You are browsing the archive for Salua Nassabay.

Making Data Relevant: Irish Health Data

- June 26, 2015 in #OpenDataIRL, Brian Costello, Caroline Lewis, Dan Alexandru Bujoreanu, Daniel "Chino" Carranza, Dave Corley, Denis Parfenov, Eoin MacCuirc, Eugene Eichelberger, Events, Flora Fleischer, Hospital Waiting List, Ingo Keck, Maker Party, Margaret Furr, Martin Kelly, Open Knowledge Ireland, Pamela Duncan, Richard Geoghegan, Robert Harte, Roslyn Fuller, Ruta Danyte, Salua Nassabay, Shawn Day, Steve White

Call for Action

Data collected on behalf of the people of Ireland and paid for by taxes should be available for use, reuse and redistribution as a right and under Service Level Agreement (SLA) in 21st century non-proprietary, machine-readable formats. PDF is not open data. Publishing reports in PDF format makes them inaccessible for processing and in effect renders the data unusable. Open Knowledge Ireland and OpenStreetMaps Ireland  call on Brendan Howlin, Minister of Department for Public Expenditure and Reform, Leo Varadkar, Health Minister and Richard Corbridge, Chief Information Officer for the Health Service Executive in Ireland, to support the efforts of the Open Data Community to increase the usefulness of publicly available Health Sector Data, by ensuring its publication in an open data format.  

Maker Party Round-Up

Screen Shot 2015-06-26 at 10.23.35 On Saturday, 13 June 2015, a diverse collection of twenty publicly-minded enthusiasts joined to explore and demonstrate the benefits of public data made open and used. The day began with opening remarks by Denis Parfenov:
We want the government to make data collected on behalf of citizens and at taxpayers expense available for use, re-use and distribution in useful 21st century non-proprietary, machine-readable formats; so it can be converted into actionable information to help society to answer real questions.

A lively and frank discussion led by Brian Costello and Eoin McCuirc from the Central Statistics Office followed – primarily detailing concrete ways to make data requests easy, fast and open. There was agreement among the wider group that making requests ‘public’ will help to make requests traceable and transparent and serve the public interest. Participants then broke into 3 focussed working groups:
  1. The OpenStreetMaps (OSM) group followed detailed instructions (link) by the OSM Community Organiser in Ireland Dave Corley and in a determined effort tidied up geospatial data for the 41 hospitals listed in the NTPF acute care dataset.
  • The intention was to use a publicly accessible, open format platform to provide a geospatial foundation for the Hospital Data Working group – but also make the same data available for anyone who cared to make use of it;
  • The result is a clean and accurate list of hospital lat/long coordinates generated by Dave Corley and available (link).
  1. The Data Wrangling group manipulated the available CSV data (which has been manually scraped on a monthly basis) on hospital waiting lists accessible to everyone:
  • The objective of this group was to transform inaccessible hospital data, published in PDF reports by the National Treatment Purchase Fund (NTPF http://www.ntpf.ie/home/inpatient.htm / http://www.ntpf.ie/home/outpatient.htm), into machine readable data formats (link to data / link to project page)
  • This is a requirement to be able to use the data to provide quality analysis into how long patients are waiting for hospital appointments for May 2015 (link to data);
  • Participants self-organised themselves into an ongoing ‘Hospital Data Working Group’ to work on strategies of making hospital data accessible to everyone via an interactive data publishing platform;
  • This group identified that a major obstacle to persons identifying the hospital at which they may get the fastest specialists appointments or where they would wait the longest is the authorities is not making this data available to the public in machine readable formats.
  1. The Social Media group discussed ways of making the activities & benefits of open data known to the wider public. Their primary recommendations are:
  • Open health data (or any sector-specific data) is potentially of interest for everyone in Ireland, but people tend to think about it only when they are personally affected;
  • A media campaign would need to create general attention first, eventually leading to seeking of more detailed information by individuals with specific interest;
  • Information needs to be provided in small, relevant and instantly recognisable pieces for it to lead to more in-depth information requests.

Inspiration from Uruguay

At the end of the productive day we established a virtual bridge with Daniel “Chino” Carranza (@danielcarranza / @DataUY) in Uruguay. Daniel shared http://datauy.org/’s inspirational story of co-creating a data driven Health Care Dashboard (http://atuservicio.uy/) which helps people of Uruguay to make an informed choice of health care provider base on data, not marketing. The Ministry of Health of Uruguay published data in Excel spreadsheets over the past 4 years. However, the number of downloads was a meagre 500 cases. By making this same data accessible in an easily comprehensible and actionable format via http://atuservicio.uy/ increased data exposure by over 7,000%! Taking the time to help the public understand the context of the data makes it active data.
With the information was published through the dashboard, the government of Uruguay started a quality of healthcare discussion, and for the first time based on data, not on opinion and marketing.   For more information: Health Data Maker Party on Storify: link Photos on Flickr: link Opening remarks: link Daniel “Chino” Carranza’s slide-deck (link); video (link); full (rough) transcript of the call (link) Hospital Waiting List project page (link) Acknowledgements: Many thanks to everyone who participated in this workshop in person and virtually: Margaret Furr, Richard Geoghegan, Martin Kelly, Ruta Danyte, Robert Harte, Pamela Duncan, Salua Nassabay, Roslyn Fuller, Flora Fleischer, Dave Corley, Shawn Day, Daniel “Chino” Carranza, Dan Alexandru Bujoreanu, Eugene Eichelberger, Caroline Lewis, Ingo Keck, Brian Costello, Eoin MacCuirc, Steve White  and Denis Parfenov Special thanks to newly openned TechMeetup.space for hosting the event and to the Department of Public Expenditure and Reform (http://www.per.gov.ie/) for sponsoring the venue and providing tasty sandwiches and healthy refreshments.  

First Irish Charity Data HackDay took place on Saturday July 26th 2014 at Tcube

- September 4, 2014 in Adrian Corcoran, Adrian O’Flynn, Allen Thomas Varghese, charity, Chris Garde, Data, Dave Corley, Denis Parfenov, Flora Fleischer, Grazia D'Aversa, Hackday, Helen Nic Giolla Rua, Ingo Keck, Ireland, Open Knowledge Ireland, Patrick Killalea, Salua Nassabay, Tracey P. Lauriault

Screen Shot 2014-09-04 at 13.27.58
When: Sat, 26th of July, 2014
Where: TCube kindly hosted the event to support the work done for the Charity sector
Organiser: Open Knowledge Ireland
Who: 13 participants: Dave Corley, Tracey P. Lauriault, Patrick Killalea, Adrian Corcoran, Allen Thomas Varghese, Ingo Keck, Helen Nic Giolla Rua, Adrian O’Flynn, Flora Fleischer, Denis Parfenov, Chris Garde, Grazia D’Aversa, Salua Nassabay
 
Main Findings from first Irish Charity Data HackDay
  1. No standardised way of tracking income and expenses
  2. Not all charities make income and expense data publicly available on their website
  3. When published, income and expense data is not published in an ‘open’ format
 
… and here is how it all went down on Saturday 26th July 2014 at Tcube:
The day started with a welcome speech by our conveners and hosts Denis Parfenov and Flora Fleisher of Open Knowledge Ireland.  A short presentation was given by Flora Fleischer. Adrian O’Flynn, our Charities Subject Matter Specialist, and the person who inspired the event, introduced the topic of spending for charities in Ireland. He highlighted recent issues portrayed in the media and explained why it would be useful for the public to be able to compare charities based on their financial reports.  Here is his presentation. Adrian Corcoran, the event’s project manager, provided detailed instructions on how to work together for the day.  He followed this outline (http://openknowledge.ie/chy-01-charity/)  
Three main datasets were used for the CharityHack:
  1. An overview working document listing basic data about the charities: (https://docs.google.com/spreadsheets/d/1sIH9NKBkpQMFMnt_0sYW9B8DSyv839EbbtLmibdml1s/pubhtml).
  2. An excel spreadsheet (CharityFinancialDataFinal.xlsx), that includes detailed data for a number of charities, which were initially extracted from the annual reports by Adrian O’Flynn.
  3. A reference document which includes definitions of the codes used in the document above (e.g. expenditure codes)
    (https://docs.google.com/spreadsheets/d/1sIH9NKBkpQMFMnt_0sYW9B8DSyv839EbbtLmibdml1s/edit#gid=861039018)
  Participants were divided into three teams “red”, “green”, “blue”, following the team label of the charities in the overview document.   Participants from each team picked a charity from the overview document.  Adrian Corcoran told us how to label the data, how to quality check the numbers within and how to verify the data with the annual reports. These data were then assembled into one shared document.  Data quality issues were reported and then corrected by Adrian Corcoran.   Any issues were reported during this process and the more serious problems were shelved for later. Once this work was completed, only data that passed data quality processes were included.  
The charity data used for this CharityHack exercise are as follows:
  1. Data were extracted from the annual financial reports of 24 of the largest charities in Ireland (€1Mil+).  NOTE This represents a small sample of the thousands of charities in Ireland.
  2. These are charities known through their public brand awareness for the fiscal year 2012 (see the definition).
  3. ONLY charities where funding represents charitable funds collected voluntarily from the public were selected.  This excludes charities that primarily rely on large institutional funds (e.g., Irish Aid Grants, HSE Grants).
  4. Only charities that have been independently audited and who have published their financial statements on a publicly accessible websites were used.
  5. Only charities who followed the Statement by the Accounting Standards Board on the SORP Accounting and Reporting by Charities: Statement of Recommended Practice were used (Section B: Resources Expended). Not all organizations in Ireland follow this standard. It is only because of this standard that it was possible to derive standardized financial data from the financial statements in the annual reports. The data here reflect resources spent on Governance, Fundraising and Charity.
  6. The list of charities was derived from the survey conducted by the Irish Charity Engagement Monitor (ICEM).
  7. This final dataset was quality checked by the participants at the 2014 Charity Hackday on July 26.
  8. The original dataset was created by Adrian O’Flynn.
  It is a small sample, but this illustrates the power of open data.   The dataset that was generated on the HackDay, and which is being continuously developed, can be found on the  Open Knowledge Ireland website. Finally a first draft of a future website was created, where charity data can be be displayed in an easy to understand fashion. Thanks to Barry Alistair (TCube), Adrian O’Flynn (Charity Subject Matter Expert), Adrian Corcoran (Project Manager), and Denis Parfenov & Flora Fleischer for organising the Charity HackDay. And many thanks to all the participants who lent their time and skills! We could not have made as much progress in one day without you!
Photo report of the day:
https://www.flickr.com/photos/123736148@N04/sets/72157645554728899/ Pictures are courtesy of Dan Alexandru who kindly joined us on the day to capture all the fantastic work happening! Thanks Dan!