Monitoring progress in TB control
The Millennium Development Goals, together with the Stop TB Strategy launched in 2006, have broadened the scope and aims of TB control. In 2004, however, the year in which the data for this report were collected, NTPs were focused on achieving the targets of 70% case detection and 85% treatment success under DOTS. From the data presented in this and preceding reports, we estimate that the detection rate of new smear-positive cases by DOTS programmes has continued to accelerate since 2000, reaching 53% globally by the end of 2004. Only six HBCs (Democratic Republic of the Congo, Myanmar, the Philippines, South Africa, Thailand and Viet Nam) had reached the detection target, and the estimate for at least one of these countries (Democratic Republic of the Congo) is uncertain. No WHO region had reached 70% detection by the end of 2004. If the observed rate of acceleration continues, case detection will exceed 60% by 2005, but will fall short of the 70% target.
The acceleration in case-finding since 2000 has been achieved both by improving detection within established DOTS areas and by expanding geographical coverage. Up to 2001, there appeared to be a ceiling on case detection (smear-positive and all forms of TB) at 40–50%: the series of case notifications showed that, although the detection rate by DOTS programmes had increased substantially between 1995 and 2001, the number of patients reported from all sources was more or less stationary. That was because DOTS programmes were recruiting and reporting TB cases that would have been reported anyway. The new data for 2004, when put together with those for 2002 and 2003, show that this ceiling has been breached.1 This has happened predominantly in the South-East Asia and Western Pacific regions, where DOTS programmes are recruiting TB patients from new sources, including clinics and hospitals in both the public and private sectors.2
The discussion of case detection in this and previous reports has focused on smear-positive cases, largely because the target for DOTS implementation is defined in these terms. Several WHO reports have, however, emphasized that, in the Region of the Americas and in the European Region, many TB cases are reported through the public health system but from outside DOTS programmes. This implies that target rates of case detection could be achieved relatively easily in these two regions by implementing the procedures required under DOTS, including the more frequent use of smear microscopy in the European Region. In other parts of the world, especially the Eastern Mediterranean Region, case detection must be improved by finding more patients in total, for example by increasing the number and diversity of clinics and hospitals that report TB cases.
Anticipating the development of new and more sensitive diagnostic tools, and pursuing the above comparison of detection rates among regions, we have compared various approaches to estimating case detection. A comparison of 25 European countries in 2004 shows that the proportion of culture-positive cases detected was typically lower than the proportion of smear-positive cases detected. We conclude that culture is seldom used as the principal or sole method of diagnosis in European countries, but rather as a supplementary or complementary method of diagnosis.3
On the other hand, the proportion of all TB cases detected in the European Region (diagnosed by all methods – smear, culture, radiography, clinical examination) was mostly higher than the proportion of smear-positive cases detected. There are two possible explanations: either the reported numbers of TB cases of all forms (numerator of the detection rate) are disproportionately high, or the estimated incidence rates of all forms of TB (denominator) are disproportionately low. The first explanation includes the possibility that smear-negative TB is over-diagnosed; the second implies that smear-negative cases – pulmonary or extrapulmonary – are underestimated. The observation that detection rates for all forms of TB exceeded 100% in several countries does not distinguish between these two alternatives.
Whatever the explanation for the pattern in the European Region, it is different from the pattern in the Region of the Americas, where smear-positive case detection rates were almost always higher than the detection rates for all forms of TB. Understanding the variation among case detection statistics is likely to be important in evaluating TB epidemiology and control in the Region of the American and in the European Region, and perhaps elsewhere in the world.
Outcomes of treatment
Although the cohort of patients treated under DOTS has grown from 240 000 in 1994 to 1.7 million in 2003, treatment success has edged closer to the 85% target, falling just short of it in 2003 (82%). The global average has been held below the target mainly by the African and European regions, where high proportions of patients fail treatment or die, or are lost from DOTS cohorts. HIV/AIDS and MDR-TB are major obstacles to TB control in Africa and eastern Europe, respectively, but incomplete cohort data from these regions show that programme management also continues to be weak. Eight HBCs had met the 85% target for treatment success based on the 2003 cohort. All of them are in the South-East Asia or Western Pacific regions, with the exception of Afghanistan where the case detection rate by the DOTS programme is relatively low.
Among HBCs, only the Philippines and Viet Nam had met the targets for both case detection and treatment success by the end of 2004. Given the delay in assembling data from around the world, the final assessment of whether these targets were reached globally by 2005, and in which countries and regions, cannot be made until the end of 2006.4 However, it is possible that the targets were reached in 2005 in the Region of the Americas and in the South-East Asia and Western Pacific regions. The HBCs that are most likely to have succeeded are Cambodia, China, India, Indonesia and Myanmar, besides the Philippines and Viet Nam. However, the 2005 reports from each of the regions and countries that appear to have met the targets will require careful verification.
The progress made in global TB control by the end of 2005 depends greatly on what has been achieved in eight countries that were inhabited by 61% of the patients who were undetected in 2004. For this reason, Bangladesh, Ethiopia, Nigeria, Pakistan and the Russian Federation will be under close scrutiny, in addition to China, India and Indonesia.
Whatever the results for 2005, it is clear that NTPs must continue, from 2006 onwards, to improve case-finding and treatment success within the framework of the new Stop TB Strategy. The targets of 70% case detection and 85% treatment success are milestones, not end-points. They should be regarded as minimum requirements for all countries, all regions and globally.
Epidemiological trends and the impact of TB control
Where DOTS has been intensively implemented in the past five years, we expect to find evidence that incidence is beginning to decline. That evidence may be obscured by the continuing efforts made by NTPs to improve case-finding. However, some countries, or parts of countries, have apparently had high and stable case detection and treatment success rates for at least five years, and yet there are no indications that national case notification rates are falling. Viet Nam is a conspicuous example, and case reports in this country are now being examined for signs that incidence is falling in at least some age groups or in some parts of the country. In other countries, notably India and the Philippines, case notification rates have fallen for some periods during the past 10 years, but it is not certain that these trends reflect a real decline in incidence (rather than failing surveillance or improved diagnosis, for example) and, if so, whether the decline is the direct result of TB control. To help quantify the impact of DOTS and other factors that influence TB epidemiology, all countries should carry out detailed analyses of trends in case notifications – disaggregated by age, sex, place and other patient attributes – thereby making full use of the wealth of routine surveillance data that are available.
Based on data aggregated at national level, the TB incidence rate was, by 2004, falling or stable in seven out of the nine epidemiologically different regions of the world defined in Figure 5. Incidence rates in eastern Europe (mostly countries of the former Soviet Union) and Africa (countries with low and high HIV rates) increased during the 1990s, but appear to have peaked in Europe around year 2000, and have since fallen. While case notifications are in decline in Europe as a whole, they continue to increase in some eastern European countries, or in parts of these countries. There is no way of predicting when incidence will peak and at what level in African countries, but the rates of increase slowed markedly during the 1990s. Because the epidemic is growing more slowly in the African and European regions, it is also growing more slowly globally. The worldwide incidence of all forms of TB reached 140 per 100 000 population in 2004 (8.9 million new cases, including those who are HIV-positive), and was growing at about 0.6% annually.
Besides the trends in case notifications, changes in TB epidemiology can be measured through sequential population-based prevalence surveys of infection and disease. Such surveys are logistically demanding and costly, though surveys have recently been carried out in Cambodia,5 China,6 India7 and Indonesia,8 and more are planned or under way in Eritrea, Myanmar, Somalia, the United Republic of Tanzania and Viet Nam, among other countries. While these surveys have provided, or are likely to provide, important additional information about the impact of TB control, routine surveillance will continue to be the principal source of information for all countries.
Since few national, population-based surveys of TB prevalence and deaths have been done, and since TB death registrations are far from complete, we have made indirect estimates of progress towards the targets of halving the 1990 prevalence and death rates (Table 1). Prevalence and death rates, like incidence rates, have been rising in Africa, and more steeply in African countries with the highest rates of HIV infection. They have been falling in five out of the six WHO regions (i.e. excluding Africa), and in six of the nine subregions of the world shown in Figure 5 (i.e. excluding two regions of Africa and eastern Europe). The net effect globally, in our assessment, is that prevalence and death rates have fallen between 1990 and 2004.
The epidemiological forecast for 2005 and beyond is set out in the Global Plan. Even if the targets of 70% case detection and 85% treatment success are narrowly missed in 2005, the recruitment of well over 20 million patients by DOTS programmes in the past 10 years gives enormous momentum to the new plan. The $56 billion plan demands that 50 million patients be treated between 2006 and 2015, reaching case detection rates that will be greater than 75% worldwide by 2010, and over 80% by 2015. These improvements in the number of patients treated, when implemented with other components of the Stop TB Strategy, should reverse the rise in TB incidence by 2015, and halve prevalence and death rates globally (if not in Africa and eastern Europe). The plan must be fully implemented from 2006 onwards, and the targets for epidemiological impact achieved by 2015, if there is to be any chance of eliminating TB by 2050.
1 The apparent ceiling is discussed in: Dye C et al. What is the limit to case detection under the DOTS strategy for tuberculosis control? Tuberculosis (Edinb), 2003, 83:35–43.
2 More information about collaborations between public and private practitioners and institutions in TB control can be found at: www.who.int/tb/dots/ppm/en/index.html.
3 For further data, see Tables 10 and 13 of: EuroTB. Surveillance of tuberculosis in Europe. Paris, Institut de Veille Sanitaire, 2003 (www.eurotb.org).
4 The 2007 report in this series will give case detection rates achieved by the end of 2005, and treatment success rates for patients who are enrolled during 2004 and who complete treatment during 2005. These data will form the basis of WHO’s declaration to the 2007 World Health Assembly, stating the number of countries and regions that met the 2005 targets.
5 National Center for Tuberculosis and Leprosy Control, Ministry of Health, Royal Government of Cambodia. National TB prevalence survey, 2002, Cambodia. Final report, August 2005.
6 China Tuberculosis Control Collaboration. The impact of tuberculosis control in China. Lancet, 2004, 364:417–422.
7 Chadha VK et al. Annual risk of tuberculous infection in four defined zones of India: a comparative picture. International Journal of Tuberculosis and Lung Disease, 2005, 9:569–575.
8 Soemantri S et al. Reduction in the prevalence of pulmonary tuberculosis in Indonesia, 1980–2004 [in preparation].