I do not know if that is true, but you are on the phone. I'm going to stop in a few moments to ask for a comment. You're at the bottom are other interesting organisms, like Pseudomonas aeruginosa is not one of the top 10, but it is an important organism, especially if there's an outbreak. The Malthus influenza also an interesting organism. So I'm going to stop for a moment and ask a question. Do you have an explanation why these numbers dropped at the end of the year? My guess is that you stopped entering the negatives. Hello, John. First half year's day. Second, I'm not sure that October, November and December, the old data is not included or entered. Maybe Zalalem will clarify more, but the number of the sample is low because of, I think, it's not entered the whole data for the last year, because as you see, the number of the sample size is below drop. I'm sure that October, November, December is the same as the previous numbers. So maybe Zalalem will explain more. You see that the pathogens, the numbers are relatively stable. So the number of E. coli is about the same, Pseudomonas is about the same. The only two that dropped are the XXX and the XPA. So Zalalem, did Zalalem during the call? No, John. Currently, they don't register the negatives. I think that's the trend. Maybe that will be the case. Zalalem is not really around. Okay, sure. That's fine. Basically, if you asked about Geneva, they would tell you they would love to have the negatives. It's really your decision. How much time and effort is it? What is the value? And the answer is there is value, but is it worth all of that entry? So as you can see, that your team has entered results for 1,800 samples. But from those 1,800, over 1,000 were no growth or no significant pathogens. So it's your decision. And I really very much want all labs to enter the positives. The negatives is a recommendation, but it is a lot more work. Why do we want the negatives? Because sometimes you just simply want to answer the question, how many blood cultures did we do? So if you enter the negatives, then you can calculate that. But there is another way. And this is what, this is actually what they do in Europe. So in Europe, the Earsnet project asks them to report the positive bloods for some priority pathogens. But then they have a separate question. How many blood cultures did laboratory 1 do in the year? How many blood cultures did the others do? So they just do an estimate. So they don't have manual, they don't have detailed statistics on the negatives. But they just ask each lab, how many, how many blood cultures did you do each month? If possible, not how many blood cultures, but how many patients had blood cultures. You know, because a lot of times with blood cultures, you'll take two at the same time. One from the left arm and one from the right arm. So if you have a hundred blood cultures, that might be 80 different people. So if your group has an interest in the negatives, if your group is no interest in the negatives, forget about it. If your group has some interest in the negatives, you can either manually enter all the negatives, which is what happened here for the first part of the year. Or you can ask the negatives in a different way. You can just say to the lab, how many blood cultures did you do last month? And then they just tell you the number. That way they can tell you the number because they usually know from purchasing inventory. They just go to the notebook and they just count quickly. One, two, three, four, five. And they say, okay, last month we did 50 blood cultures. So this is a way to get the negatives without entering them manually into who net. Okay. So, so my job here with the looking at the data is to understand what the laboratory is doing. And then I will have some high priority recommendations and low priority recommendations. And then just low level comments, observations that people may or may not want to act on. So that is section C about the organisms. I will note that. That's John. Yes. John, I have a question. Can you hear me? So what is the relevance of the negative, be it manually or inventory or whatever the method we used at local and at national level? Which level is very important to capture the negative data? Right. Well, it's right. It's, it's, well, okay. So in short, there is a value. Is the value work the amount of work that it would take to do it? And that's a decision that's a discussion to have, because basically you want to convince the labs that their time is being well spent. If you cannot give them a good reason, then they might not do it. Or they might do it, but they're going to complain. So they need a good reason. And then it's up to them to decide whether they agree with your reasoning. So what is the value of the negatives? Well, a lot of the value of the negatives is not, it's actually how many were tested. For example, if laboratory, if laboratory one tests 50 blood cultures every month, but laboratory two, let's assume that hospital one and hospital two have the same size. For example, hospital one has a hundred beds, hospital two has a hundred beds. Maybe hospital one does 50 blood cultures a month, but hospital two only does five blood cultures a month. And you want to tell the hospital two, why in the world are you doing so few blood cultures? Or as hospital one, and they'll say often we don't have the resources. It's not part of our training. They don't believe the results. The physicians don't know how to take a blood culture. And the laboratory doing a lot of blood cultures. Blood cultures are very important cultures. And laboratories that do not do a lot of blood cultures often are missing important diseases like sepsis, meningitis, where a blood culture is an important element in the diagnosis and in knowing the treatment options. So this, okay, I guess the general topic that I'm referring to is called diagnostic stewardship. Are people doing microbiology samples? Are they collecting microbiology samples? Are they collecting the right samples? Are they collecting a few? Are they collecting many? Unfortunately, a lot of places, they collect a lot of sputum and wounds and neurons because it's easy to do. But they're not doing blood. They're not doing cerebral spinal fluid. They're not doing abscess punctures like deep abscesses because they're difficult and more challenging. So for diagnostic stewardship, I want to see what are the samples are they taking? That's one question. So hospital one maybe does 50 blood cultures a month and hospital two takes five blood cultures a month. At this point, I don't really care if they're positive or negative. I want to know are they doing the tests? If they've done the test, yes, I want to look at if they're positive or negative. But one of the reasons I want the negatives is not because I'm interested in the negatives, but I am interested in how many tests they did. For example, if they have no positive blood cultures, is that because they tested, is that because they took no blood samples? If they took no blood samples, then I know why they have no positives. If they took 50 samples and they have no positives, then something strange is going on. In my laboratory and a lot of lab, in a lot of medium and high resource labs, about 10% of blood cultures are positive. About 10% of those are real positive. I'm sorry, about half of the, about 10% are positive. About half are real positive and about half are contaminants, like coagulants, negative staff or diphtheroids, lactobacillus. So in a medium and high resource place, often about 10% are positives. In a low resource place, the percentage is often higher for two reasons. One is bias. They only take samples from the sickest people. So you have a high positive predict. Yeah, there's high specificity. You're not taking a lot of people, like if we have somebody with a fever, we'll take a blood culture sample. Whereas in a lot of low resource places, if a person has a fever and sepsis and low blood pressure and very sick, then they will take a blood culture. So one reason a lot of low resource countries have higher percent positives is because they're only taking the sickest people. Also it is possible the rate of sepsis is higher. And the other reason is sometimes the rate of contamination is higher. So what is the rate of contamination? What is the rate of no growth? I kind of need the negatives and the no growths to calculate that. So there is value for diagnostic stewardship to know how often did they do the test. So laboratory one tests 50 blood cultures a month. Laboratory two takes five blood cultures a month. I want to have a discussion with laboratory number two. Why are you not taking more blood cultures? So we're not talking here about resistance and disease epidemiology. We are talking about the utilization of laboratory services. What we would like to see is maybe in January, if we talk about laboratory two, maybe they took five blood cultures in January and then maybe 10 in February and then maybe 20 in March. So we can see that after we give them feedback, they are starting to realize the importance of blood cultures. They're starting to make more purchases. They're putting the blood cultures on the floor. They're doing training with a phlebotomist on the proper way to take a blood culture. So in short, there's not a lot of intrinsic resistance in the negatives except to get the total denominators. One way to get the total denominators is to have them enter the negatives into who net as you did here. 1000 of these results were negative. That's one way to get the negatives. The other way is to don't enter the negatives into the computer one at a time. It's just to go to ask the lab and ask them at the end of the month how many bloods did you do? How many urine did you do? How many wounds did you do? You're just asking them the total statistics because then you can answer the question. How many blood cultures in January? So you don't have the age and the gender and the location, but you do have that total number. How many blood cultures do you have? So I do give different recommendations to different countries. I mean, I don't give different recommendations, but it's not my job to decide what your priorities are. So it's something with the IDDS colleagues, EPHI with the hospitals. You want to look at what is the value? What are the expectations? What are the resources? What are the time commitments? One thing that would be valuable for you to do in terms of informing this discussion is to time somebody in the lab who knows when a data entry, how much time does it then to enter one negative result? And how much time does it take to enter one positive result? Of course, it takes more time to enter a positive than a negative because you have the antibiotic results, but let's say it takes one minute to enter a negative, the patient ID and the sample and the date. So a lot of that, all of that's the same whether it's positive or negative. So let's say it takes maybe one minute to enter a negative and maybe one and a half minutes to enter a positive because you still have to put the antibiotic results. This allows you to start estimating how much time they need to enter data. For example, to enter 100 negatives, does that take two hours? I'm exaggerating. To enter 100 negatives, does that take a half a day? Does that take two days? Does that take three days? Does it take the week? So you can start estimating the time commitments. So here you can see that they stopped entering the negatives. They stopped entering most of the negatives, so you can estimate if they wanted to go back to October, November, December and enter the negatives, you can immediately estimate how much time it's going to take. If we look at these numbers, I'm going to round it off, but they have approximately 90 a month. So they have approximately 90 negatives a month. So October, we're missing about 90. So basically if it takes one minute a negative, that's an hour and a half for October, an hour and a half for November, an hour and a half for December. So that's basically five hours. But don't trust me about the one minute. You'd have to do your own estimates and your own calculations. Does that help you? Because I'm just giving you general guidance, and it's really a question for the network coordinators, the funders and the participants to come up with an agreement as to what is valuable and useful and the best way to collect those data. You're right. The problem is they always complain about time, lack of human resource and all these kind of things. Still we have this kind of problem and discussion among labs and with the National Coordination Team. That's our problem. So my question is really how can we convince them to really enter the negative data for their purpose, for their facility purpose? Otherwise, we cannot convince them. They assume that the data will be used at national level. And instead of telling them this, I just want convincing idea, convincing points to convince them to enter also negative data. Still, if we ask them every month manually, still they'll complain about that. No, I understand completely. Of course, that is also one of the advantages to have a lab information system like Polytech or one of these others because they're entering all the data for other reasons. They're entering the data for clinical reporting. So people, this is also another thing that a lot of countries do. For a lot of countries, for people with an LIS, they request the negatives because they can simply download the negatives. I want to say the negatives. I mean, they really download everything. They're not going to download the negatives separate from the positives. So for a lot of countries, the laboratory is using an LIS, download everything. On the other hand, people doing manual data entry, they ask the labs, please just enter the positives. So that is an approach. So one of the things we're interested in, one thing which is a core requirement, is that whatever we do be sustainable. Enter the future for the next two, five, 10 years. Regarding the negatives is entering the negative sustainable. So one thing you might want to do is enter the negatives two months a year. So if there's value in the negatives, maybe not do it 12 months a year, but maybe just enter it two months a year. And also, as I said, I kind of like the idea if people are complaining a lot, I don't want laboratories to drop out. Because some laboratories drop out because they just don't have the time. So for these labs, you might want to avoid manually entering the negatives into Hoonat. And for those labs, it's just asked them the total number. Just ask them, how many blood cultures did you do? And they don't even have to tell you positive versus negative. I mean, they could if they wanted. But I'm mostly interested, how many blood cultures did you do last month? If they go over to the notebook, and they just start counting, they say one, two, three, four, five, they could tell you the number of blood cultures in a month in probably just a couple of minutes. So if labs are complaining significantly about the data entry for the negatives, I would recommend you consider this recommendation that people with an LIS request the negatives. People without the LIS request that they enter the positives, and then just request that they give you a rough count of the negatives. And they could just do the negatives for the priority specimen types. Your blood, urine, pus, whatever you want to define, sputum, or respiratory. Okay. So you're asking for a reason to convince them. Even if I were there, I'm not sure that I could convince them. I give you a reason, but also some backup plans. Thank you. John, there's a question from Rogers. He put it in the comments. He's asking, how can quality indicators like recovery, possibility rates can be monitored without recording negative samples on the same issue? Right. So basically what you can, so if that is valuable, what you could do is just get the denominator. You can ask them, you can ask them, how many blood cultures did you do? You have the positives in Hoonat. And you might see that in the month of January, they had 10 positives, because they manually entered the 10 positives into Hoonat. And then you just have the laboratory ask the simple question, how many bloods did you do? They can tell you, oh, you know, we did 50 bloods. So they don't have to enter the negatives into Hoonat. They just tell you the denominator at the end of the month. Thank you. Okay. Another way is to pay them. If you pay them to enter the negatives, then they will enter the negatives as a requirement for receiving funds. That's a reason. But usually we're not funding the labs. We want this to be sustainable. You know, you might want to give them money early, like with the EARSnet project, not for data entry, but the EARSnet project in the first two years of EARSnet, the European project, is they did give countries a halftime salary for a national data manager. They only did that for two years, but it was just kind of to get things started. Because obviously it's more work at the beginning than it is later. So the European Union did have money to pay for a halftime data manager, one data manager for each country. So that was an incentive to get the project going. So your positivity rates, contamination rates, blood culturing rates, you do need the negatives for that, but you don't need the negative details one at a time. You just need the simple total count denominator, which is much easier to calculate just by going through the notebooks quickly than manually entering them into EARSnet. In Uganda, we decided early on to enter all the negative tests into EARSnet from the beginning. Of course that helps us to get the denominator. So it is hard work. It is hard work. But we decided early on to enter negative samples. But it is hard work. It is quite hard work. And do you have some idea? Like here we have about 90 negatives. Well, if I add up, I'm going to go back to sheet number one, or B, but go to the bottom. So January was 140. February was 200. I'm looking at the numerator here, 200, 240. I'm going to round that off. So they do about 200. They were doing about 200 samples a month. So Rodney in Uganda, is it about 200 a month? Or is it less than that or more than that? So it differs from from site to site. In some sites have the testing capacity. Some sites have up to 200 a month. Others could have about 50 per month. So we have about eight labs using EARSnet. So some of them are as low as 50 a month. Others with better capacity are about 200 samples a month. Quite a few of them. But most of them are in the range of 50 to 100 a month. Right. Right. So this idea about entering the negatives, to a large degree comes down to how many are there. For a small lab with a small data volume, this is maybe an extra half an hour or an extra hour a week. But for a large data with a large data volume, it is maybe three to five hours a week. And you know, that's hard when they're so busy with everything else. Fortunately, in a lot of the world, the larger labs have a lab information system, so they just download everything. The smaller labs often have a smaller data volume. So before making final decisions, it would be good to get a sense of how much time does it take to enter a negative and get an estimate of how many negatives there are per month for the different facilities. And so are we asking them to do three hours of work a week for the negatives or half an hour of work or a day of work? I'm just exaggerating because I really don't have a good sense. So this kind of analysis in front of you was very quick and easy to do, but it does help you to try to get an estimate of how much work it's going to take. And as I said earlier, the long-term strategy for everybody should be to get a lab information system. Once they have a lab information system, usually you log in the sample when the sample arrives, meaning that it's just part of normal laboratory processing. A sample arrives, you don't know if it's positive or negative, you put the blood into the system. And then when it's done, you just put positive or negative and then you spit out the clinical report. So you are doing the data entry for purposes of clinical reporting, not specifically for epidemiology. So the long-term goal for everybody should be to have a working LIS. Once they have a working LIS, hopefully we can get the positives and the negatives at the same time. So we just get down to everything at once. People doing manual data entry. So people with an LIS, I'm not that worried about the workload because everything is already in the system. People doing manual data entry, it's a question of how much work is it. So how many negatives, how many bloods, how many urine, or maybe what you could do is enter all the negatives for the bloods, but don't enter the negatives for the urine. The urine is often a very high, how about this database? In this database, I'm going to go to the specimen type. Where's my specimen type? Oh, I guess I didn't tell you. Oh, okay. I guess it's something we didn't put in. So the specimen type, maybe because there are so many. So I can't actually show you the specimens on this screen. So what some people do is they do a lot of details for the bloods, but they don't do a lot of details on the urine simply because the volume of urine is often very, very high. I was working in one project and about 70%, no, no, about 60% of their database was outpatient female urines. So of course, outpatient female urines is a very high proportion of the data and the data entry. Okay. Okay, I'm going to continue with section D called antibiotic results. And I can see enterococcus ficalis EFA and we'll put the long names in later. 100% are vancomycin susceptible. Here at the top it says percent susceptible. Staph aureus 71% are oxycylence susceptible. It's about 29% resistant. Vancomycin is 81% susceptible. There is a problem here. Vancomycin resistance almost does not exist in the world. So here you have a 19%, 81% are vancomycin sensitive. 19% are not sensitive. There's some mistake. Vancomycin sensitive almost in the world does not exist. In the last 20 years of the United States, we've only had 16 people. So there's some laboratory mistake with the vancomycin resistance. Also the vancomycin disk is not a reliable test. Let's look at imipenem. E. coli imipenem is 100% sensitive. Clepsiella imipenem is 87% sensitive. So you can see very nicely in eddy-phi isolates, there were no carpet pen and resistant E. coli. But Clepsiella, there were 13%. Any questions or comments? I'm going to go through a bit more quickly because I'm talking a lot about the interpretation, but I want to make sure you feel comfortable in actually doing the analyses. Okay, I'm going to go to section e-microbiology alerts. High priority. Let me just- I have a question, please. Yes? What was the reason behind the vancomycin was 81%. It's related to the interpretation or the quality control. I'm doubt on the data entry because, as you know, vancomycin resistance stuff not exist. So what's the reason behind in detail on your experience? Thank you. Sure, sure. And that's- and you can see also the value of these standard reports. I didn't have to spend any time putting all this together, but very quickly I have a lot of information that leads to new questions, new things that we want to verify. So that's the whole idea of doing these standard feedback reports. So again, these are more quality reports. A lot of times looking at quality and epidemiology, some things are clearly quality issues. Some things are clearly epidemiology issues. A lot of times in the middle we're really not sure. This example of vancomycin resistance stuff, that's a quality issue because in the world it almost does not exist. Basically, there's a mistake. The mistake could be in data entry, but as you said, I don't think so. They might make a mistake once in a while, but they're not going to make a mistake 19% of the time. So in general, it could be a typing mistake, or it could be maybe it's not staph aureus. Maybe it is enterococcus vicalis. So enterococcus vicalis can be vancomycin resistant. Maybe they're simply a mistake in the organism identification. I don't think that's the case here because as you can see the enterococcus vicalis was 100% sensitive. So it would be strange that all of the vancomycin resistant enterococcus have been misclassified as staph aureus. So one is mistaken data entry, but I don't think so. One is a mistaken organism identification, but I don't think so. The other reason is a mistake in the antibiotic tests. First of all, this is not a valid test. They did the vancomycin disc test. The vancomycin disc test was an authorized test in the 1990s before vancomycin resistant existed. But once vancomycin intermediate and resistance started to appear rarely, in Japan it started, then United States, in the late 1990s, they realized that the disc test was unreliable. The main problem with the disc test is that the resistance strains were accidentally, no, not accidentally, they were incorrectly. So the vancomycin resistant strains were incorrectly being called susceptible. So there's no, it would be misleading to tell the doctor it's vancomycin susceptible when the test itself is an unreliable test. So since around the year 2000, CLSI and UCAST, well UCAST didn't exist then, but the Europeans agreed, the vancomycin disc is not a reliable way to find vancomycin resistance in staff. The vancomycin disc is a valid way to find enterococcus resistance, VRE, but not a reliable way to find staff resistance. So the fact that they did the test is a mistake. It is not a reliable test. So basically, if you really are worried about vancomycin resistance, what you really need to do is an MIC. So what you might want to do in the app, like what we do is we just do vancomycin MIC and everything because we have a machine that does MIC tests. In a low resource level, what they might want to do is the vancomycin disc test. It is not a reliable test, but if you do find resistance, then don't report it to the doctor. Under those circumstances, report, do an MIC test to confirm it. And what I suspect you'll find is that the vancomycin disc resistant result is incorrect, and the vancomycin MIC test is susceptible. So you might want to do the vancomycin disc test, but I will not believe it if it is resistant. If it shows resistance, then do not report it to the clinician because it's probably wrong. Do a confirmatory test. And you're doing a confirmatory test by E test is expensive if you're doing it all the time. But if you're doing it infrequently, then it's not that expensive if you just do it a few times once a month, once every other month. So your question is why do we do it? So it's 19%. That's a lot. This is a common. This is happening way too much. If I saw two or three percent resistant, it would still be wrong. It's a minor issue. But 19% being resistant is there's a big issue here. It's important to recognize that vancomycin is a very large, heavy molecule. It's just big. Penicillin is a little tiny antibiotic that diffuses very easily, ending it with very big zone diameters. For example, 25 millimeters for most organisms and most antibiotics is susceptible. But for steph aureus, penicillin 25 millimeters is still resistant. This antibiotic diffuses very far and very fast. So vancomycin, on the other hand, is very heavy and by its molecular weight does not move a lot. And because of that, the zone diameters are small, the breakpoints are small, and there's a lot of potential for mistakes because of the method. If the medium is too thick, if the medium is too thin, if the pH is wrong, if the inoculum is too heavy, if the disc has too much antibiotic or too little antibiotic, all of these can have an impact. For example, the vancomycin disc is supposed to have 30 micrograms. But what happens if you have a bad quality disc? Maybe the vancomycin disc that's supposed to have 30 micrograms only has 20 micrograms. So even though on the label it says 30, the disc itself only has 20. And this happens with old expired disc because the disc goes bad. And even though it says 30, there's only 20. So the zone diameters, the bacteria is going to look more resistant than it really is because the disc has less antibiotic than it's supposed to. So you end up with false resistance because the disc is a bad quality disc and does not have enough antibiotic. Or alternatively, the medium is supposed to be, the auger is supposed to be four millimeters thick. But I've seen this, I've seen auger that's five millimeters, six millimeters, or two millimeters. The auger is too thick or too thin. And what happens if the medium is too thick? What happens with the vancomycin disc is that the vancomycin is supposed to diffuse horizontally to spread out. But if the disc is too thick, a lot of the vancomycin goes down. So a lot of the vancomycin doesn't go out, it goes down. So the zone diameters are smaller than they should be. So there is some quality control problem. I do suspect it's either a problem in the disc or the medium or the pH or the inoculum. How would I be able to distinguish these things? That's where we use our quality control results. There is ATCC 25923, which is a quality control staph aureus. There is also ATCC 25922, which is an E. coli. The whole purpose of these QC strains is to verify that you're doing everything correctly. I visited one laboratory in the Middle East that did not have a good idea of the concept of quality control. They showed me their clinical bacteria and their clinical bacteria 15% were staff aureus vancomycin resistant, similar to what you have here. And I said, that's not right. You can't have 15% vancomycin resistant staff. And they said, yes, but we're doing our QC and our QC is 20% vancomycin resistant. And I'm thinking to myself, it is. Now, I didn't say that. I said, your quality control strain is not resistant. The fact that you are getting 20% of your quality control resistant means if the quality control is not working, there's something wrong. The disc is wrong. The medium is wrong. The inoculum is wrong. Something is wrong. So the first thing I would recommend you do is just go back to your QC results or the QC results in range or out of range. If the QC results are out of range, that's obviously telling you that the problem is not the clinical bacteria. The problem is the reagent. So the problem is the test method. And this is where you would like somebody come in, look at your quality. So one thing you can do is just verify all of your QC results. The second thing that you do is you, next time you find a vancomycin resistant strain, what's currently happening is people are reporting it and that's a mistake. Next time somebody has a vancomycin resistant strain, say, wait, this is not possible. Let me retest it myself. Let me have somebody else retest it. Let me use the new disc, the new reagents, the new medium, and let me retest it. If one of your laboratories has vancomycin resistant staff and they say it's resistant, then they should send it to EPHI for EPHI to test. If EPHI, under every possible thing they can do, you still find it's vancomycin resistant, that's where you send it out to somebody else. Send it, well, you might want to send it to ICL, see if they have the same results, or send it to the US CDC or talk with Africa CDC. Africa CDC doesn't have a lab, or contact WHO, send it to the South African lab, send it to some other good quality lab, because if you have vancomycin resistant deaf or deaf in Ethiopia, that's an international concern. I don't think that's the case here. I do think these are quality issues. So in short, so we can't do anything with the old data because these isolates are all thrown away. So even though this has 81% vancomycin susceptible, if I'm going to report this out to the world, I would say 100% susceptible because I simply do not believe that the 19% are real. So I would falsify the data, but it's not really falsifying it, it's calling it what I believe. I believe in Ethiopia it's 100% vancomycin susceptible, and that's all I can do with the old data is I just have to go on my belief. But for the prospective data, the next time you have it, it's 90% of the time, don't simply report it. You need to investigate thoroughly. Is the QC correct? Let me retest it. Let somebody else retest it. If we really think in Ethiopia it's resistant, then send it to some lab outside of Ethiopia, because maybe it is real. If it is real, that would be global news. Okay? Okay, if no other comments, I will move on to microbiology alerts. So here are high priority alerts, high priority species or high priority resistance. And a lot of these are high priority because of their public health importance, but some of them are high priority because it's probably a quality control issue. So here at the bottom, I'm also surprised. Okay, let me go back to the organism, stephorius. Okay, stephorius, there are only 91 of those. I also don't know how often they tested vancomycin. The thing that I noticed here, the thing that I noticed under high priority alerts, 90% sounds like a lot, but in fact, it's only two. So in other words, so two divided by 91, two out of 91 is approximately 2%, not 19%. So this 90, I'm feeling a little bit better now, because here it says 81%, but that's not 81% of the 91 organisms. Because this is so important, I'm going to take a little break from my quick analysis. Let me go back to normal analysis, percent resistance, stephorius. Let me choose this one month of data. Such an important issue. I do want to wrap this. I do want to make sure we have a good conclusion. Let me test it. Okay. So here we have the results. Let me go to vancomycin. Where's the vancomycin? So this is the detailed reports. Okay. So let me choose something like the oxacillin. Let me put this in alphabetical order. Okay, oxacillin. So oxacillin, so at the top of the screen, you see it says 91 isolates. We have 91 isolates of stephorius. From those 91 isolates, we tested oxacillin 69 times. We tested penicillin 55 times. We tested nitroferentine once. We tested tobromycin twice. We tested vancomycin only 16 times. Okay. So let's see. Yeah, and I even have a few more comments on this. So there is a limitation in my standard report is it's not giving enough detail. Here I have a very good idea of what's happening and the problem is not as bad. It's not as bad as we were thinking. It's not 90%. So I'm happy we looked at this in more detail. So this is raising a different issue. I'm going to sort this by number. Here it says number. I'm going to click on that once, click on it again. I'm going to go to the top. So here you can see sephoxitin. They tested 71 times. Oxacillin 69 times. Trimethoprim sulfa 68 times. Cipro. So these organisms, these antibiotics are tested most of the time. This is your standard panel. Here at the bottom, they only tested nitroferentine once. Tobromycin twice. You see it's 100% sensitive, but it's not 100% out of 91 isolates. It's 100% out of 2. So one thing we're missing on that other screen is the denominator. So for vancomycin, they only tested it 16 times. So it's not 81% out of 91 isolates. It's 81% out of 16 isolates. 6% it says resistant. 6% it says intermediate. And then we have this other one. 6% is question mark. The question mark is interesting. Let me go to the vancomycin. Okay. Well, there's a lot of different issues to discuss here. So let's see. Okay. Here, look at this top row here. You see that they tested, okay, they tested clindamycin 75 times. That's what this number 75 means. The 64 here is the number with measurements. So here there were 75 clindamycin results with 64 measurements. Here we had 55 penicillin results, but only 33 measurements. I'm kind of guessing at the beginning of the last year, maybe you were entering the letters R, I, and S. But then later in the year, you started to measure. So for example, with the vancomycin, you tested it 16 times, but only one of them had a zone diameter. So 15 times, 15 times, it was, 15 times, it was simply the letters R, I, and S. Once you actually put a zone diameter. And the zone diameter is related to this question mark here. I'm looking at my graph, and you see the zone diameter was 6, completely resistant to vancomycin. Of course, I don't believe it's true resistance. But the zone diameter is 6, there are no break points. Let me look at the table. Look at the break points. Where are my break points? There they are. So you see the break points, 14 to 17, greater than or equal to 21, 13 to 17. But there are no vancomycin break points. So if there are no break points, it's not a valid test. So basically, even though 90% sounds like a bad problem, it isn't a bad problem. This, let me redo this one more time. I'm going to options here. Here under options, you see under my percent resistance, I'm looking at the percent of isolates. I'm going to change it to number of isolates. Number of isolates. I click on OK, begin. So we are no longer looking at percent R, I, and S. Let me just shift this around a little bit. Shift this around. Let me hide this. Let me hide that. Okay, good. So here what you see are not percent RIS. We don't need this, and we don't need this. We have, for example, penicillin. We have 55 isolates tested for penicillin. 47 were resistant, eight were sensitive. So 47 plus eight is 55. Oxicillin, we have 69 results. 20 were resistant, 49 were sensitive. 20 plus 49 is 69. From the 69, 13 of them had a zone diameter. Cefoxitin, 71 results. 70 of them were zone diameters. Okay, I'm going to go to the Venkamaicin. We have 60 isolates. One was resistant, one was intermediate, 13 were sensitive, and one of them was no interpretation. That's what question mark means, because there's no breakpoint. So even though I was saying 90% looks bad, in fact, there was only a problem with two isolates. This one is resistant. This one is intermediate. I don't believe it, but it's only two isolates. And then this one, they measured it, but it's not a valid test because there are no breakpoints. So in short, I was worried that you have 19% of resistance, but the reality is it's not really that bad as 19%. Most of the time, you do not test it. And this is a drug, you have 91 staph aureus, but you only tested the Venkamaicin 16 times. The fact that you tested it was wrong. And further 16 times, one they said resistant, one they said intermediate, one they said six millimeters. So there were only a problem with three isolates in the year. So the problem is not as bad as I thought. Okay, so that's where this detailed view is giving me more insight than the summary review. And we will be adding more of these details to the summary review, you know, just to make it more of the information there in front of you. So I hope that explains some of it. So there is a problem, but there are two problems. One is that at some point last year, they were doing some Venkamaicin disc testing. Once they got an R, once they got an I, once they did put the zone diameter, I'm glad they put the zone diameter, but it's still not a valid test. So in all of last year, out of the 91 staph aureus, they did an incorrect 16 times. And twice it was not sensitive. So it is a problem, but it's a small problem, not a big problem. And my guess is that you stopped testing Venkamaicin disc. So a lot of times you see this a lot of times the problems are at the beginning of the time period. Later in the time period, people learn to do better to doing appropriate testing. So the problems are still there in the early data, but they're not there in the older data. Any questions or comments on that? Your question related issues. So I hope the separate discussions were helpful. Yes. I'm agreeing with you. Last year, we don't have enough amount of Venkamaicin. The Venkamaicin disc was failed repeatedly because of the poor brand. So they have stopped to test for those 91 isolates. That's why the reason behind. So maybe it's not tested because of repeatedly the quality control was failed of Venkamaicin especially. Great. So there are only two cases where the result was bad. The R was bad, the I was bad. So for those two isolates, further investigation is warranted. On the subject of number tested, this is a very important graph here. I'm sitting looking at how often did they test penicillin? How often did you test oxacillin? In fact, testing staff orders by oxacillin disc is no longer recommended. Instead, the recommendation is to test the sephoxidin disc. So I see you have a lot of data for, well, first of all, I see there's a lot of variability. There's nothing that you test all the time. You test glendamycin, that's the best. Sephoxidin is about the same, oxacillin about the same, Cipro and SXT, penicillin a little bit behind it. So these are your core six drugs. You also test gentamycin about half the time. Isitromycin, erythromycin, tetracycline. You very rarely test tobramycin, daptamycin. We have these even twice. There must be an MIC or something in here. Nitroferantoin, chloramphenyl. So what we really want to encourage all of your labs to do is to maybe decrease the number of antibiotics, but increase the completeness of testing. And a lot of this has to do with gearing towards them through a certain priority set and also working with procurement to make sure that people have the discs of good quality. So basically, when I look at this, I'm just going to ignore the tobramycin, daptamycin. I'm not going to ignore them completely, but that's a friend to them. We only tested it once. I'm glad that it was sensitive. It's 100% sensitive, but still it's only one isolate. So you really want to steer towards people towards a minimum priority agreed set that everybody does to their best of their ability. You can also do additional drugs. Daptamycin is not, daptamycin resistance is extremely rare. Linese lid resistance is extremely rare. So I would not necessarily recommend it for first decline testing. But second line testing, you know, EPH, I might want to do this. So the problem with tobramycin, daptamycin, is that tobramycin is the results are not going to be representative. Okay, so the results won't be represented. So what I'd love to see here is I'm going to show you basically the my, the WHO test data. I'm going to remove your Ethiopia data. And instead, I'm going to put the WHO test data that you already do have. It's included as part of HUNET. I'm going to show you the same graph for staph aureus. So I'm looking at the same organism, but you can see a different hospital. And they test they're very consistent. They test the same antibiotics all of the time, except for natuferantoin, because it's not because of stock outage. The reason is because it's a urine drug. So in this hospital, their policy is always test one, two, three, four, five, six, seven, eight, nine. So every step for is they test nine drugs. If it's a urine, they will also do nitroferantoin. So this is what I'd like to see, consistent testing of a standard set. This hospital also has something interesting with their E. coli. It's a little bit more complicated, but it's still very systematic. You see there's one, two, three, four, five, six, seven. There's seven drugs, one, two, three, four. There's seven drugs that they always test. There are three drugs, AMC, nor, and night, nitroferantoin, that they only test in the urine. There are three, PIP, Fox, and SIP, that they test in the non-urine, like blood. And there are three, septicidine, amiccation, and turbromycin, where they do second line testing. So even though this graph looks complicated, it's still good testing. There's seven drugs that they always test. If it's a urine, they do these three. If it's not a urine, they do these three. If it's resistant to a bunch of the first line agents, then they will do some supplemental second line testing. So even though this is, so here there's no stock outage. They've done this on purpose. They've done this delivery. Always do this, always do this in urine, do this if it's not urine. If it's resistant to the first line agents, then do the second line agents. So even though this looks messy, they did this deliberately. So I'm going to go back to your dataset and let me go to, here it is. We're just looking at the time. We still have 25 minutes. Number tested. So, oh, that's the E. coli. Let me change this back. Well, yeah, we can discuss it for the E. coli. So for the E. coli, I see that you always test meropenem. So let me look at the table. So here at the top, you see it says 126 E. coli. It's at the very first line of the screen. There are 126 E. coli. You tested SXT 122 times, Cipro 119 times, Meropenem 118 times. These three, you more or less always tested over 98% of the time. But then you have Tobromycin, Nitrofranzoine, Peptazo, then you test about two-thirds of the time. And then if I go down further to the very bottom, Zythromycin, Penicillin, some of these were probably typing mistakes. You should not be testing Penicillin on E. coli. So this could easily be a typing mistake. It's just one or two. Ertypenem, they only tested four times. So you do have to be careful about biases because they didn't test it very often. So you have a lot of antibiotics here. I would kind of recommend decreasing the number of antibiotics but increasing the testing of those antibiotics, okay? And to try to standardize that around the country. Usually for a gram positive, you maybe have a minimal set of like five or six, you know, for a gram negative, maybe like eight or 10. But it does come out to much realistic. Fortunately, these discs of everything we do in microbiology discs are one of the cheapest things. So improving medium and, you know, or identification and quality control, all those things are expensive. Buying more discs is usually one of the cheaper things because, you know, discs, it's like two cents for each disc or something. Of course, different countries, the costs are different. So this is what your testing pattern looks like for E. coli. There are too many drugs and the drugs that are tested are not tested systematically. Like here we have seftraxone and sephotaxine. You don't really need both of them because it's basically the same drug. Imipenem, you don't have a lot of imipenems. You do have a lot of meri-penems and that's good. You don't have to have both. The drugs are a little bit different but they're mostly the same. So trying to standardize people in this common direction. I want to show you a different issue that I noticed here. Look at the sephotaxine here. You see that there are two different sephotaxines. The C-T-X means sephotaxine. The D means this diffusion. The five means five micrograms. The 30 means 30 micrograms. The second one is correct one. If you are doing CLSI testing with sephotaxine, you are supposed to use the 30 microgram disc. The 30 microgram disc has break points. The five microgram disc has no break points. So here somebody made a mistake. Either they tested the wrong disc but I don't think that's the case because to test the wrong disc you have to have a five disc. I don't think you have a five disc. I'm guessing that there was some laboratory configuration issue. Somebody at the beginning who maybe did not understand who in that configuration created a sephotaxine five microgram disc. They entered three results but then they did the correct one, the 30 microgram disc. So this is a different problem. This is a different kind of problem. I don't think this is a laboratory test problem. I think there's a Hunnet configuration problem. Somebody accidentally created a sephotaxine five microgram disc when the correct one is the 30 microgram disc. So that's why here we have two different sephotaxines. The CTX on the left is the wrong disc potency and the second one is the correct disc potency. So that's a different kind of problem. It's not a test problem. I think there was a small mistake in the laboratory configuration. I also noticed that I noticed that for something else. I noticed that for for the E. coli that's the only problem which is great. There was one for the staph aureus. I'm going to go to the staph aureus and there was a problem with the azithromycin. Let me do this alphabetically. So you see there are two different azithromycines. Let me make this narrower. Let me make this narrower. Let me make this narrower. I'm trying. It's not, doesn't want to go narrower. That's fine. That's enough. So here you see there are two different azithromycines. There's a 15 microgram disc which is the correct one. It has break points. There's a 30 microgram disc which is not right. There are no break points. So 32 times they have the correct disc potency. Four times it's an incorrect disc potency. So what we could do is we could fix this. We could clean this up. You know just to put these azithromycines together. Same thing for no actually the others look okay. I don't see. So there's a small problem there that they just they put an incorrect disc potency. They put a couple of results and then they put the correct one. The correct one has most of the data. 32 results. Any questions on some of these? I keep on jumping to different points because you know there are different issues that I'm looking for all at the same time. Is there any standard to either decrease or increase the number of drugs for each pathogen? For example you say for E. coli they tested seven drugs. Is there any standard to make it five or eight or nine or is it consensus in each lab? Let's see. I'm going to close out of this and I'm going to go and we've discussed this before. So the goal is not to get a certain number of drugs. The goal is to choose the right drugs. So I'm sure I'm going to go to Google. CLSI free. So free resources. I'm waiting for the website to come up. It's going a bit slow. I don't know why it's going slow. So in short and I've discussed this previously. CLSI has some general recommendations for what you should be testing. But it's just a general that nobody does exactly what CLSI recommends. People do more. People do less. It depends on what's available. It depends on their resistance problems. It depends on how much money they have. If you have more money, you know, test more discs or do these M.I.C. panels, which have a lot of antibiotics. The website is not coming up. That's okay. I have a copy of the document. So if I go to this, if I go to here, if I go to C drive, I do have these documents. There are commercial documents. So I, you know, purchase these in January. Released. Good. So there's something here called table one A. Okay. Why is it having trouble finding it? Okay. Okay. There it is. Okay. Let's table two A. Let me go back. Okay. Almost there. That's it. Okay. So here you see is at the up in the green area at the top, it says table one A suggested antibiotics for testing for non fastidious organisms at table one A. Then there is table one B suggested testing for fastidious organisms. So here you see hemophilus, myceria, strep, that table one B. And then finally, there's table one C, which is table one C suggested anaerobic grouping. So gram negative anaerobes, gram positive anaerobes. Okay. So let me go back to table one A, which is the normal bacteria. Table one A. So here you see, at the beginning of this, you see enterobacteriales, like E. coli, clepsiola, pseudomonas, staph, and enterococcus. That's the first part of the table. I'm going to go down further. I think I went too far. Right. And then here you see this. So, so, so here you see everything on one page. So at the beginning, you have enterobacteriales, pseudomonas, staph, and enterococcus. At the second part of the table, you have the non fermenters, acinetobacter, oposite, synomonas, acinetobacter, burcal, deris, genitropomonas, multifilia. Like some drugs, some things like genitropomonas, multifilia, they only recommend three drugs. For acinetobacter, they recommend a lot of drugs. Let me go back to the beginning. Where's the beginning of the table? Table one A. Here it is. So here you see for staph, lacoccus. Okay. So group A is recommend, always test, always report. Group B, always test, selectively report. So if I look at all of the enterobacteriales, that's a lot of drugs. If I look for the staph, it's a smaller number of drugs. And there are a lot of these drugs, like septaroline, vencomycin, has to be by MIC testing, linazolid, that's more for a high resource place or a low resource place as supplemental testing. So the full group of enterobacteriales is a longer list than for staph. For staph, they're just fewer options. You see there are different footnotes here, like for the oxacillin. They're saying oxacillin by MIC, if you look at the footnotes. But suffoxitin by disk. So this is the global recommendation in a general sense for everybody in the world. But you should look at this, but you should make your own list for Ethiopia. So in Ethiopia, like a lot of places, like CLSI recommends testing penicillin for staph, but a lot of places don't bother with penicillin because resistance is usually 85, 90, 95%. So a lot of people test, a lot of people don't test it because it's usually resistant. It is useful epidemiologically, but resistance is so high, it's usually not a useful treatment option. So I suggest that you look at this list to inspire what an EPHI Ethiopia list would be. And the staph list is usually a shorter list, like here azithromycin, clindamycin, oxacillin by MIC, suffoxitin, and maybe not the penicillin in the SXT. That's really just one, two, three, four drugs. Yeah, and then maybe some of these others, like the tetracycline or rifampicin. And then you want to decide what are drugs, what drugs are your clinicians using. If your clinicians do not have access to lenezolid, you don't need to test lenezolid unless if you have an epidemiological scientific interest, it is interesting. But you want to correlate this with what the drugs are that your physicians have. So the staph, there are not a lot of drugs that your physicians have to choose from. Enterococcus, it's even a shorter list. Ampicillin and vancomycin are really the core drugs. And then if you have lenezolid, deptamycin, etc. The endobacteria, the endobacteria, you see there's so many. You want first generation, second spore, and third generation sephalosporins, hemipenins, carbopenins. So there's a lot more choices going on there. So when I said that staff is like maybe five and grand negatives, number eight, I'm not aiming for a specific number of drugs. I'm just reflecting the drugs that doctors want to choose from how many choices they have. They have a lot of choices for treating grand negative infections. They do not have a lot of choices for treating grand positive infections. So I think as a network, if you come together, you can EPHI proposal list and then and then bring the network people together and or share the list with them and say, do you agree with this list? If it's a urine, then do this. If it's not a urine, do this. And I think that you can come up with an Ethiopia consensus for minimal testing and of course if people would want to do more, they're welcome to do more. Does that make sense? John, I understand. Good. And then in terms of their test practices, I want to start giving people scores. I don't want to penalize them if they don't, I want to give guidance. Maybe for a staff or as we have five minimal recommended drugs and maybe let's say they have 100 staff. So if they have 100 staff and five minimal antibiotics, they should have 500 results. Out of 500 results, maybe they have, you know, 490 results, meaning excellent. They tested almost everything. Or maybe they only have 100 or maybe they're doing great on penicillin and oxacillin, but they're doing bad on clindamycin because they ran out of discs. So if you, so I want to come up with the minimal set that we can use for scoring. Please do this minimal set because we're going to score you on these on completeness of testing. If they do not have complete testing, often it is not their fault. It's two things. Do they know what they should be doing? And secondly, are they having trouble? Are they having trouble getting the discs or the discs are failing QC? So I want to score them on like five minimal tests and encourage them to go in a good direction. I don't want to penalize them. Like if they're not testing Linazolid, you know, if they are testing Linazolid, it's nice to know that. But if they're not testing Linazolid, I'm not going to give them, I'm not going to take that off of their score. I do, I will want to take off of their score. If they are testing things that they should not be testing. You know, for example, if they're testing vancomycin on E. coli, that's an obvious mistake. I mean, it's an ineffective drug. Don't test vancomycin on E. coli, you're just wasting money. But there are other things that are not so obvious. You should not test amoxicillin on E. coli because there are no, even though amoxicillin is commonly used for clinical therapy. If you look at this list on the left, you do not see amoxicillin. You do see amoxicillin, clogilinic acid, but that's different. That's augmentin. That's a different drug. So CLSI does have recommendations for ampicillin, but they do not have recommendations for amoxicillin. So I want to penal, I want to score them on their ability to do minimal testing. And I also want to penalize them if they're testing inappropriate drugs that do not have any break points. Okay? So, and my, the whole goal of the early analysis, as all of your labs are going to have different kind of quality control issues. So we have spent most of our time today, today discussing primarily quality completeness issues, because those are really where you want to start. Of course, we're getting to the epidemiology issues, but you don't want to interpret the epidemiology issues until you feel that you have good quality data to work with. And often what happens is you have good quality data for E. coline steforius, because they're easier organisms, but maybe unreliable data for strep pneumonia enomophilus, because they're more difficult organisms. Similarly, there are certain drugs that are easier than others. Imipenem is a very unstable drug. So sometimes you end up with false resistance, because the imipenem disk has gone bad, or the vancomycin disk, it's a difficult disk, because it's a heavy molecule. So sometimes what you will find is you will have good QC results for ampicillin and erythromycin. So I do have more trust in those data, but I'll have less trust in the imipenem and vancomycin results, if I, especially if the QC results are not good. Other questions? So, in answering your question, yeah, look at the CLSI document. You can get it online. But this is just a general principle. Don't do exactly what they're suggesting here. This is meant for the U.S. audience. The concepts apply everywhere in the world. So section A is always test, always report. Second section always test, selectively report. Third section is called group C. Here you see group C. This is supplemental optional testing, like second line testing. Like in the United States, we don't usually consider carmphenicol, because it's potentially harmful, but sometimes it is usually still effective drug. And then you see group U or urine drugs, like nitroferantoin, trimethyprim, all by itself. Usually you would only consider utilizing these for urine infections. So at the top, I showed you the E. colisodomonas stuff. And then here at the bottom, it continues with esonidobacter, ferclderias, genotrophomonas, and other non-fermenters. I'm going to close the CLSI document. I'm going to go back to Hunet. And there's only 10 minutes left. I just want to return to my standard report. Okay, I'm going to go to data analysis, quick analysis. Standard report. I haven't discussed these other three. So that can be a good subject for next call. Because everything I am showing you now with the standard report is valuable, but you cannot change it. It is what it is. I can say screen Excel. I already showed you the Excel when I started this call. Let me just go back to the screen version. I'm sorry, I double clicked on this. Against the analysis, yes. Okay, begin analysis. Whoops, I accidentally, this is only funny with my mouse. Against the analysis, yes. Okay, begin analysis, great. Okay, so it's very fast. It can be slow if the data sets larger, of course, but this is one year of your EPHI data. So we talked about a high level summary, B data fields about validity and completeness of data entry, and a summary of like male versus female, hospital 001, hospital 01. At some point in the future, we can have a call on cleaning. I've highlighted a point here, but I haven't shown you how I can standardize that. Right now we have two lives, but it's not too different. It's the same life. So this is some valuable epidemiology data and quality data. Organisms are the most common in other organisms, useful for epidemiology, looking at positives and negatives, antibiotic results, microbiology alerts. So this, we ended up on the detour because the vancomycin non susceptibility that 90% was not as bad as I thought. In fact, there were only two. So even though you have 91 step aureus, they only tested vancomycin 16 times. One was resistant. That's here. One was intermediate. That's here. So there was only a problem with two of the isolates. And then medium priority alerts like ESBL, MRSA. At the hospital will often be interested in these medium priority alerts. At the national level, usually you're just going to be interested mostly in the high priority alerts. You know, you are interested in the percent MRSA, but you do not need to see a list of the MRSA. This is a list. I don't think you need to see their list of MRSA, but you do want to see. So I think you want to see the statistics for the medium priority alert, the percent MRSA, but you only want to see the list of the high priority alerts. And of course, different people will want different things. These are my two vancomycin non susceptible staff. Here's a vancomycin non susceptible staff. Regulus negative. Here are my carbapenem resistant E. coli's and Klebsiella's and intrabacter. So if you're getting data. So at the national level, if somebody sends you a data file, at the end of one year, you have a relatively small number of high priority alerts. What is this? This is maybe 25 isolates. So 25 isolates out of 1800. So if you get this on a monthly basis, it helps you to find quality control problems as well as important new resistance that you want to be aware of and confirm. Low frequency alerts. There's nothing there. So I'm not going to talk about it right now. Lab configuration. You know, the lab configuration, this is more about your Hoonet configuration. You have the following fields. You have 49 antibiotics. You have no, there are no breakpoints for these drugs. So all of these are incorrect dispotencies. I already pointed out to you, azithromycin, the 15 microgram disc is correct. The 30 microgram disc is incorrect. Cephotaxine, the 30 microgram disc is correct. The five microgram disc is incorrect. So basically, why are there no antibiotics? Either there are no official breakpoints or because you chose the wrong dispotency. So basically, I'd kind of like to delete these because they are incorrect. But they do have a little bit of data. You know, for example, something like this, I don't think you have, I think this was a complete mistake. Cephapim is at a back-ten. That's a brand new drug. I don't think that you are purchasing that it would not be a good use of resources. So some of these are columns that probably have no data. So I just like to delete it. If there's no data, there's no sense in me analyzing the data because there's no data. On the other hand, cephotaxine, there were a couple of incorrect results. I'd like to move those over to the correct results. So we can talk about this later. What we are showing you now is how to find problems. On a different call, I will show you how to fix problems. How to delete columns that you do not need. How to merge columns that you want to put together. How to change that 001 to 01 so that they match. Column G, column H, no issues there. Column I. In that data file, you sent me who not found no invalid results. So congratulations. I'm not talking about the microbiological quality, but there was no gender equals W. So there was nothing wrong in terms of your data entry, you know, that who it was able to find. So I'm going to click on okay. So, and if I just say here Excel, export to Excel, begin the analysis, it does exactly the same analysis. It does take a bit longer because it has to save it to an Excel file. Found a mistake that we'll have to fix that. I'll worry about that later. It's like everything will work except for the last thing. So that's about the end of the time. Are there questions? We focused today on the, we focused today on the, the, the predefined who net report. Next time I think it would be good if we focus on the user defined reports. I go to EPHI. I go to data analysis, quick analysis. Here you see or I should alert report organism alerts, patient and sample statistics. If I click on edit, you see there are a lot of macro. So we'll talk more about macros on the next call. I want to do sex by laboratory, age group by laboratory, location by laboratory, specimen by laboratory, or organisms. I want to see organism by lab, organism by date, organism by sex, organism, I want to see percent RIS for step four is the poli. So these are configurable. You can decide what you want here. For example, if you don't want this, you can just remove it. You can add and remove macros to match what you would like things to be. Let the little preview enter into our next call. Also, I sent you everything by email. So feel free to jump ahead and start looking at what I sent you by email. If any questions, then you can let me know. Questions? I don't have a question, but did you send the email yesterday? I didn't receive that email. It's right here. So this is the email. These are the people I sent. So the subject is who net data management online training part six. Thank you. Oh, you found it. Great. Great. Great. So just as a reminder, the word is a general concept and I laid out a lot of what I said is described here as a word document. And then you see the standard report. That was the focus of today's call. And then you see two, three and four. Those are the user defined reports that I created for you, as well as the macros I used to create those. The macros, if I open up the macros. So here you can see all those macros. We discussed macros previously. If you want to use these macros, you simply need to, the file that I sent you, save this to your computer and then unzip it. And then you put them into the who net macros folder. So this is where you put macros, who net macros. So if you want to get ahead and play things around, you can simply unzip those macros into this folder and then you can start using them immediately. But you don't have to. This is just if you want to jump ahead. Or if you go to who net, there's also a folder called documents. You're under documents. There is a document here called, where is it? Macro and Excel reports. So this macro and Excel reports can also give you further background. But there's no need to do that before the next call. The only reason you would do that is if you wanted to kind of just jump ahead. But these will be things that you can review after the next call.