Map displaying BCG vaccination policy by country. A: The country currently has universal BCG vaccination program. B: The country used to recommend BCG vaccination for everyone, but currently does not. C: The country never had universal BCG vaccination programs. doi:10.1371/journal.pmed.1001012.g002 

Context in source publication

Context 1

... of these changes to the BCG policies in various countries, it is necessary to not only know the current BCG vaccination policies but also past policies and applicable changes when dealing with adults who received BCG vaccination in childhood. Since the publication of the M. tuberculosis genome, comparative genomic studies have documented that BCG vaccine strains have evolved and differ from each other and from the original BCG first used in 1921 [9]. Because these genetic differences affect antigenic proteins, these changes may translate into differences in efficacy and effect on the tuberculin skin test (TST) [10,11]. Work done by Ritz and Curtis looking at global BCG strain variations demonstrates the diversity of strains used by different countries and even within the same countries [12]. They found that 44% (83/188) of countries reported using more than one BCG strain type during an interval of only 5 years. This highlights the importance of docu- menting BCG vaccination practices for both clinical and research purposes. Clinicians cannot be expected to know BCG practices in all countries, and immigrants themselves may not know about the vaccination policies in their countries of birth (most adult immigrants are unlikely to retain childhood vaccination records). Information on diversity of BCG policies between countries and across time may be helpful for better interpretation of TB diagnostics as well as design of new TB vaccines. To our knowledge, there is no single, comprehen- sive, searchable database of BCG policies and practices (past and present) across the world. We developed the ‘‘BCG World Atlas: A Database of Global BCG Vaccination Policies and Practices,’’ a database containing BCG information from each country across all world regions (http:// www.bcgatlas.org/). Figure 1 shows the homepage of the Atlas. To make this resource practical and useful for clinicians, public health practitioners, and researchers alike, our database captures both present and previous policy and practices within a country, as well as any applicable changes. Detailed information on past and present BCG vaccination policies and practices were collected from as many countries as possible by one of three methods. First, short respondent-completed questionnaires were sent out to at least two individuals in each country. Questionnaires were sent to experts in TB research, TB control programs, or public health/vaccination programs. Whenever possible, an attempt was made to collect two completed questionnaires from each country in order to validate the data. Questionnaires were available in English, French, and Spanish, and were designed to capture both current policy and actual BCG practices, as well any applicable changes that had occurred over the last 25 years. Detailed questions asked for information concerning past as well as current practices, the timing and nature of changes to the policies and or practices, repeat, multiple, or booster shots, information concerning tuberculin skin testing in conjunction with BCG vaccination, influence of HIV on the decision to vaccinate, and vaccine strain differences. A total of 89 completed questionnaires were received over a 2-year period. Second, data were abstracted from published papers, reports, and available government policy documents retrieved through literature searches on PubMed and via the World Wide Web. Third, we used immunization data available from the World Health Organization Vaccine Preventable Diseases Monitoring System (. int/immunization_monitoring/en/global- summary/ScheduleSelect.cfm), which provide basic information on all vaccines currently in use in each country [13]. Based on the data generated by all methods, countries were grouped into three main categories: A), the country currently recommends universal BCG vaccination at a certain age; B), the country used to recommend universal BCG vaccination but currently does not; or category C), BCG vaccination is recommended only for selected high-risk groups or was never recommended. The beta version of the Atlas went live in the fall of 2008 with completed questionnaires on BCG vaccination from 62 countries. Since that time, more data have been added and several improve- ments have been made. As of October 2010, we have collected data concerning BCG vaccination policies and practices for 180 of 209 (86%) countries worldwide that we approached. The database is available as an interactive Web site at . bcgatlas.org/, where information for a particular country’s BCG policy, along with its estimated World Health Organization (WHO) TB incidence statistics, can be viewed alongside a graphical map. Among the 180 countries with available data, 157 countries currently recommend universal BCG vaccination, while the remaining 23 countries have either stopped BCG vaccination (due to a reduction in TB incidence), or never recommended mass BCG immunization and instead favored selective vaccination of ‘‘at risk’’ groups (Figure 2). Complete questionnaire data were available for 77 countries, while remaining data were extracted from published sources. Many countries began BCG vaccination programs in the 1940s–1980s, though some countries such as Romania and Uzbekistan report vaccination campaigns as early as 1928 and 1937, respectively, while some sub-Saharan African nations such as Nigeria and Sierra Leone only began BCG vaccinations in 1991 and 1990. Nine countries have ceased universal BCG vaccination programs; Spain and Denmark were among the first, stopping in 1981 and 1986, respectively, while Austria and Ger- many had stopped by 1990 and 1998. The remaining countries, including the Isle of Man, Slovenia, UK, Finland, and France, all ceased their BCG vaccination campaigns between 2005 and 2007. While these countries may have ceased mass universal vaccination programs, many do continue to provide BCG vaccination selectively to high-risk individuals, including those involved in high TB risk occupa- tions and/or travel, and infants born into high TB risk environments. We identified 49 countries that reported changes to their BCG vaccination policy in the past 20 years. Twenty-seven countries reported major changes to their BCG policy within the last 10 years. Of particular interest is the large number ( n = 33) of countries that had multiple vaccination programs in the past, but have since ceased revaccination, and now use a single BCG vaccination schedule (Table 1). These revaccination policy changes were as recent as 2007. Sixteen countries continue to give an additional BCG vaccination after the initial BCG, known as a booster vaccination (Table 2), while Kazakhstan, Belarus, Uzbekistan, and Turkmenistan continue to recommend three BCG vaccinations, with the third given between the ages of 12 and 15. Multiple revaccinations may lead to a delayed hypersensitivity reaction also known as the Koch response (phenome- non), where a person previously infected with M. tuberculosis is reinfected intracu- taneously, resulting in a local inflamma- tory reaction marked by necrotic lesions that develops rapidly and heals quickly [14]. Changes in vaccine strain were the most frequent type of change reported ( n = 42), and many countries have employed several strains over the course of their BCG vaccination program’s existence, with countries reporting strain changes as recently as 2008. Additional variations in BCG vaccination administration are seen across countries. Currently, eight countries recommend TST post–BCG vaccination, and two other countries had this policy but have since ceased. Estimated national BCG coverage ranged from 70% to 100%; however, frequently these estimates were not available or were several years out of date. Finally, 19 countries that did not recommend universal BCG vaccination did report BCG vaccination for certain at-risk groups, most frequently health care workers and infants living in high-risk TB settings. These variations highlight the impor- tance of mapping these differences across regions both for clinical purposes and research. The Atlas is an interactive Web site that allows users to select and view information concerning a country’s past and current BCG vaccination policy either by clicking on an interactive map or by selecting the country of interest from a drop-down list (Figure 1). The Web site is available to the public and is free of charge. Over the past year (during its beta phase), we have recorded over 6,000 visits to the site, with a steady increase in traffic over time. While novel diagnostics have been developed for latent TB infection (LTBI) [15–19], the TST continues to be the most widely used diagnostic test worldwide [20]. False positives can occur in BCG-vaccinated individuals, complicating interpretation of test results [21]. However, research suggests the timing of vaccination plays an important role [9,10]; in a meta- analysis, Farhat et al. found BCG vaccination at infancy has only a minimal effect on TST specificity, particularly if the TST is done more than 10 years after the BCG was administered, whereas BCG later in life or if given more than once led to more frequent, larger, and pronounced TST reactions [21]. The Atlas may help clinicians interpret TST by providing the information necessary to assess whether the TST is a valid diagnostic tool in a particular patient, or when alternative diagnostics may be preferable. Newly available interferon-gamma release assays (IGRAs) are more specific than TST because they are not affected by previous BCG vaccination [19,22]. Recent meta-analyses show that the specificity of IGRA is high in all populations, and will be of greatest utility in BCG-vaccinated populations [19,23]. For example, TST may be less specific for LTBI in Japan [24,25] (Figure 1) or other countries that revaccinate with BCG or have recently ceased revaccination programs (Table 1); in these settings, IGRAs may be more specific than the TST. Conversely, ...