Click for next page ( 50


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 49
5 Diagnosis of Drug-Resistant TB Key Messages • apid diagnostic methods would permit more immediate initiation R of effective treatment, thus reducing the amount of time that MDR TB patients are infective. • olecular-genetic methods, including gel-based biological micro- M chips, can reduce diagnostic intervals to as little as 1−2 days. • aboratory information management systems allow use of diag- L nostic results to maximum advantage and monitoring of treatment results. • he effectiveness of these laboratory information management T systems would be enhanced if information could be shared in common public health databases even if the information manage- ment systems were based in different technology and software platforms. Presentations on the diagnosis of drug-resistant TB addressed rapid diagnostic methods, the use of biochip technology, and the need for improved laboratory capacity. 49

OCR for page 49
50 DRUG-RESISTANT TUBERCULOSIS IN RUSSIA RAPID DIAGNOSTIC METHODS1 CTRI of the Russian Academy of Medical Sciences uses both conven- tional and newer, rapid methods of mycobacteria identification (see Box 5-1 for an overview of current TB diagnostic methods). The conventional methods used are fluorescent microscopy and culture, which can require up to 10 weeks for culturing and an additional 4 weeks for drug susceptibility testing. During this time, physicians and patients must wait to determine how to treat the patient’s TB, and this delay provides an opportunity for the disease to spread. To shorten this time, CTRI has been using commercial products that rely on culturing and on molecular-genetic methods. These products can determine which strains are resistant or sensitive to specific drugs through colorimetric methods, with automatic detection and no use of test tubes, and reduce the time required to obtain drug susceptibility results to 6−13 days. The molecular-genetic methods rely on detection of mutations in the DNA of M.tb. that convey drug resistance. One such method uses biochips developed in Russia (see the next section). It involves the extraction of M.tb. DNA, two-stage polymerase chain reaction (PCR), hybridization with amplicons labeled with fluorescent marks on the biochip, and detection of results using an analyzer with subsequent computer processing. A comparison of biochip and culture data showed a concordance of 95 percent for rifampicin resistance, 88.5 percent for isoniazid resistance, and 87 percent for fluoroquinolone resistance. These are reassuring numbers, said Larionova, since not all mutations responsible for resistance are found using biochips. Other molecular-genetic methods used to detect drug resistance involve DNA-strip or related technologies. They rely on DNA extraction, ampli- fication by PCR, hybridization on strips, and visualization and estimation of results. These methods are highly safe, easy to use, and cost-effective. Results are available within 1−2 days and can be obtained from either solid or liquid media. A comparison of results from biochips and DNA strips demonstrated full concordance. The above technologies allow for the detection of MDR and XDR TB during a patient’s examination and the administration of adequate chemotherapy regimens to shorten the sputum conversion period, improve treatment outcomes, and prevent the spread of disease. Experience at CTRI indicates that 64 percent of new MDR TB cases convert after 2 months of treatment and 87 percent after 6 months of treatment. New methods are also being used to diagnose infection with nontuber- 1 This section is based on the presentation of Elena Larionova, Central TB Research Institute, Russian Academy of Medical Sciences.

OCR for page 49
51 DIAGNOSIS OF DRUG-RESISTANT TB BOX 5-1 Some Diagnostic Methods Currently in Use for TBa Microscopy smear. Experience has shown that microscopy can de- tect TB, but the sensitivity is variable and can be very low. Culture/phage based. Culturing bacteria takes longer than a smear test but is more sensitive. Smear-negative but culture-positive tests allow for earlier treatment and a reduction in transmission. Molecular, bacteria based (e.g., PCR). Many reports on the per- formance of PCR in diagnosing TB have appeared since 1985. Positive results from these types of tests do not guarantee live bacteria, and repeatability issues have arisen. Specificity and sensitivity depend on the kind of sample, the kind of test, and the manufacturer. PCR may not be much better than culture for “difficult samples” such as pleural fluid or urine, and the sample preparation method can generate problems. Ac- cording to Paul van Helden, Stellenbosch University, PCR also can be very expensive unless a cartridge-based test is used, and even the cost of a cartridge-based PCR test, at about US$40 or more per person, is unaffordable in the developing world. PCR has a number of applications beyond the diagnosis of TB. Scientists and clinicians can use it as a basic research tool, to assign isolates of M.tb. to a particular strain, and to obtain drug resistance information. Furthermore, many applications can be automated to reduce costs. In the future, for example, multiple fluorescent probes might generate considerable information simultane- ously, said van Helden. DNA based. As mentioned in Chapter 2 (Box 2-1), since the work- shop was held in Moscow, a new, fully automated DNA test (Xpert MTB/ RIF) for TB has been validated and subsequently recommended by the WHO for broad implementation as the initial diagnostic for individuals suspected of having MDR TB or HIV−TB coinfection. The test simultane- ously detects TB and rifampicin drug resistance (a reliable indicator for MDR TB) in sputum. WHO reports that FIND has negotiated a reduced price for 116 low- and middle-income countries (including South Africa, Russia, India, and China) of US$16.86 per test cartridge. The test pro- vides results in 100 minutes, allowing proper treatment to begin imme- diately (WHO, 2010d). SOURCE: IOM, 2011. aThe information provided in this box was originally presented by Paul van Helden, Stellen- bosch University, and summarized in the IOM Drug Forum’s second workshop on the subject of drug-resistant TB held in 2010 in Pretoria, South Africa (IOM, 2011).

OCR for page 49
52 DRUG-RESISTANT TUBERCULOSIS IN RUSSIA BOX 5-2 New Methods for Species Identification of Nontuberculosis Mycobacteriaa Infection by nontuberculosis mycobacteria is the cause of mycobac- teriosis, a frequent opportunistic infection in patients with AIDS. New liquid culture-based automatic systems and molecular-genetic systems have made it possible to identify both NTM infection and the species of NTM involved. The gold standard for NTM detection today is sequencing of the bac- terial genome, but not all laboratories are equipped with DNA sequenc- ers. Restriction fragment length polymorphism analysis using PCR is much more commonly employed, and several commercially available test systems exist. Mass spectrometry is another technology used to identify NTM spe- cies. Smirnova is involved in an effort to identify NTM strains by direct protein profiling using mass spectra for ribosomal proteins. She and her colleagues have collected 35 strains of NTM and have characterized each using several methods, including microbiological, PCR-based, and mass spectrographic techniques. Accumulation of mass spectrometry data for conserved proteins will allow for the creation of a database that can be used for species identification. Little information about mycobacteriosis is available in Russia, said Smirnova, and species identification of mycobacterial cultures is rarely performed in bacteriological laboratories. In addition, no molecular- genetic systems are available for the quick and inexpensive identification of NTM. Spectrographic determination may offer a way to fill this gap. culosis mycobacteria (NTM) and to identify the species of NTM involved. Box 5-2 summarizes a presentation on this subject. In the discussion period, Maria Y. Giovanni of the National Institute of Allergy and Infectious Diseases (NIAID), U.S. National Institutes of Health (NIH), noted that NIAID supports a comprehensive program of research and development in diagnostics. She also suggested that all diag- nostics still need to be faster, easier to use, and less expensive. Kathleen Eisenach, University of Arkansas, pointed out that the Institute of Tropical Medicine in Antwerp does quality control testing for drug susceptibility tests worldwide. A panel of isolates, including MDR TB strains, is circu- lated to laboratories once or twice a year, which provides an opportunity for proficiency testing. “These coordinated efforts help us have confidence

OCR for page 49
53 DIAGNOSIS OF DRUG-RESISTANT TB Many other questions surround the diagnosis and treatment of NTM infection. How can drug susceptibility testing for NTM be performed? Who should treat patients with mycobacteriosis, where should they be treated, and how, given that mycobacteriosis is not TB? Physicians are insufficiently educated in the diagnosis and treatment of patients with mycobacteriosis, and delays in diagnosis and treatment can lead to severe and progressing disease. In the discussion period, Carlos Pérez-Vélez of the National Jewish Hospital, Denver, Colorado, commented on the treatment of mycobac- teriosis at the hospital, which typically has 10 such patients who have been referred from around the world. An interdisciplinary team of pul- monologists and infectious disease specialists manages these patients, who usually require prolonged courses of multiple antibiotics. Success for these patients depends on susceptibility testing, the management of adverse effects from antibiotics, and pharmacokinetic studies of indi- vidual patients. “The same doses in people of similar weight [and other characteristics] can vary, so sub-therapeutic dosing is really a problem in nontuberculous mycobacterial infections,” said Pérez-Vélez. Many pa- tients with MDR TB actually have an NTM infection that is resistant to isoniazid and rifampin, and some people have a mixed infection, he noted. These mixed infections can be very difficult to diagnose, since one bacterium can outgrow and overshadow another. aThe information presented in this box is based on the presentation of Tatiana Smirnova, Central TB Research Institute, Russian Academy of Medical Sciences. in the drug susceptibility testing that is being performed at all levels in all countries,” she said. Several participants discussed the prospects for diagnostics that use sputum or some other sample and do not involve culturing. Coetzee noted that several technologies do not require culturing and have produced good results. However, they do not work as well in HIV-infected patients, who have a lower smear-positive rate than other TB patients. Several participants compared the cost of diagnostics with the cost of treatment. Given the expense of second-line drugs, even quite expensive diagnostics may be cost-effective. “A day of capreomycin, even at discount rates, costs $3 to $5,” said Cassell. Therefore, she suggested, it makes no sense to wait for a $1 test. Accurate diagnostics also could obviate the need for expensive follow-up tests.

OCR for page 49
54 DRUG-RESISTANT TUBERCULOSIS IN RUSSIA Keshavjee pointed out that airline security systems now rely on mass spectrometers to detect bomb-making chemicals within a few seconds. In an ideal world, a TB diagnostic would produce results instantly so that the proper treatment could be initiated at the point of care. Even if diagnostic technologies are not perfect, they can suggest a treatment at the point of care, and that treatment can be modified once more definitive results are available. BIOCHIP TECHNOLOGY FOR TB DIAGNOSIS2 Gel-based biological microchips were developed by Andrei Mirzabekov at the Engelhardt Institute of Molecular Biology in the early 1980s, and much progress has been made on their further development since then, said Zimenkov. Today, biochips are based on three-dimensional gel pads on a plastic surface rather than two-dimensional glass surfaces, which increases the sensitivity of analysis and leads to excellent discrimination levels. Point mutations are detected by DNA hybridization on the biochip, with fluores- cence intensities being compared to determine whether tested DNA bears a particular mutation. Software is user-friendly for medical personnel, and the biochip analyzer has been certified in clinical trials. In 2001 the Institute of Molecular Biology received support from the International Science and Technology Center (ISTC) to study the applica- tion of biochips in TB diagnostics for fast discrimination and strain typing of MDR TB in Russia. Biochips were developed to detect mutations leading to resistance to rifampicin and isoniazid. The chips were able to identify more than 95 percent of rifampicin-resistant TB strains and more than 80 percent of isoniazid-resistant strains. Since being certified in 2004, the bio- chip has been used in more than 10,000 analyses. The use of biochips has had a major impact on treatment. When bio- chips were used to determine second-line treatments in MDR TB cases, healing, or bacterial conversion, was approximately twice as rapid as when classical methods were used, according to Zimenkov (Kuzmin et al., 2006). Similar results were obtained in a comparison of three groups: one con- sisting of people resistant to one or two drugs who received the standard treatment, one consisting of MDR TB patients who received the standard treatment, and a third treated after biochip analysis (Morozova, 2008). A more recently developed biochip detects mutations involved in resis- tance to fluoroquinolones in about 80 percent of strains. Thus far it has been used in more than 3,000 analyses in Russia and other countries. New 2 Thissection is based on the presentation of Danila Zimenkov, Engelhardt Institute of Molecular Biology.

OCR for page 49
55 DIAGNOSIS OF DRUG-RESISTANT TB chips being developed detect additional mutations involved in fluoroqui- nolone resistance. Biochips also are being used to perform automated analyses of repeated units in the DNA of M.tb., allowing different strains to be distinguished. For example, the technology can distinguish bovis strains from Beijing strains in about a day. The technology can now distinguish among more than 100 strains of the bacterium, making it possible to determine the extent to which strains are mixed in patients. This technology is being extended to produce a biochip for mycobacteria differentiation. In the discussion period, several workshop participants expressed inter- est in the biochip, which was developed in part through a partnership with Argonne Laboratory in the United States. Jeffrey Drazen, New England Journal of Medicine, urged that quantitative research techniques be used to study the biochip to determine its sensitivity and specificity relative to standard diagnostic techniques. Drazen said that this type of information would move the field from descriptive to quantitative research and would be a positive step forward in terms of the scientific development of diag- nostic techniques. NEED FOR IMPROVED LABORATORY CAPACITY3 WHO’s Global Laboratory Initiative (GLI) has identified the need for an urgent and massive scale-up of TB laboratory services. According to the GLI, “The global lack of TB laboratory capacity constitutes a global crisis, requiring a paradigm shift in providing laboratory policy guidance, quality assurance and knowledge creation within a global and integrated laboratory network.” Some of the challenges and successes of improving laboratory diagnostic capacity are illustrated in Box 5-3. Several critical issues surround global laboratory capacity, Nordenberg said: • L aboratory capacity is desperately insufficient. • L aboratory capacity-building efforts rarely consider data and infor- mation management. • L aboratory programs are focused on specimens and therefore have information system requirements very different from those of clini- cal or public health programs. • E merging diagnostics can change surveillance methods or the sen- sitivity and specificity of both TB and drug-resistant TB testing, 3 This section is based on the presentation of Dale Nordenberg, Novasano Health and Science.

OCR for page 49
56 DRUG-RESISTANT TUBERCULOSIS IN RUSSIA which in turn affects surveillance trend estimates and outbreak control. • T here is a critical need to integrate the information systems of laboratories, clinics, and public health programs. • T here is a critical need as well for “operations” systems to track such activities as infection control programs and therapeutic supply chains. BOX 5-3 Diagnostics and Laboratory Infrastructure in South Africaa Dr. Coetzee discussed South Africa’s latest efforts to organize labo- ratory services for the effective diagnosis of drug-resistant TB. In South Africa, simultaneous infection with HIV and TB has created a widening gap between smear positivity and TB cases. As a result, smear micros- copy is rapidly failing as a diagnostic tool and will soon become unusable, Coetzee said, emphasizing that new diagnostics are desperately needed to combat drug-resistant TB. In 2006 WHO held a meeting in South Af- rica that generated a recommendation to develop and implement rapid diagnostics throughout the country. In response, South Africa invested heavily in line probe assays (LPAs).b Currently, line probes are running in 15 laboratories, and they will be available in 10 more before the end of 2010. However, this scaling up of LPA capacity is an extremely difficult task. A significant challenge is the country’s workforce constraints, in particular the small number of molecular biologists. The probes also were placed in laboratories with good ventilation and working environments, and such facilities are among the top few percent of African laboratories, said Coetzee. The performance of the implemented line probes has been high. How- ever, the reading of the line probes had to be standardized. “We found inter-observer errors in PhDs, so you can imagine what the rate would have been in technicians,” said Coetzee. The line probes are scanned and interfaced with the laboratory information system, which has pro- duced standardized results. The National Health Laboratory Service has about 350 laboratories, and all of the public-sector laboratories have been consolidated into one organization. This consolidation has enormous benefits for surveillance. The development of an algorithm for the early detection of MDR TB was a politically challenging process, Coetzee said. In the end, however, a policy was established that a line probe will be administered to every new smear-positive patient. If a patient is smear negative and culture positive,

OCR for page 49
57 DIAGNOSIS OF DRUG-RESISTANT TB The challenge, said Nordenberg, is that these issues apply not just to one or two but to thousands of laboratories. To stop the spread of TB globally, the world needs rapid, accurate diagnostics, particularly in resource-poor settings. Also needed are drugs that will shorten treatment, be effective against both susceptible and resistant strains, be compatible with antiretroviral therapies, and improve the treatment of latent infection; a vaccine that is safe and effective for children, adolescents, and adults, the culture will undergo a line probe. This is an expensive process, but less so than first-line drug susceptibility testing. The infrastructure in much of South Africa is rural, with scattered electricity and water supplies. However, cell phone reception is often available. The country has begun using cell phone printers, through which messages can be exchanged. With the rollout of the line probes, TB facilities throughout the country began using a new form that makes it possible to follow patients longitudinally through the system and monitor their adherence to the TB control program. Information from the forms is migrated to a central data warehouse in Johannesburg to produce a single consistent view of the data. Information includes all demographic data, all test results, all drug susceptibility test results, and billing infor- mation. The new system is still being piloted, but Coetzee said the hope is that it will improve adherence to policy. Making the transition from specimen-based to patient-based data has been difficult. Demographic information is often inadequate or inac- curate. Data security and the user interface required considerable work. Some of the data contained in approximately 12 million records had to be checked visually. The system also is being used for other infectious diseases in South Africa, such as cholera and meningitis. Notification of an outbreak is automatic through web-based portals. Reports can be generated by province, by district, by subdistrict, or by clinic. South Africa is now en- hancing its reporting for national health programs focused on HIV, TB, cervical cancer, and sexually transmitted diseases. It also is working to improve the spatial reporting of infectious diseases to assist various health authorities. aThe information in this box is based on the presentation of Dr. Coetzee. bLine probe assay technology involves the isolation of DNA from sputum specimens for the rapid detection of MDR TB.

OCR for page 49
58 DRUG-RESISTANT TUBERCULOSIS IN RUSSIA including people with HIV; and efficient and sustainable information supply chains (discussed below). The New Diagnostics Working Group of the STOP TB Partnership has identified several critical attributes of TB diagnostics (WHO, 2009), finding that such diagnostics should: • s implify and improve the detection of TB, including smear-negative, extrapulmonary, and childhood TB, through increased sensitivity and specificity and improved accessibility; • o ffer simple, accurate, safe, and inexpensive tests that can be per- formed at the point-of-care level of the health care system and produce same-day results; • e nable more effective monitoring of TB treatment for both latent and active cases; • r apidly identify resistance to both first- and second-line TB drugs; and • r eliably identify latent TB infection and determine the risk of pro- gression to active disease, enabling the rational use of preventive therapy. Nordenberg suggested that information should itself be seen as an intervention. The supply chain needs to get the right therapeutics to the right patient. If a diagnosis is not used to drive treatment, a major oppor- tunity is being missed. If cases are identified earlier, less resistance will develop, patients will be less costly to treat, and fewer people will spread the infection. Thus, the return on investing in information infrastructure is very high. “Without that investment, all the investment in diagnostics is compromised,” said Nordenberg. The data produced by diagnostics can have a major impact on pub- lic health, but they need to be collected, managed, and shared. Doing so means moving across what Nordenberg called the “information chasm” from laboratories to epidemiological and public health impact. Crossing that chasm requires laboratory information management systems. Such a system is a tool that supports the work of laboratories as opposed to that of clinicians or health epidemiologists. A laboratory information management system needs to perform a diverse set of functions—test requisition, test receipt documentation, sample management, testing and validation, report distribution, report receipt documentation, test scheduling, sample collec- tion, chain of custody, reagent management, quality assurance, and others. These functions are not in the domain of a clinician or epidemiologist. With influenza, for example, various types of tests can be performed on a given specimen. A laboratory information management system has to document and be the repository of results from this complex analytic process.

OCR for page 49
59 DIAGNOSIS OF DRUG-RESISTANT TB About 80 percent of the 50 public health laboratories across the United States have a laboratory information management system, after about 7 years of implementation efforts. Each state is implementing its own system, while approximately five to seven different primary systems cover the 50 states. The laboratories perceive that they have differing needs with regard to public health priorities, bench methodologies, and technologies. This is also the case internationally; some countries in Africa, for example, are implementing two or three different systems, even within the same city. With the emergence of new technologies and new diagnostics, more- over, local laboratories are conducting more of their analyses locally and sending fewer specimens to regional or national laboratories for analysis. As a result, local laboratories are starting to build their own data-sharing networks. These networks need to be integrated with clinical, public health, and other research activities. They also need to be able to handle not just TB but also other infectious diseases. A public health laboratory is typically responsible for a broad spectrum of programs, and a laboratory informa- tion management system needs to support that mission. Nordenberg used influenza as an example. More than 60 tests are used to describe and manage an influenza epidemic. To facilitate data exchange, the national laboratory community has made more than 500 specific data-related coding decisions. Nordenberg emphasized the importance of being able to share diagnostic data, which requires such collaboration. Even though laboratories are working with different technology platforms, they are identifying common data that they will share, so the information kernel is standard. This kernel, derived from a description harmonized by a community, must adjust dynamically as new scientific methodologies and technologies emerge. Even before the first version of a product has been fully implemented, new versions can appear that overlap with the previous versions. At any given time, multiple versions may be in use. Only through ongoing collaboration can a systems approach produce an alignment of vision, mission, and execution. Nordenberg emphasized that a public health information supply chain is not an abstract concept, but something that must be engineered through a comprehensive, logistics-based approach. In particular, a disciplined approach to data and information provisioning will enable measurement of the quality and impact of the data, allowing performance to be improved. A systems approach, in contrast to an ad hoc approach, can produce several additional capabilities. A systems approach is scalable, so that it can meet the need for robust information supply chains in thousands of laboratories across the globe. It is intentionally designed to be dynamic so that it can be optimized through continuous performance improvement. And it provides the ability to integrate data across laboratories, clinical programs, and public health programs. A systems approach transcends a

OCR for page 49
60 DRUG-RESISTANT TUBERCULOSIS IN RUSSIA vertical, single-disease approach and can support diverse health care priori- ties and programs. Nordenberg outlined additional important capabilities of a laboratory information management system: • i t must be sustainable; • i t must be cost-effective; • i t must leverage and build local expertise; • i t must be driven by public health and science, not just by technology; • i t must be governed by stakeholders; and • i t must provide a clear path to robust capability while accommo- dating a diversity of baseline capacity. An information supply chain is distinct from the other supply chains required to manage a TB or MDR TB program, such as people, hard goods, diagnostics, therapeutics, reagents, and so forth. Nordenberg noted that information is often overlooked when laboratory capacity is being built. As a result, systems are frequently developed in an ad hoc manner, and there is a divide between the technology platform and the ability to get the data where they are needed. The components of an information supply chain are information prod- ucts, a source of raw materials, human resources, and standard operating procedures, each of which must be articulated with the others. Information products are motivated by the questions that drive decision making, and all supply chains must reflect a clear idea of the products they need to produce. The sources of raw materials are the data systems that provide the data for information products. Human resources are the staff that build and operate the systems and develop the information products. Standard operating pro- cedures are the processes used to manage the data to produce information. “The information supply chain is a complex endeavor,” said Nordenberg. “It’s much easier to buy a piece of technology, or buy a system, install it, collect the data, and then hope you get out of it what you need. But usually you are going to be disappointed.” To realize the full benefits of information systems, the laboratory, clinic, and public health entities in a given area need to work together. Nordenberg illustrated this point: “If you can get a laboratory result in 2 days or in 3 weeks, but then it takes another week to get the result of that test to some- body, that’s going to have serious costs in terms of delayed diagnosis and increased spread.” Especially with TB, the cost of a broken information supply chain can be calculated in both human and financial terms. Nordenberg offered several recommendations regarding information supply chains:

OCR for page 49
61 DIAGNOSIS OF DRUG-RESISTANT TB • C ountry plans for information supply chains should be developed to support TB and MDR TB control. • L aboratories should collaborate multinationally and sustainably to develop shared best practices. • P lans should be developed to migrate laboratories at all levels of information technology capability to the same target capability. • A tight linkage should exist between diagnostics development and data activities. • T echnology adoption and information capability should be tracked to guide programs. • M etrics related to time from specimen acquisition to diagnosis and treatment should be developed and tracked. • I nformation plans should be assessed annually to respond to new developments in such areas as diagnostics, drugs, and intervention programs. • H IV and TB data systems should be integrated to support the man- agement of coinfected patients. • E ducational programs should impart a clear understanding of the specific needs of laboratories versus clinical and epidemiological activities. During the discussion of information systems, Coetzee, Nordenberg, and Renzhong Li of China’s Center for Disease Control and Prevention discussed the degree of integration between the clinical and laboratory sys- tems in different countries. In South Africa, said Coetzee, the systems are not integrated. Even large hospitals have systems that handle only admin- istrative and not clinical data. A new patient management system is being implemented for lower-level clinics, which will help integrate the treatment of HIV infection and TB. Also, there are limited interfaces between hospital management systems and patient management systems. Much HIV testing is no longer done at laboratories but at the point of care, which means that previously available surveillance data have been lost. In China, said Li, data from patient care are available but not from the laboratory system. A workshop participant asked how data from private practitioners in many countries can be integrated in the same system, since the same platform probably will not be used to communicate results. Nordenberg observed that the same situation exists in the United States, where a wide diversity of independently run systems exists in both the public and private sectors. He reiterated that effort is focused on creating an information ker- nel that is constant so the data can be shared even though the technologies differ. He also noted that the web-based patient-level system in China is remarkably successful, reaching thousands of different entities—from the

OCR for page 49
62 DRUG-RESISTANT TUBERCULOSIS IN RUSSIA county, to the prefecture, to the province, to the national level—across the country. A participant pointed out that the ideal situation is to get data from laboratories to clinicians rapidly so that treatment can begin. In some countries, this information flows via cell phone, although provisions must be made to ensure privacy. Also, private practitioners and laboratories need to be integrated into public systems. Nordenberg responded that the most efficient way to scale up com- munications is through web-based systems. Implementing thousands of information systems across laboratories of varying capacity is difficult. But the problem of connectivity can be and is being solved through web-based communications. In the short term, cell phone infrastructures can suffice, particularly if systems are set up using passwords to protect privacy, accord- ing to Nordenberg. It is also important, he said, for laboratories to share best practices. For example, if different diagnostics have differing sensitivity and specificity, how does that affect estimates of incidence? How can data be shared within or across countries? How are reporting forms designed? A workshop participant asked whether the widely varying results of drug susceptibility testing in China argue for the use of individualized rather than standardized treatments for MDR TB. Nordenberg asked whether the laboratory results are fed into the patient information system. Li responded that the patient data are referred to the TB dispensary. Nordenberg also pointed out that, despite the excellent patient information system, the lack of an integrated laboratory system in China makes it difficult to look at population trends in resistance. The challenge, he said, is to run a program focused both on resistance patterns and on patients.