Map the gap: missing children with drug-resistant tuberculosis

Abstract

Background: The lack of published information about children with multidrug-resistant tuberculosis (MDR-TB) is an obstacle to efforts to advocate for better diagnostics and treatment.

Objective: To describe the lack of recognition in the published literature of MDR-TB and extensively drug-resistant TB (XDR-TB) in children.

Design: We conducted a systematic search of the literature published in countries that reported any MDR- or XDR-TB case by 2012 to identify MDR- or XDR-TB cases in adults and in children.

Results: Of 184 countries and territories that reported any case of MDR-TB during 2005–2012, we identified adult MDR-TB cases in the published literature in 143 (78%) countries and pediatric MDR-TB cases in 78 (42%) countries. Of the 92 countries that reported any case of XDR-TB, we identified adult XDR-TB cases in the published literature in 55 (60%) countries and pediatric XDR-TB cases for 9 (10%) countries.

Conclusion: The absence of publications documenting child MDR- and XDR-TB cases in settings where MDR- and XDR-TB in adults have been reported indicates both exclusion of childhood disease from the public discourse on drug-resistant TB and likely underdetection of sick children. Our results highlight a large-scale lack of awareness about children with MDR- and XDR-TB.

Multidrug-resistant tuberculosis (MDR-TB, defined as TB resistant to at least both isoniazid and rifampin), among children is an unrecognized epidemic. Like drug-susceptible TB, MDR-TB is curable, but treatment requires the use of longer regimens with more toxic drugs, few of which are available in child-friendly formulations.¹ Although an estimated 32 000 children fall sick each year with MDR-TB,² few of these are diagnosed and treated. An analysis of age-disaggregated surveillance data collected in 85 countries during 1994–2011 found that 50 (59%) of these countries reported no child MDR-TB cases,³ although many of them did report cases of MDR-TB, presumably in adults, during this period.⁴ The microbiological diagnosis of drug-resistant TB in children is complicated by insensitive diagnostic tests, limited diagnostic capacity in many countries with high TB burdens⁵ and the inability of young children to expectorate sputum.⁶ These diagnostic challenges, as well as historic misperceptions that children are not likely to be important contributors to or victims of the global TB epidemic,⁷ have contributed to the invisibility of the MDR-TB epidemic among children.⁸

The lack of published evidence documenting numbers of children treated for MDR-TB, as well as treatment practices and outcomes, is an obstacle to efforts to advocate for better diagnostics, treatments, and policies for children with MDR-TB. As publicly available data on numbers of MDR-TB cases reported by different countries are not disaggregated by age,⁴ it is impossible to determine how many children are diagnosed with MDR-TB each year. A systematic review of studies reporting outcomes among children treated for MDR-TB found only eight studies, comprising a total of 315 children.⁹ With such sparse published evidence on both the magnitude of the problem and its potential solutions, it is difficult to convince companies to invest in new diagnostics for drug-resistant disease in children, investigators to include children in clinical trials of second-line regimens and governments to reform policies that limit access to effective treatment for children with MDR-TB.

As global TB policies tend to focus on adults, there is a profound lack of awareness about the MDR-TB epidemic among children. In the present study, we sought to describe the current state of awareness about children with MDR-TB by comparing documentation of child and adult MDR-TB cases in the published literature.

METHODS

The numbers of MDR-TB cases reported each year by governments of individual countries and territories (referred to from here on as ‘countries’) are publicly available through the World Health Organization (WHO).⁴ Based on these data, we compiled a list of countries that reported at least one notified MDR-TB case during 2005–2012.⁴ We then used a systematic search of the published literature to identify child and adult MDR-TB cases in each country. We repeated this process to identify child and adult cases of extensively drug-resistant TB (XDR-TB), the subset of MDR-TB strains with additional resistance to fluoroquinolones and second-line injectable agents.

For each country, we attempted to identify a publication documenting at least one MDR-TB case in a child aged ⩽14 years, and one documenting at least one MDR-TB case in an adult aged >14 years (Figure 1). In our search for pediatric MDR-TB cases, we first referred to two publications documenting country-specific reports of MDR-TB in children: one based on WHO surveillance data³ and the other a systematic review.² For countries in which no pediatric MDR-TB cases were indicated in either of these publications, we systematically searched for publications published through 1 July 2014 in PubMed, EMBASE, LILACS, Web of Science, BIOSIS and WHO regional databases. Together, these databases include citations for a variety of published literature, including research publications, case reports, editorials and conference abstracts.

FIGURE 1.

Search strategy for identifying publications, by country, for adult and pediatric MDR- and XDR-TB cases among countries that reported MDR- and XDR-TB cases. MDR-TB = multidrug-resistant tuberculosis; XDR-TB = extensively drug-resistant TB.

No reviews or surveillance summaries explicitly focused on adults with MDR-TB. We referred to the aforementioned systematic review of MDR-TB in children,² which also documented adult cases in some countries. For the remaining countries, we systematically searched the published literature for publications documenting adult MDR-TB cases. We used an analogous process to search for XDR-TB cases in both adults and children in all countries that had reported at least one XDR-TB case by 2012.¹⁰

We used search terms that included combinations intended to capture publications documenting MDR- and XDR-TB cases (e.g., ‘multidrug resistant tuberculosis,’ ‘MDR-TB,’ ‘drug-resistant TB’) and country names. In our search for child cases, we also included terms intended to capture children (e.g., ‘child*,’ ‘pediatric’). The complete search strategy is documented in Appendix Table A.1. We did not restrict publications on the basis of language. For publications in English and Spanish, we contacted authors by e-mail for additional information if the publication identified MDR- or XDR-TB cases but we were unable to determine the age of subjects.

For each of the four categories of cases (child MDR-TB, adult MDR-TB, child XDR-TB, adult XDR-TB), we considered a single publication sufficient to document the presence of that type of case in a given country. If our search yielded multiple published reports for a country, we recorded only the most recent English-language report, or the most recent report in any language if no English-language report was found.

No ethics approval was required for this study, as the data used were publicly available and contained no personal identifying information.

RESULTS

Of the 216 national and territorial governments that reported TB case data, 184 (85%) reported at least one case of MDR-TB during 2005–2012.⁴ Through our literature searches, we found publications documenting adult MDR-TB cases in 143 (78%) of the 184 countries (Table and Appendix Table A.2). By contrast, we found publications documenting pediatric MDR-TB cases in 78 (42%) of these countries. We were thus unable to identify any pediatric MDR-TB cases for 45% of the countries for which we were able to identify adult MDR-TB cases through our systematic literature search. There were no countries with publications documenting MDR-TB cases in children but not in adults.

TABLE.

Countries and territories with publications documenting MDR-TB cases in adults and children ^*

Figure 2 shows the countries categorized according to whether any MDR-TB case was reported during 2005–2012, whether we identified any adult MDR-TB case through our literature search and whether we identified any pediatric MDR-TB case through our literature search. The proportion of countries for which we were able to find publications documenting adult MDR-TB cases but not pediatric MDR-TB cases was highest for the WHO African Region (69%) and lowest for the European region (19%) (Table). We found publications documenting adult MDR-TB cases for all of the 27 countries classified as having a high MDR-TB burden,¹⁰ but found no publications documenting pediatric MDR-TB cases for 7 (26%) of these countries.

FIGURE 2.

Adult vs. child MDR-TB cases in the published literature: documentation of adult vs. child MDR-TB cases in the published literature among countries reporting ⩾1 case of MDR-TB during 2005–2012. Note that some smaller countries and territories are not depicted. MDR-TB = multidrug-resistant tuberculosis.

By 2012, XDR-TB cases had been reported in 92 countries.¹⁰ We identified publications documenting adult XDR-TB cases for 55 (60%) of these countries, and publications documenting pediatric XDR-TB cases for 9 (10%) countries. We found no pediatric XDR-TB cases for 84% of the countries for which we identified adult XDR-TB cases through our literature search. Figure 3 shows the countries categorized according to whether any case of XDR-TB was reported by 2012, whether we identified any adult XDR-TB case through our literature search and whether we identified any pediatric XDR-TB case through our literature search.

FIGURE 3.

Adult vs. child XDR-TB cases in the published literature: documentation of adult vs. child XDR-TB cases in the published literature among countries reporting ⩾1 case of XDR-TB by 2012. Note that some smaller countries and territories are not depicted. XDR-TB = extensively drug-resistant tuberculosis.

DISCUSSION

In almost every household where an adult has TB, children, who are even more susceptible than adults to developing TB disease,¹¹ are exposed. However, although we found documentation of MDR-TB cases in adults for over three quarters of the countries whose governments reported any case of MDR-TB by 2012 through systematic searches of published literature, we were unable to find published reports of MDR-TB in children for over half of these countries. We were unable even to find published reports of MDR-TB cases in children for a quarter of the countries with a high MDR-TB burden. The size of the discrepancy varied by geographic region, and was even greater in our search for XDR-TB cases. Our results are consistent with large-scale deficient disease awareness of children with MDR- and XDR-TB. This deficiency has major implications for national and global TB policies, as children (age ⩽14 years) make up approximately 25% of the global population and may comprise 20–40% of persons with TB in high TB incidence settings.¹²

The pronounced absence of childhood MDR-TB cases in the literature is worrying, as the published literature is an important forum for the dissemination of information, the discussion of public health policy, and the testing of new ideas. In a time when MDR- and XDR-TB are gaining worldwide attention, the absence of child cases in the published literature can only perpetuate the invisibility of children already suffering from these forms of TB. Of course, there may be reasons for this absence: research publications often exclude children, and publications are generally not a priority for resource-limited TB programs struggling to provide treatment. However, as programs strive to improve the diagnosis and treatment of children with MDR- and XDR-TB, the lack of published evidence in these areas may prove an obstacle.

While the absence of cases in the published literature may not indicate an actual absence of diagnosed cases, the magnitude of the discrepancy between published accounts of adult cases and published accounts of child cases raises strong suspicions about underdetection. Given the large number of adults worldwide with untreated or inadequately treated MDR-TB,¹³ transmission to children in their households is inevitable, and child MDR-TB cases are to be expected wherever adult MDR-TB cases are reported.⁸ Recent estimates have suggested that each year 600 000 children worldwide require evaluation for household exposure to MDR-TB,⁸ and that 30 000–50 000 require treatment for MDR-TB.^2,8 However, many of these children are likely going undiagnosed. Although global data on children diagnosed with MDR- and XDR-TB are lacking, widespread underdetection of child MDR-TB cases is suggested by the finding that 59% of the countries from which age-disaggregated surveillance data on MDR-TB were available reported no child MDR-TB cases,³ although many of these countries did report MDR-TB cases, presumably in adults.⁴

The major limitation of the present analysis is that a review of published literature is unable to distinguish between the effect of underdetection of MDR- and XDR-TB in children and the effect of any publication disparities that may exist. Many factors could contribute to a publication bias against reports that include children with drug-resistant TB. For example, research studies often exclude children; analyses of routinely collected data on patients are frequently limited to patients with bacteriologically confirmed disease, which reduces the likelihood of children being included; and children are commonly excluded from drug resistance and prevalence surveys. In addition, countries with high TB burdens (and likely higher relative burdens of childhood TB cases¹²) are often poorly represented in the published literature; the ratio of TB publications to TB cases is orders of magnitude higher for Western Europe, the United States and Canada, than for Asia, Eastern Europe, Africa and Latin America.¹⁴ Many countries for which we found no publications documenting any MDR-TB case were countries for which few TB publications existed. However, true underdetection of MDR-TB in children is also likely in many of these countries given the limitations in the current diagnostic capacities of their TB programs. Public availability of age-group-specific data on reported MDR- and XDR-TB cases would allow a more careful assessment of the relative contributions of underdetection and publication bias to our findings.

Another limitation is that our analysis only included publications that were indexed in a systematically searchable format; although available databases cover a variety of publications, including conference abstracts, bulletins and perspective pieces, some types of publications, such as advocacy materials and national tuberculosis program reports, are generally not included. While information on children with MDR- and XDR-TB may appear in these other types of publications, we had no way of systematically searching them. A recent advocacy effort to collect information about the experiences of children with drug-resistant TB found stories of children with MDR-TB from 10 of the countries for which we could not find any publications in our search.¹⁵

However, despite the limitations inherent in using published literature as an indicator of case detection, our results add weight to the suspicion of underdetection of MDR- and XDR-TB in children. Reasons for this underdetection are varied, but multiple steps can be taken to overcome existing obstacles to the recognition of child cases of drug-resistant disease. First, providers require more training to recognize the manifestations of TB disease in children, which can be diverse and non-specific.¹⁶ In addition, as children frequently have paucibacillary disease and cannot produce sputum, programs should employ child-focused methods of specimen collection and diagnostic technologies that are more sensitive than smear microscopy.¹⁶ Even with these methods, however, cultured isolates for drug susceptibility testing (DST) will be unavailable for the majority of sick children.⁶ Both individual providers and TB programs should therefore presumptively treat sick children who meet clinical criteria for TB disease and drug resistance in the absence of bacteriologic confirmation.^6,17 Finally, household contact investigations of adult MDR- and XDR-TB patients should be carried out to promptly identify and treat children with drug-resistant TB.⁸

Underdetection of childhood drug-resistant TB cases should be an indicator of weaknesses in TB programs, such as the failure to perform contact investigations or the underutilization of sensitive diagnostic techniques. To rigorously quantify underdetection, accurate counting of children with MDR- and XDR-TB will be necessary. If programs count and report only bacteriologically confirmed cases of drug-resistant TB, a substantial proportion of children with MDR- and XDR-TB will not be included, as the majority of children with TB will not have a bacterial culture available for DST.⁶ Surveillance systems should therefore allow for the reporting of probable MDR-TB cases in children, with diagnosis based on clinical evidence of disease and exposure to an MDR- or XDR-TB case.¹⁸ Furthermore, universal reporting of MDR- and XDR-TB cases by age group¹⁹ and public accessibility of these surveillance data are necessary to broaden our understanding of the magnitude of the drug-resistant TB epidemic in children.

CONCLUSION

The absence of information about child MDR- and XDR-TB cases in countries where adult cases have been reported has major implications for national and global TB policies. Recognition of this large-scale deficiency in awareness should spur more systematic work to improve surveillance for and estimates of TB cases in children, which are necessary to allow programs to project unmet needs and allocate resources. In addition, it should serve as a warning that children with drug-resistant TB in these countries are going undiagnosed. Clinical awareness, aggressive methods of specimen collection, use of sensitive diagnostics and active case finding through contact investigation are all needed to ensure that children with TB—including those with MDR- and XDR-TB—are promptly recognized and treated.²⁰ Finally, policy makers, providers, investigators and advocates should all be aware that the discourse about how to end the toll of drug-resistant TB must include children to ensure that this vulnerable population is not neglected.

APPENDIX

TABLE A.1.

Terms used in systematic searches of published literature

TABLE A.2.

Table of identified references by country ^*

TABLE A.2.

(continued)

TABLE A.2.

(continued)

TABLE A.2.

(continued)