Overview of EMR Options and Roadmap

As we note on the ICT Home page, “the installation of a Hospital Management System is a large project that has yet to get off the ground. There are multiple approaches to take, depending on how ambitious we want to be, feature-wise. Short-term needs would also be addressed by a very different system than Long-term ones.”

The shortest-term goal is simply to set up networked file sharing for the hospital's offices so that admin, billing, and records can be centralized. Currently, much of this data is being passed around on USB sticks. These are usually Office documents (Excel files for records, Word for reports) and Quickbooks files for the accounting. Other needs around this:

• Scripts to convert the excel spreadsheets into Quickbooks transactions.
• Automatic backup and possibly RAID.

In the meantime, the most important thing to collect are copies of all the hospital's forms (for the input side of things) and the reports that are generated (for the output side of things.) This is the first step we'll need to take to start modeling the operations of the hospital to try to implement a system for hospital management.

Eric did a good amount of research and testing of Hospital Management Systems and concluded that his best shot was to try to get an installation of OpenMRS running for the HIV Pediatrics Ward, which is on the Jubilee center side of the campus. The EMRSystemsResearch page has the current set of notes on all of the other systems considered, and the tradeoffs involved. In a nutshell, we're choosing a software package that currently used primarily in developing world HIV clinics. It thus has the right feature set for the HIV pediatrics ward: highly structured data input for accurate reporting (needed for the national program, Global Strategies, the Gates Foundation, and other funders.)

At the moment, however, OpenMRS is less suitable for the entire hospital's medical records, as it lacks features such as appointment scheduling and billing. Those features are being discussed actively on the OpenMRS boards; if they are developed in the near future, we can expand OpenMRS installation to the entire hospital; otherwise, a different software package may be appropriate.

Eric completed a basic installation, including the data model for one of the intake forms. This will be passed along to Steve Conrad and Erik Jerde for their trip to Goma, but OpenMRS may be out of scope for their visit, as well.

Assorted notes:

• We have buy-in at the pediatrics ward through Dr. Vindu, who is great. She will be leaving in a year for further medical training, however, so we will need to make sure that the system is solid before she goes.
• Hardware: We will eventually need a server, preferably dual-core, plenty of hard drive space, and some sort of backup drive. In the meantime, we will be able to get away with using a moderately-powered laptop.
• An OpenMRS installation will also require one or two additional paid data entry clerks to fill in old data from old records and keep up with new patients coming in. Checks will be needed to ensure data integrity.

• Input forms:
  • form_-_ordonnance_medicale.pdf
  • form_-_vnpa-pvs.pdf
  • form_-_vnpa-avs.pdf
  • form_-_fiche_du_malade.pdf
  • form_-_fiche_des_taches_du_la_visite.pdf
• Reporting:
  • scans not yet obtained
  • Generally, this is of aggregated data, by various breakdowns.
  • Some specifics:
    • Entry of longitudinal data (see notebook)
    • Nice to have, but potentially difficult with OpenMRS: entry of ordinal data (patient # in given month)

A full-fledged hospital management system is a very large project, most likely requiring a man-year to complete. Hospital-wide, these are the kinds of things that seem to be unsupported by OpenMRS, at the moment:

• Scheduling and calendaring
• Billing for patients
• Some way to integrate information systems in the rural clinics with those at HEAL Africa, to better maintain lines of communication as well as tracking patients and outpatient statistics (may be supported by OpenMRS)
• HL7 protocols to prevent future migration problems in terms of communicating with other hospitals (may be supported by OpenMRS)

We have no conclusions, yet, on which system will provide for HEAL Africa's needs. Research and notes on options is kept in EMRSystemsResearch.