Peru has the highest incidence of tuberculosis (TB) in South America and the eighth highest burden of multidrug-resistant TB (MDR-TB) in the world . Despite demonstrated success in the treatment of MDR-TB with an outpatient approach , it remains a problem in metropolitan Lima. Important questions remain about containing further spread of drug-resistant TB in high burden settings of this sort. In 2006, we were funded by the United States National Institutes of Health (NIH) to implement a study of tuberculosis epidemiology in Peru to determine mechanisms of MDR-TB transmission (Grant numbers: U01 AI057786, U19 AI076217). Implementing this study required a secure information system that would collect data from a target goal of 16,000 subjects, each followed for at least one year.
Traditionally, clinical research projects have built custom information systems for each study. Depending on the nature of the study, this approach can be expensive and time consuming, as each system needs to be meticulously tested and documented. In 2008, Lang et al. estimated that proprietary commercial systems for research data management can cost hundreds of thousands of dollars for a medium sized study, cost prohibitive in developing countries . Projects that include longitudinal follow-up and multiple visits, forms, and laboratory results are particularly complex and expensive to support. Creating an effective information system to manage such data requires careful database design, and often necessitates significant adaptations to both design and software as new data items and workflows become apparent once the study has started.
Across study types there are common functional requirements for data collection systems; these include patient-based data collection, cohort analysis, scheduling of appointments or tests, patient identification using bar codes, linking to laboratory information systems, and ensuring security and confidentiality. Software tools created for these requirements can potentially be reused in new studies, reducing time, costs, and quality control issues. Recently, a number of clinical data management tools have become available in developed countries to standardize the management of clinical research data collection [4, 5]. Research studies in developing countries face additional challenges in infrastructure, staffing and expertise, and are usually dependent on outside technical expertise and/or expensive software licenses for data collection systems. A similar situation exists with medical software systems in developing countries such as Electronic Medical Record (EMR), laboratory, and supply chain management systems .
When planning began for this study, we had extensive experience in developing information systems to support the clinical care of MDR-TB in Peru, including management of laboratory data . We were keen to use a standard platform to develop the research information management system. In this paper we describe the customization and use of an open source EMR system platform called “OpenMRS” to support this large epidemiological study. We also describe the steps required to adapt and tailor the OpenMRS framework for this study, and our experience using the system for more than two years.