The first Electronic Medical Record Systems

The first generation of EMRs were extensions of medical billing systems in large US hospitals, and over the last four decades, they have been used as tools to organize and store medical data. EMRs are widely accepted as important tools to support high quality health care in the US, Europe and other developed countries. Evidence shows that using EMRs that include decision support systems improves quality of care and both reduces medical errors and unnecessary medical investigations.

Experience with the use of EMRs in developing countries is much more limited than it is in the US and Europe, but there is now considerable interest in using medical information systems to support the treatment of HIV and TB in Africa, Latin America, and Asia. Some examples of use in the developing world include:

• The Regenstrief Institute in collaboration with Moi University in Kenya developed an EMR for general patient visits to clinics in western Kenya. This system was subsequently modified to support the care of several thousand HIV patients.
  
  
• Baobab Health Partnership in Malawi has developed an EMR system using innovative, low-power touch-screen PCs for data entry and display. This system is now used to support the care of more than 7,000 HIV patients in the Lighthouse clinic in Lilongwe and has been chosen by the national HIV program for use throughout the country.
  
  
• Careware®, an HIV medical information system developed for US patients, has now been deployed in Uganda and is planned for use in other African countries and in Latin America.

EMR systems used at PIH sites

As of March 2008 nearly all of PIH's sister organizations in the countries mentioned below are using either the PIH-EMR 1.0 or PIH-EMR 2.0, a customized version of the medical record system framework, OpenMRS.

PIH-EMR (Peru)

In 2001, Partners In Health developed the PIH-EMR, a web-based medical record system to support the treatment of patients with MDR-TB in Peru. This system is built with an Oracle database and uses the Apache web server, Java server pages, and the Tomcat servlet engine. It now has records on over 25,000 patients, almost 7,500 of whom have started or completed treatment for MDR-TB. The system is designed to collect baseline data on patient assessment, bacteriology test results, drug regimens and patient outcomes.

It now includes:

• Extensive tools for data analysis
• Drug forecasting and communication tools to improve the quality of laboratory data and reduce treatment delays

In addition, we have also created a system using Personal Digital Assistants (PDAs) to enter information into the PIH-EMR. PDAs are used to track over 3,500 monthly laboratory results from more than 120 health establishments throughout Peru.

The data collection team travels to the health center or laboratory and enters information into the PDA. They then synchronize the PDAs with the computers in the home office, transferring the data into the PIH-EMR. This method of data collection has reduced delays in data entry and has increased efficiency in data entry by 40 percent and reduced data entry errors.

Use of the PIH-EMR in Peru has been evaluated in several studies showing it reduces both the workload for those responsible for data entry, documentation and reporting, and delays in MDR-TB diagnosis and treatment.

e-Chasqui (Peru)

In order to provide direct web access to critical laboratory data in Peru, the PIH-EMR now includes a component that permits web-based entry of tuberculosis laboratory results at all regional and central laboratories performing these tests. The component, termed “e-Chasqui,” includes applications that assess quality control, generate aggregate reports, notify health centers of new results or contaminated samples by e-mail, track enrolled patients, and follow the status of pending laboratory tests.

This system was initially deployed in the national TB laboratory, two regional laboratories and twelve health centers. Because of high user satisfaction and heavy use, e-Chasqui is now being used in two district laboratories and 34 health centers. Staff at an additional 25 health centers will be trained on e-Chasqui in May 2008. To date more than 65,000 laboratory samples have been entered and tracked in the system. e-Chasqui will eventually serve a network of institutions providing medical care for over 3.1 million people. An impact evaluation is currently being performed on the system.

e-Chasqui benefits include:

• Connection of laboratories and health centers by email
• Constant access to laboratory information for health center staff
• Tools for data quality improvement
• Reporting functions for laboratory personnel
• Clinician alerts for high risk patients

HIV-EMR (Haiti)

In 2002 the PIH-EMR was modified to support the treatment of HIV in rural Haiti. The HIV-EMR is now used in nine clinics in rural Haiti and collects data on patients’ clinical histories and examinations, drug regimens, laboratory tests (including CD4 counts), follow-up data including opportunistic infections, and medication side effects. Data for over 10,800 HIV patients are now stored in the EMR. We have recently added a patient tracking system that will reveal whether patients are missing follow-up visits or if their medication has not been picked up. The HIV-EMR also includes a web-based pharmacy system that tracks all medications and supplies for the clinics, including non-HIV drugs.

The HIV-EMR is accessed by satellite internet in each of the clinics where it is used. It both generates reports on patient treatment and detects patients whose treatment may be delayed or incorrect. Given Haiti's erratic power supply, the HIV-EMR allows staff to enter data while disconnected from the internet. It is then synchronized when the internet is re-connected.

OpenMRS and PIH-EMR 2.0 (Rwanda, Lesotho and Malawi)

By 2005, PIH wanted to refine its original EMR system into a new open source version. The technical components of PIH-EMR systems – the data collection forms, patient summaries, general purpose reporting framework and web-based design – were crucial pieces of a system that had to be developed to widen its replication potential. These tools are needed at all PIH sites and those of our collaborators, and have been requested by many other organizations. By developing them using open standards and open source software, even organizations with significantly different requirements would be able to modify them efficiently for their use.

Just as we believe that the benefits of healthcare should be delivered to those who need them most, so should the benefits of technology.

Significant resources are being invested into large-scale treatment programs for HIV and tuberculosis, the world’s leading infectious disease killers in developing countries. It is critical that these programs are managed well and can demonstrate successful patient outcomes and sustained health systems improvement. Establishing a uniform architecture will greatly facilitate the coordination of large-scale HIV and TB treatment programs. PIH saw a huge opportunity to refine its EMR as a new, open-source version, one that is flexible, scalable and easily adaptable by others in the field.

The original OpenMRS system was created as a collaboration between PIH and the Regenstrief Institute; a community of developers and implementers in more than half a dozen African countries are now contributing to its constructionl.

Special features of the PIH-EMR/OpenMRS

• Open source software which enables the system to be as widely accessible as possible by sites with limited funding.
• Open standards for data exchange such as HL7, allowing exchange of patient records with other medical records systems, including use of a WHO-supported HIV record standard.
• Synchronization tools to link separate OpenMRS installations. This is useful for sites that do not have internet access or have very slow connections. They can run their own copy of OpenMRS, enter and view data on their local networked server and get a fast, efficient response from the system. In the background OpenMRS sends all the data updates to the main server at a central site. The central server does the monitoring and reporting for all sites.

• The concept dictionary which allows users to add new concepts to customize the system for their individual needs.
  

  The Concept Dictionary allows users to enter new concepts which help to develop data collection forms and store data.

• Modular architecture that allows a programmer to further customize the system. New modules can be designed
  and plugged into the existing system without disturbing the core foundations.

• Use of the Java platform and other standards make this system viable for the foreseeable future.

• Fully customizable. The system can be modified to perform almost any task.
  
  
• Runs on Linux, Windows, or Mac OS and the entire system can be run on a large server or on a laptop computer.
  
  
• Includes extensive built-in reporting tools.
  

To learn more about the OpenMRS, please visit the web site and try the demo; we will soon be posting a specific demo of the PIH-EMR/OpenMRS. While we are pleased to share the software which does not require programming, it does require an experienced IT professional to install and configure it and back up the data, as well as staff for data entry and quality control. We encourage you to use the software but also to be prepared to support the needed investments in infrastructure and personnel described later in this section to develop and maintain it.

In August 2006, the Rwanda site began use of OpenMRS; in November 2006 we started rolling out OpenMRS at PIH Lesotho, and in late 2007, PIH Malawi began using it.

In addition to the features mentioned above, the PIH-EMR 2.0 contains the following functionalities:

• A patient register, programmed to give reports that fulfill internal and national reporting requirements
• A clinical tool that prints patient data summaries on a daily basis
• A database for operational research
• A drug prediction tool to assist in procuring medications