Implementers Lessons Learned

Evan Waters – PIH Malawi

• Challenges – can’t really upgrade because application  running the database, not integration level

  • Application development is on core features, we need customized things that work well in hospital setting, data collection up front and not having a place to do something with the data (M&E, reporting, point-of-care w/ interactivity of data)

• Solutions – get confidence in the system and then build

• Success – we have a lot of data, but data is not as reliable (soln – don’t do everything that is possible, but do things to a minimum to get it done)

Gilbert – AMPATH

• It is scalable and robust enough to make information

• Challenges – viruses, networks breaking down, no major issues with the system, lack of basic knowledge for daily running and maintenance of the system

• Solutions – need IT support that is local, people need to know how to install and get services running, need knowledge in customization to fit forms to the situation that way it is easily adapted for use

PIH Rwanda

• Using OMRS since 2006

• Successes – OMRS is not only based on data entry people, clinicians are using it and are being trained so they can look on the machines in the consultation rooms, having patient lookup and patient summary; instead of only producing reports, they can go under their own username and have patient summary and looking at clinical history

• Challenges – sometimes you can’t find the electricity in rural areas, collaboration

• Solutions – PIH implemented generators and solar power

Andy Kanter – MVP

• Primary care, rural implementation

• Challenges – electricity, connectivity and bringing data together (have not explored Sync yet), need to provide feedback immediately to users (not yet solved how to in primary care setting to provide value to ppl immediately coming in for chronic and pediatric care), integration with the moving target of MoH reporting and reducing redundancy (soln – possible automation), HR capacity (networking, maintaining computer)  (soln – deep freeze to lock computer and take care of viruses), train providers and people who  maintain the servers (led to more centralized approach such as forms for central place for data entry), interaction with the community (rapid SMS – identifying pregnant women in the community), access to data

• Soultion – solar power

• Success – pivot table to access data on systems such as Microsoft Excel

John – AMPATH

• Challenges – connectivity (depend on synching once/wk which gives problems for MoH reporting), capacity building (clinics to have ppl generate reports on their own)

Septarshi – HISP India

• Challenges – train people and have them use the applications , trying to understand what real-use case is in district hosptials

• Solutions – build new modules – billing, pharmacy, participatory design mockups (implement it for a week and collect information on experiences….slow process but more fruitful)

• Success – good community around it and will help the system build up quickly, to get good feedback you need to deploy something and make it sustainable for people to use (sustainable = 1mth or so)

Roger – CDC – Panama

• Challenges – geting used to the entire stack of applications and breadth of OMRS (steep learning curve), this was a national system (no ramp-up time), inability to have easy way to do iterations, hard to have multiple ppl working on the same things, expect more of customization process than development process (not quite the experience we had when we went into it)

• Solutions – find a network or central model/design process

Hamish – PIH

• Solutions – start with the goals and what you want from it

• Challenge – number of people you need to do data entry

  • Solution –

Implementation Advice

• Provide right data at the right time

• Clean up the data that you have before you corporate data for other programs

• Collect data that you are user someone will use and is really needed

• Paper based forms are not necessarily the best template for html form

• Understand the system environment

• System manager function is important – have right ppl trained and capable

• Make sure your data collection tools meet the usage requirements

• If it’s a large network, have good data manager

• Identify customer, know their needs, respond to needs

• Need good support system

• Don’t build concepts from scratch, use what exists in concept server

• Provide early and often feedback on data back to ppl doing data collection

• Involve the clinicians and have them use OMRS more often to get more information that you need

• Have good training for those who will use OMRS to have better quality data

• Don’t try to replace your whole paper-based system with the system, think about raw essentials you need and scale from there (patient registration, lab orders, pharmacy, obs)

• Don’t operate it under faulty IT issues, have in place a system that gives feedback so you know data is useful enough and change it if you need to improve it

• Don’t listen to what people are using the system tell you – have them show you

Steven – WHO

• Suggests for people to share, blog experiences because many have the same situation

Andy Kanter

• Questions (unanswered)

  • Virus issues

    • Connectivity is the root – if a computer isn’t always connected, it doesn’t receive updates and then you have multiple flashdrives being used

Frank Tilugulilwa - University Computing Centre Ltd - Tanzanoa

• Support issue to clinicians in remote areas

• Possibilities of overhauling these sites to Linux servers instead of Windows for stability e.g virus, clashes etc

Andrew

• Getting qualified people to work in remote areas

  • Roger – some countries have 1yr of service where you can have IT students come work for you and they receive stipend from the government

  • Steve

    • ICDL – International Computer Driver License

      • Interested in health tracks and scaled down versions in other settings

      • IT Infrastructure Library (ITIL)? - a global framework for IT Best Practices