 Hello and thank you for having me, Sarah I'm suspect that maybe you don't like me very much. She's asked me to talk about an EMR straight after that talk so bear with me. I would also mention I'm on succumbent from the NHS and I'm sure that has nothing to do with it. Okay so I'm here to talk about an evaluation that I did in February of the electronic medical record for emergencies and we did this in Bokoro in Chad. Okay so just briefly well I was going to explain to you what an electronic medical record is I think that's probably redundant it's a computer program that allows you to access and work with individual patient data and this was a particular kind of patient record that was developed initially for use in the Ebola crisis. It's designed to work in the sorts of locations that MSF sometimes works and you can see an example here on the screen where there is unreliable power there is no internet and the work that was done in Ebola was then taken and redeveloped with the vision being that this would be something that could be taken to the field could be configured at short notice for a number of conditions where a clear view of key patient information would help you improve care and the possible use cases that we were thinking of were things like happy cholera Ebola malnutrition. Okay so this is the problem that the development team were trying to solve in the field your paper records often look like this repaired some of the data is difficult to read they get dirty torn lost and this was the solution you can see on the left the is this the pointer yeah so on the left is an example of the patient screen if you like where you can see key patient data and on the right is one of the forms in which you enter data so it's important to note that this electronic medical record was designed to replace paper it was to be used instead of paper at the patient bedside and the pilot that we did um was uh an attempt to see how well that works okay and just to briefly summarize exactly how that it works on the left you can see a little server it's an intel Edison chip running with a database uh an open medical record software open MRS I'm sure you've heard of and on the right you can see the tablets which are connected via a Wi-Fi network to that server uh and therefore everyone in possession of a tablet if it's working properly will have an up-to-date view of every patient in the facility so the development team were there to pilot the software my job was to evaluate the readiness of this software for deployment in the sorts of settings that we're talking about so um i'm not going to go through all of this slide it was simply to say that there was a very careful process we went through to try and ensure that we were covering the key elements of this technology and you can see that we were looking at training needs its ability to be configured its stability its acceptability and the ease with which people could access data amongst other things so we're talking a little detail about the methods because these are kind of important we had an audit there were 25 key requirements for the system that covered um different aspects of the system could it work in an offline environment uh did it do everything that the paper record could do did it have the necessary safety features and those audits acted also as a safety check that were done before during and after the pilot i had a list of key um uh key tasks that occur in malnutrition care and i was to look at those when people were using paper records and then when they were using the electronic record and see what differences we could observe i did interviews with staff and to really find out their views and experience of using the technology we designed a comprehension test to test the hypothesis of whether actually that rather attractive user interface that i showed you would that allow you to access data more quickly and more accurately than the paper records uh and i also did something called a delphi questionnaire essentially a way of generating consensus among experts who were geographically dispersed and we did that to try and um assess the technical strengths and weaknesses of the system so to give you some results the last the audit we conducted um when we came home you can see that a number of the key operational criteria were met okay and these included particularly the ability to replicate what was on paper records and some of the key offline features unfortunately also a number of very important criteria were not met and actually um because of the results of this audit halfway through the pilot we changed our plan from one of complete implementation and removing paper entirely to a dual entry using paper and electronic records to maintain patient safety and again as much learning as we could from the pilot um and probably one of the key points here is that um the the the stability and the ability of the server to handle the load of data that a normal number of patients would generate was actually not uh as good as it needed to be uh we did the comprehension test and actually we found no significant difference between paper and the electronic record in terms of the ability to access information and the accuracy with which that information was processed there are no standardized tests for this we devised our own we did it in the field but that was as good as we could do uh interviews gave us quite interesting information when I interviewed people absolutely everybody I spoke to was desperately keen to use the technology and I think this was it was obviously a real source of personal and national pride in an environment um in Chad where most people have had no exposure whatsoever to technology uh to have the opportunity to do that was obviously a big plus for them but when we observe people's behavior you could see that engagement varied some professions were more engaged than others interestingly nutritional assistants were probably the most engaged uh and particularly for those in senior positions as the workload increased perhaps inevitably given that we were double entering data the engagement may be decreased staff said that they felt that they had a requirement for a training need of perhaps a month to become really comfortable with using this technology uh and that does not sit with the vision of this which was for something that could be rapidly deployed in emergency um and because we had 13 patients over 12 days on this system uh we couldn't sensibly assess whether or not there was a benefit to patients from using this the results of the Delphi questionnaire um we sought the views of six of the seven development staff who'd worked on the project while I was involved with it um and there are a number of key technical strengths the system was designed from the ground up to work in an offline environment and you can install back up and restore the software from a usb stick which is an innovative feature which worked extremely well during the pilot you can also configure the forms the pieces of data you want to record how they displayed or that can be configured via an excel spreadsheet uh and that although it was a draft version also worked well there were however some serious weaknesses to the system it had been developed by a number of different developers uh often in somewhat of a hurry and as a result we had a very complex code base which contributed to the instability of the system in addition uh as I've mentioned before the server didn't have sufficient power to deal with that workload and that's somewhat dry point about synchronization actually translates to can you trust the information that you see on your screen when you're treating a patient um if you have a server holding a database of patient information you need to know that the information on your tablet is 100% reflective of the information on that database otherwise you are potentially going to make decisions based on inaccurate information and on a small number but important number of occasions we observed some problems with that so to conclude what went well I think we demonstrated or the team development team demonstrated that an electron medical record for use in emergencies that could function in an environment without power without access to the internet could be rapidly configured is technically feasible some of the innovative features worked well and the user interface is very attractive and has uh some of that will be used in other work ongoing in terms of my study limitations as I mentioned we only implemented this kit for 12 days with 13 patients so there's a limit to the inferences that can be drawn particularly around impact on patients however it's important to note that the EMRE as is is not ready for deployment for the reasons listed there what I wanted to do was then place this in the wider context of the use of electronic medical records um in humanitarian settings within operational center Amsterdam or MSF Amsterdam we are using electronic medical records in certain circumstances already for example outpatient care of TB and HIV and that's been shown to be helpful and effective this particular piece of kit um is a very specific use case aiming to replace paper and to be used to be configurable for an emergency and I think you know this work started with Ebola we broadened its scope to several other scenarios and we have come back I think to a very clear use case for where we think something like this might be useful where you have a need for a very small number of data fields where a standard data collection app uh would not be sufficient uh where paper presents serious problems for example with a severe biohazard you can't get the data out of the red zone to analyze it to then understand what constitutes effective care uh and perhaps where we have poor existing outcomes which would justify the investment and the risk involved uh finally all the code that has been all the work that's been done on this is on an open source platform and it's freely available to anyone who wishes to interact with it uh I leave you with this my background is in the NHS I've been in the NHS for 10 years and I'm on a year long succumbent to MSF um and I guess I leave you with a question as to how does MSF we have very different attitudes towards risk in the NHS compared to NSF and that's one of the things I've observed so in MSF you have a high tolerance for risk partly because you have to because of where you or we work in the NHS we're much more risk averse and how does MSF innovate but protect the interests of the people that it serves how does it do new things that are required without placing unnecessary risk on those patients that's the correct I'd like to leave you with um I acknowledge everybody involved in the project and thank you very much for your time and attention