AV-Portal 3.23.3 (4dfb8a34932102951b25870966c61d06d6b97156)

Tupaia - Health Resource Mapping

Video in TIB AV-Portal: Tupaia - Health Resource Mapping

Formal Metadata

Tupaia - Health Resource Mapping
Alternative Title
Tupaia - Health resource mapping in Asia-Pacific
Title of Series
CC Attribution 3.0 Unported:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
Release Date

Content Metadata

Subject Area
Tupaia - Health resource mapping in Asia-Pacific
Mathematics Wave Existence Mapping Personal digital assistant Texture mapping Chain Video game Pattern language Navigation Row (database) Physical system
Point (geometry) Mobile app Group action Overlay-Netz Service (economics) Open source Dependent and independent variables Multiplication sign Source code Information systems 1 (number) Mathematical analysis Quality of service Staff (military) Machine vision Information technology consulting Independent set (graph theory) Different (Kate Ryan album) Single-precision floating-point format Visualization (computer graphics) Energy level Multiplication Computing platform Overlay-Netz Source code Dialect Dependent and independent variables Mapping Information File format Multitier architecture Mathematical analysis Electronic mailing list Usability Database Staff (military) Multilateration Data management Uniform resource locator Visualization (computer graphics) Chain Computing platform Self-organization Quicksort Routing Physical system Row (database)
Ocean current Mobile Web Texture mapping Information State of matter Point (geometry) Archaeological field survey Mobile Web Source code Moment (mathematics) Projective plane State of matter Archaeological field survey Neuroinformatik Bit Digital signal Abelian category Physical system
Wechselseitige Information Group action Presentation of a group Transportation theory (mathematics) Distribution (mathematics) Multiplication sign View (database) Decision theory Source code Archaeological field survey Open set Mereology Quality of service Computer Front and back ends Data management Facebook Cuboid Office suite Physical system Area Source code Texture mapping Arm Building Moment (mathematics) Menu (computing) Staff (military) Instance (computer science) Digital photography Data management Chain Configuration space Right angle Quicksort Physical system Slide rule Functional (mathematics) Server (computing) Mobile app Graphics tablet Service (economics) Social software Open source Mobile Web Mass Content (media) Staff (military) Field (computer science) Wave packet Spreadsheet Goodness of fit Green's function Energy level Computer worm Gamma function output Summierbarkeit Computer architecture Dependent and independent variables Key (cryptography) Information Copyright infringement Weight Lemma (mathematics) Archaeological field survey Projective plane Mathematical analysis Content (media) Planning Stack (abstract data type) Software maintenance Loop (music) Visualization (computer graphics) Software Personal digital assistant Revision control Routing
Uniform resource locator Email Arm Texture mapping Service (economics) Software developer Range (statistics) Source code Moment (mathematics) Self-organization Website Right angle
thank you very much and thanks for having me here today bye I guess I would start by saying I'm not necessarily a mapping legend I'm not a legend at anything and I'm certainly not a data legend I'm a pharmacist originally and very much just into health system strengthening I'm still kind of stumbled into this role in in a lot of ways to meet just a need that we had around health supply chains so hopefully it's not too many techie questions but this is my favorite question of all what does tip I I mean tip I was a Pacific Islander he was a navigator who joined Captain Cook's crew in 1769 and sailed with Captain Cook through the Pacific helping them find islands that they didn't know existed and in a lot of cases which 2/pi himself hadn't even been to but just knew of their existence and was able to find them using the stars bird life and wave patterns he died on his way back to Europe in 1770 I always say he died of the record shows he died of malaria or dysentery but those two diseases don't resemble each other at all so I suspect he just died of something and and that's what they wrote down so it's kind of named in his
honor and it's a data aggregation analysis and visualization platform that allows us to overlay different health data sources and present them in a mapping lead format so we have a data collection app like everyone does that allows a single data entry for H is or what we call package of essential health services or other clinic assessments and we can route it from that app into different databases but our preferred source for data is just passive data that exists already in hrs databases or in allymeier logisitics management information systems so stock management for pharmacy and and consumables the
idea is to break down data silos the idea for to Peyer came out of a concept that I was working in Solomon Islands in 2008 and this DFAC consultant came up and said that she'd just come back from a clinic in Western Province that had node and we were well I was like what creek was it you know you know and she said I will I don't I can't remember the name of the clinic we went to a few but this one didn't have any you know any medicines I said well it must have had some medicines you know they've usually got some when you go like which ones did it having which one did it not she didn't know they hadn't made a record but could we do something about it in the pharmacy vision and this is very very common I mean that was that was kind of a funny and extreme example but we constantly had people coming in and telling us about some problem that they'd found in some clinic there was very poor record of it they'd looked at one item and it wasn't available couldn't tell us anything more and we was supposed to be running the supply chain but not only that not just making sure stock was on the shelf probably but it was in used properly that staff had been trained in how to use these medicines that patients were receiving what you were supposed to receive for it and generally that the usage of medicines match the epidemiology in the country none of which we had any data about we didn't even have a list of facilities in Solomon Islands at the time we do now so that was the idea to break down those data silos and create a single source for that sort of information but also to do as a regional platform and that appeals to multilateral donors like D Fed and UNICEF and whro in these groups across things like disaster response disease outbreak tracking obviously health supply chains and other things like medicines quality or the other one is service provision the quality of service provision the facility is supposed to be providing X to what level are they providing that that service and we can provide that in tailored way so it can it can detect who is logged in and present them a different dashboard to to someone else logging in and it is now free and open source it's always been free but actually when we applied to speak at this conference it wasn't open source so I kind of um I did explain that to the conference organizers but since then actually just a couple of weeks ago we've released the platform open source the data is not openly available because this is data that we collect with the permission of governments and we have data sharing agreements with them that are quite specific about what we're allowed to share and what we're not we we do share publicly the location of facilities there the services that they provide and the staff that work there but we have you know like over half a million data points on all sorts of things I'm about these facilities and a lot of that information is obviously quite sensitive and governments don't want us to just publicly put it out there but we try to we try to expose as much as we can otherwise everything is kind of password-protected at various tiers of
access this is our current state it's in six countries at the moment some of which you may have heard of some of which not curable Solomon Islands Vanuatu Tonga Cook Islands and Tokelau it's a map approximately 900 health facilities so far we've had 14,000 surveys submitted through the app but as I say we also collect data from from other sources and primarily those sources are M supply M supply Mobile and D H is - for those of you who work in health systems we've won awards I suspect sometimes were the only people that apply to those awards I the this is
the URL if you really want to help us out today go and visit it and have a look diners love sight hits so so get your mobiles out and and go to to fight org but um but if you want to know more about the project obviously info doctor PI dot org sits behind that and kind of describes a little bit more about the
project this slide is probably a remnant for people who are really interested in health supply chains this is one of our key sources of data and we partnered with M supply which is a New Zealand company from the start to do two pi m supply is used in about 30 countries around the world so it's it's a lot bigger than 2 pi er they have just actually converted from being a private company to a not-for-profit they've now just started the M supply foundation and moved all their IP across into that amazing bunch of people over in New Zealand but probably don't need to talk to too much
this is our basic architecture I'll talk to kind of like the bird's eye view on this please don't ask me any specific questions that aren't literally on this slide because I get a little lost in the details that's it behind that but this is how its set up so we have multiple and various sources feeding data into 2 pi ER and the interesting part of that I guess is what is to fire so we use an aggregation server to kind of pull all that data together which is a di is to instance thi Stu is is a big open-source HIAS project run by the University of Norway I think in Oslo and so we use that to kind of pull everything together we then run it through a config server just to kind of do some of the analysis that I'm d-h is to can't and then the front-end Aziru is a react front-end we use matte box and open Strait map and all that stuff is on through leaflet but that's kind of the boring part the more interesting part is what do we what do we mean by like this what how does the data get used what do we mean by data visualization this is an example of some supply chain data that we pulled out of Solomon Islands this is just a just dump out of their arm stock management software and all this data is largely correct that's interesting someone who's running that supply chain would be interested in the minutiae of this information and may well like to work with the spreadsheets but mostly not and that is just incomprehensible to most people this is the availability of various medicines across you know 300 facilities this is this is one way we visualize it so we would sit down with the health supply chain manager I don't know this is a laser pointer but um but and we and this is a heat map of medicines availability and Solomon Islands so the top province up there is a province called Choi's or they're performing really really well lots of green everything looks good geographically the province of Isabel just here is very very similar in fact it's closer to the capital Honiara which is just here so you would expect it to have better medicines availability but because we can visualize the data really nicely we can pick out a little bottleneck there and that's something that we can work with the provincial pharmacy officer on in targeting that area they've got very very limited resources but they do have some resources to go and do touring provide extra training maybe there's a transport problem and this allows us to kind of get to the heart of that problem through visualization this is working fridges so with this we would go to the national cold chain manager who's putting together their annual plan of of which facilities will they get to they're never going to get to every facility in a year so we could work with them and say up in we'll say in Western Province here around gayssot they a little cluster a really really efficient trip for one of your maintenance teams would be to go and address that cluster there of fridges that aren't working rather than spreading yourself really thinly say over mulato province which would be an inefficient trip not to say that they're necessarily going to set all their plans you know from a map but it does allow you to sit down with them and make sure they're making kind of good decisions this is what I was talking about the quality of service provision so this actually comes out of Tonga and this data is probably an example of more sensitive data so please don't take photos of this data we do have permission to show these slides at presentations but not sort of permission to share this data more widely so this is services that facilities in Tonga are supposed to be providing and this is the level to which they're providing it now they use a traffic light system which is really simple to comprehend and we really like it now you can see there's obviously there's a lot of our engines and Red's but behind each one of them sits an action plan and in each of those action plans that describes or what do they have to do to reach a certain you know will to flip to grain basically is it new staff is it staff training infrastructure equipment or whatever else so we can then help them to visualize that and this chart here summarizes that for all of Tonga and it takes all of those action plans and aggregates them into a single chart that says what are the action plans saying about flipping all these services to green and at the moment the bang for the buck in Tonga is staff training it's not in recruiting new staff it's not an infrastructure the biggest bang for the buck is in staff training so we can sit down with the Ministry of Health in fact in this case the World Bank sits down with the Ministry of Health and helps them do their long-term spending plans based around these it's not to say that you don't invest in infrastructure it's just to say the the weight of spending over the next three to five years should be on staff training that's a super useful way for them to target their resources you know to tangible kind of outcomes this shows the availability of inpatient beds across facilities in a couple of provinces in Solomon Islands this is useful more for disaster response so if for example you had a tsunami down on the southern coast of Malaya this would help an incoming Disaster Response Team a medical response team usually coming out of Australia and in the case of the Pacific to identify where they can route mass casualties - we've also got a new functionality coming that that so we have a Disaster Response portal and we're kind of improving it one will kind of show if a facility has been lost what is the impact of that and do they need to set up for example a field hospital this was first deployed after cyclone Gaeta in Tonga earlier this year and it actually helped Australian response teams who were literally loading planes in Darwin to stand down because the data showed actually the impact on facilities was not as great as was feared and that managed to save Australia a lot of money get us some a lot of a lot of nice attention at the time so that was um so that was that was a really kind of practical example of data use and then there's all sorts of other things but I won't go through them necessarily with you we have a social health fee this is a way to encourage data collection we've tried to gain a fire our data collection out we call it Facebook for health care workers so it's got an automatic news feed when you can submit a survey completed so that appears in the feed and then we can put content into that feed and we can also push notifications to people through the app to say for example there's a diarrheal outbreak near you these are the steps you should take to UM to act on it and my favourite part of the app you can earn coconuts and pigs for collecting data and there's a leaderboard of who's got the most um who's got the most coconuts the idea being to kind of create a community around data there's a big problem with health sectors in that we're constantly asking primary health care workers to collect data which they never see again it just routes up and it never comes back down so the idea was to try and build a community around it so they kind of feel part of it like all these things - piracy is a partnership between the whole loop of agencies but mostly funded by de fat I'm probably probably 90% funded by owned by de fat through the innovation exchange and that is that I think that leaves is that right on timing I'm about yeah a couple minutes for questions Thanks yeah we tried to UM we try to
partner with them well with the whole range of organizations as much as possible especially were actually super useful let's say you've got a whole arm called data for development I think it's called is that right yeah yeah legendary guys there's a guy they feel bright who I don't think we would have got this up with that some of his advice so we've shared all our all the data that we can share we've given to them and so for example facility locations is now all data that they post on their Maps for those of you who haven't seen it I actually sent Phil an email saying he should be here because SPC's maps of the Pacific services health education but census data in particular they're just unbelievably rich data sources yeah they're really young they're really good but um but increasingly our engagement with SPC's I guess more around acting on the data rather than don't so much do active stuff with Phil at the moment although they're about to um especially is about to release their population like their census data through like publicly facing API sites which has been kind of in the works for a while at the moment our population data is pretty weak it's basically based on going to facilities and asking what is your catchment population we'd like to do that more dynamically by by just pulling in that from the census data from SPC so fingers crossed it was supposed to be by the end of the year but I don't know where there are two with it