OpenMRS in Hospital Settings

Facilitator:  Wisdom Atiwoto

10 Sept 2010

Definition of hospital

  • Anything with an inpatient facility (definition in Africa)

  • Secondary hospitals – require referrals

  • Tertiary hospitals – require referrals

  • Small office outpatient à health center (can receive emergencies) à district hospital (30-40 beds, OB/GYN services, Peds, Surg, Int. Med) à regional hospital  àteaching hospitals

  • What are others doing in hospitals?

    • AMPATH – District hospital

      • Used for clinical care

      • MTRH interested in implementing OMRS

      • Hospital now using patient registration

      • Goal is for teaching hospital to use it for clinical care, put terminals around the hospital for clinicians to look up information while on rounds

      • Many systems already in place but they are using OMRS to do patient registration, unification of patient identification across departments

      • Motivation of MTRH to start use OpenMRS – 250K patients in AMPATH clinic, (right next door to MTRH), want to integrate systems with AMPATH

      • Trying to push for using REAL data to meet the needs of local people so that there is feedback to the providers to improve the quality of data entered into the system

  • Ghana

    • Driving force is health insurance claims processing and other payment mechanisms

    • Questions usually asked: Can it help us with health insurance claims? Can it help us send reports to the district and regional health administration?

    • Billing needs are urgent

    •  

  • Mozambique – in 5 districts

    • Limited connectivity

    • In the main hospitals, not the rural hospitals

      • About 10-15 beds – Infirmary

      • 5-10 beds – maternity

      • Patients do not move between wards

    • Implemented to do HIV treatment

  • Nigeria

    • One general hospital – inputting all the registers into OpenMRS

    • Using 2 computer – record keepers entering data into the registers

    • 1 computer in delivery room – midwives writing info on labor cards first and then entering it into the computer

      • If they interact with the computer, midwives can see antenatal visits

    • Planning on expanding to 2 more hospitals (~15 beds and ~400 beds) – one teaching hospital

      • Treat implementation of the system as if each were an individual clinic (some get point of care some don’t)

      • Needs inpatient care module

      • Patients move between wards while in the hospital

      • Entire hospital needs to wired and networked, multiple wards may need to see the data at the same time

      • Steps to implementing OMRS in hospitals

        • Why did Ghana hospital choose an EMR?

        • Sync from the largest hospital down instead of the smallest up

        • Hospital is a large community that needs internal communication

          • Liked log in page with hospital events and information

          • Patient tracking feature

          • Staff scheduling

          • Resource scheduling

          • Ancillary services

  • What’s the minimum that is needed in order to provide services to all of the departments

  • Some labs get paid by the test and not the result so the lab won’t record the result

  • OpenMRS is not the answer to the other processes in the hospital

    • Need to do resource planning

  • Why openmrs?

    • District hospitals have 150-300 beds

    • Good for the patient, not looking for profitability

    • Private hospitals are looking for profitability

    • Implementing a hospital system that can help with resources and costs

    • A way to gather information about a group of people and associate that information to them. OpenMRS is not a good solution for supply chain management

  • What are the hospital’s needs? What are the motivations of the hospital for implementing EMR?

    • Example, OpenMRS can do minimal set of features for something like laboratory data and then when they outgrow OpenMRS, get a true lab system and then integrate it into OpenMRS

    • Using Rest Module and HL7 messages for integration

  • Compatibility discussion with other open source communities for ancillary services

  • Medical record is fundamentally the same between inpatient and outpatient

    • The types of data are different but we are still tracking medical information about patients

  • In the USA, there is a distinction between the medical record and the “hotel keeping” functionality (supply chain, billing, etc.)

  • OpenMRS is working to be able to integrate with other systems easily to help meet medical needs

  • Promotion of best practices

    • Hospital infection control – where does that fit in?

    • Lab safety – where does that fit in?

    • Incident reporting and investigation  - where does that fit in?

    • OpenMRS can export relevant data to an external system

  • For bigger hospitals – one would need “real” ADT, lab, pharmacy systems, are there other examples of modules that are needed?

    • Lab, billing, pharmacy, radiology, patient registration

  • Modules need to work for individual organizational workflows so they need to be generic

  • Expectation management

  • Resources to build modules for inpatient/hospital needs

    • Billing module – got all the hospital services as concepts with a parent concept that is set as a global property in the billing module. Then you can assign prices to each of the services in an external table. Make a set that is billable concepts. Billing clerk can choose the necessary service.

      • 2 kinds of orders sent from billing module – lab and pharmacy

      • Can identify 2 kinds of patients (above and below poverty line) – government discounts can be applied as well

      • Registration module determines income level (person attribute)

      • Customizable receipt

      • Not a detailed accounting system;

        • Data can be extracted

  • Working on financial module

  • Release date: 1 week (17 Sept 2010) to module repository

  • Are people anything working on? – need for documentation

  • Experiences in implementing OpenMRS in hospitals vs. clinics

    • Workflow complexities much higher in teaching hospitals over outpatient clinic

    • Possibly start with smaller hospitals before larger hospitals

    • Large failure rate in Nigeria at teaching hospital

      • Connectivity, hardware needs/costs, workflow complexity à failure

  • Evolve the implementation out from small wards (or small hospitals) and then let it expand

  • Possibly start with Surgery and OB/GYN and then expand from there

  • At AMPATH – start with registration

    • Workforce training

    • Had to change the overall process of maintaining record keeping throughout hospital à MAJOR work culture change

    • Lengthy discussions with record keepers prior to implementation

    • Sub-district hospitals – couldn’t have users test the system. Workers wanted pressure from the government which didn’t happen. No local leadership buy-in.

  • In Ghana – major workflow changes that may or may not be accepted by the workers à need to have input from the people who are going to be using the system

  • Need to get local leadership buy-in otherwise there will be failure

  • Blending inpatient and outpatient modules

  • Growth of OpenMRS in hospitals

  • Coordinate all departments in a hospital

  • Expectations for user interface in the medium term

    • Expect to see basic modules around ancillary services soon

    • Visits = grouping of encounters

    • Formal support for order entry

    • Medical Notes àcan store as observations, no templates exist yet. Can create handlers to create a new kind of template, not done yet.

      • Debate as to whether that should be OpenMRS core or not

      • Can a form be put together to do documentation? (limit for a field is 64K)

      • Looking to expand OpenMRS from clinics to hospitals

      • How to use OpenMRS in commercial hospital set ups

      • Challenges in implementing OpenMRS in hospitals

      • How to use in community mental health and social care in the community (challenges)

        • Observing mood and behavior daily, can incorporate concepts based on mood and behavior

        • Answers are very wordy, how is the data used? Is there information to be extracted?

        • Data that was collected was done in discreet ways (yes/no, scales)

        • Trying to develop a module for mental health care delivery in Ghana

          • Size of textual fields is limited (needs to be a narrative)

          • Trying to extract information from a long narrative is very difficult

          • How to reconcile between discreet questions (lengthy) vs. narrative (concise)

            • In form design, find the balance between discreet and narrative components

            • Why are we coding this? How are we re-using this? àthese questions help determine what should be discreet