NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Institute of Medicine (US). Addressing the Threat of Drug-Resistant Tuberculosis: A Realistic Assessment of the Challenge: Workshop Summary. Washington (DC): National Academies Press (US); 2009.

Cover of Addressing the Threat of Drug-Resistant Tuberculosis

Addressing the Threat of Drug-Resistant Tuberculosis: A Realistic Assessment of the Challenge: Workshop Summary.

Show details

5Infrastructure and Health Care Delivery Systems

The structures of both in-country health care delivery systems and international donor programs are not optimal for the effective delivery of TB treatment, especially for drug-resistant forms, in many high-burden countries. Problems include underdeveloped general public health systems; an overemphasis on institutionalized care; fragmented funding and service delivery, which impede scale-up; inadequate training and technical expertise; inconsistent and sporadic technical assistance; ineffective diagnostic and referral networks; limited information technology (IT) and data collection; and inadequate financial resources. Following a review of the merits of vertical versus horizontal programs, the workshop turned to an examination of approaches that can be taken to address these infrastructure problems and the role that can be played by IT. In addition, a detailed case study of health care delivery programs in Cambodia and Ethiopia was presented (see Box 5-1 at the end of the chapter).

Box Icon

BOX 5-1

Universal Access for MDR Care: The Cambodian and Ethiopian Perspectives. In Cambodia, the incidence of TB is currently approximately 500 per 100,000 population; MDR TB represents about 1.6 percent of new TB cases and accounts for about 2 percent of treatment-naïve (more...)


A debate exists on the relative merits of vertical and horizontal programs. Vertical programs are donor driven and focus on a single disease, such as HIV or TB. Farmer stated that some in the global health community have criticized vertical programs, saying that they may fragment care, cannibalize funding and resources, and create inefficiencies, resulting in missed opportunities to treat multiple issues in an integrated fashion. Horizontal programs strengthen health systems so that they are able to address a broad range of needs, or at least to coordinate related disease programs. For example, establishing a diagnostic program only for TB is clearly inefficient. Yet the counterargument is that building public health infrastructure takes too long, and in the interim treatment will be denied to those in need. Farmer suggested that this debate is unnecessary: vertical programs, including those funded by such sources as PEPFAR and the Global Fund, can accomplish their immediate disease-specific goals while at the same time strengthen health systems and primary health care.


Addressing infrastructure problems at the local or national level in high-burden countries could have a positive impact on the global TB epidemic. Harrington suggested that one major problem to be overcome is the institutionalization of drug-resistant TB patients in improper settings, which may increase the risk of transmission (see Chapter 3), lower the morale of health care workers, and diminish patients’ quality of life. In addition, many high-burden countries fail to use and expand existing human resources to provide ongoing follow-up and treatment support for people with all forms of TB. Through its community-based treatment program in Haiti, Partners In Health has demonstrated the need not only for trained doctors and nurses, but also for community health workers, treatment supporters, and laboratory technicians—all of whom need to be paid for their work. Farmer stated that investing in community health workers is what makes success sustainable. Many high-burden countries have high unemployment rates and a substantial supply of educated people who could take on these roles. Harrington suggested that creating systems that pay for treatment support needs to be a top priority.

Carol Nacy of Sequella, Inc., suggested a mechanism for training health care workers in high-burden countries. Developed countries could create curricula based on existing medical technology programs in the United States that could be formatted for distance education. Such programs could train individuals to perform clinical laboratory work and bring workers in developing countries up to the same skill levels as those in developed nations. Nacy also suggested creating a health volunteer corps, modeled after the Peace Corps, which would provide TB treatment in developing countries.

Finally, Farmer discussed the need for standards with respect to treatment and the organization of health care delivery, as well as metrics for measuring effectiveness and quality. The IOM and other academic and standards-setting bodies could play an important role in establishing such metrics, which would enable aid organizations to learn, for example, how vertical programs can strengthen health systems, how TB programs can deal optimally with complex forms of the disease, and how best to institute infection control programs.


Dale Nordenberg of Brigham and Women’s Hospital addressed the role of IT in dealing with the global TB crisis. An information management infrastructure providing systems that support the disparate work of the laboratory technician, clinician, and public health program manager is essential. Globally, the demand for cultures and DST is great—on the order of 60 million cultures and 6 million DSTs annually (see Chapter 4). Thousands of new laboratories are needed, but so, too, are sound information systems. As laboratory volume and complexity increase, a laboratory information management system (LIMS) is needed to manage specimens and data effectively. These laboratory data are then transmitted to (1) medical record systems that provide physicians with a complete medical record to support care optimization, and (2) surveillance systems that support public health program management, such as prediction of medication requirements for particular programs. The challenges of building such systems, and doing so cost-effectively, are substantial in any setting, but are particularly daunting in resource-poor environments.

One vital role of IT in dealing with the TB crisis is enabling countries and projects to share data. Yet the costs involved in creating such a capability are a tremendous barrier to be overcome. In the United States, the estimated cost to establish a LIMS in a state public health laboratory is about $1–1.5 million per laboratory in the first year alone, with annual maintenance costs of approximately $500,000. In addition to the cost, however, complexities are involved in the implementation of a LIMS that are most pronounced when there is interest in and a need to share data among laboratories and programs.

A strategy for developing such a system has certain essential elements. Standards for effective data sharing must be formulated—a difficult task when technologies change rapidly, resulting in differences in case definitions, tests, manufacturers, and sensitivities and specificities. Nordenberg shared some of the lessons learned from the ongoing Public Health Laboratory Interoperability Project (PHLIP)—a collaboration between the Association of Public Health Laboratories and CDC. The objective of the project is to build a community of laboratories that can work collaboratively to help meet the challenges of building a national laboratory data sharing network. The laboratories recognized that each had different approaches and capabilities to generate diagnostic data for specific diseases because of differing public health priorities, disparate methodologies used by their scientists, and implementation of a variety of technology solutions depending on their budgets and expertise. By working as a community, PHLIP continues to lead the development of public health use cases for laboratory data, supporting data requirements, and a data sharing scheme that will enable laboratories to support public health practice. The project is driven by a recognition that the creation of an efficient data sharing network is the product of close collaboration and harmonization of public health goals.

Harnish Fraser of Brigham and Women’s Hospital described a system developed in Peru to illustrate the challenges associated with building what was ultimately a successful medical record system for MDR TB used both by the nonprofit Socios En Salud and the Peruvian Ministry of Health. The standard laboratory structure in Peru is hierarchical. Information typically passes from a small clinical laboratory that may do only smears to a larger regional laboratory that performs cultures and DST, to a national reference laboratory. Results from the larger laboratories return to the smaller laboratories in hard copy—a process that can result in long delays and loss of information. Before the project began, a median of 143 days elapsed between reporting of laboratory results and initiation of a revised second-line treatment regimen for a patient.

Peru’s MDR TB electronic medical record system was also designed to assist in predicting drug supply needs. The system made it possible to model how long each patient was likely to be in treatment and to estimate what medications would be required for a group of patients. By combining the information on recruitment rate, time in treatment, and proportion of patients taking each drug based on the regimen that had been entered into the system, supply requirements for 6 months or longer could be predicted within about 5 percent or better.

Fraser also described a new open-source free electronic medical record system, OpenMRS.1 This system is designed to support the collection and management of data on any medical condition, using a data dictionary to add new items. Partners In Health has now created a version of this system that supports the treatment of MDR TB as well as HIV. OpenMRS is also designed to support electronic data exchange with other information systems for laboratory, pharmacy, and reporting purposes. Open-source software is particularly useful because it allows multiple countries and projects to share in the system development process and maintain control over the data and the system; it can eliminate delays normally encountered when one is learning to build a system and make it easy to build individual implementations on top of an existing system. To illustrate the benefits of such an approach, Fraser described an HIV treatment project that created a data dictionary, which made it possible to share data among different projects and compare outcomes across sites without the need to program the system. Following the dictionary’s creation, the source code was released with a public license so it could be downloaded. With some funding from WHO, a similar system was developed for managing TB and MDR TB patients. This effort was part of a much larger initiative that is being supported by a number of organizations, including CDC, the Rockefeller Foundation, the Canadian International Development Research Centre, and a number of other funders, to create common standards and foster collaboration.



Information about OpenMRS and the program itself is available at http://www​

Copyright © 2009, National Academy of Sciences.
Bookshelf ID: NBK45008


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (12M)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...