PUBLIC HEALTH CLASSICS
The global epidemiology of childhood pneumonia 20 years on
J Anthony G Scott1
KEMRI Wellcome Trust Collaborative Research Programme, Centre for Geographic Medicine Research Coast, PO Box 230, Kilifi 80108, Kenya
Any reflection on history, even as recent as the past 20 years, invites a humble re-evaluation of the myth of human progress. In public health, progress has been made; certainly the number of children who die each year has declined progressively. However, rereading the work of scientists who investigated the major cause of death in childhood, acute respiratory tract infection (ARI), in the 1980s evokes an uncanny resonance with present-day concerns.
At that time, the attention of the global public health community was on oral rehydration therapy, universal immunization, promotion of breastfeeding and the use of growth monitoring charts.2 Lower respiratory tract infections (LRTIs) attracted relatively little attention. With considerable perspicacity, the Board of Science and Technology for International Development (BOSTID) at the National Academy of Sciences, United States of America, defined ARI as one of six priority areas for research funding in 1983 and convened an international ARI meeting. The participants identified three prerequisites for relevant research:
1. Studies should be undertaken in a wide variety of countries to give full geographical representation to the children of the developing world and they should be standardized to facilitate international comparisons.
2. The etiology of ARI should be investigated first because it would be essential for later research on prevention and case management.
3. The international research group should be coordinated by a centre that could provide technical assistance and quality control, and could foster active collaboration between investigators.3
Over the next 5 years, BOSTID undertook such a project, involving investigators from 12 sites who met on an annual basis to agree on clinical definitions, laboratory methods, study designs and analysis plans. The results of the programme were published in 1990 in a supplement of the Reviews of Infectious Diseases. The supplement illustrates the wide diversity of research activities in the programme from community-based epidemiology to laboratory comparisons of antigen detection methods, evaluations of recent antibiotic exposure and pathological studies of postmortem specimens. The anchor of the supplement is the paper "The epidemiology of acute respiratory tract infection in young children: comparison of findings from several developing countries" by Beatrice Selwyn on behalf of the BOSTID investigator group, reporting a standardized analysis of the epidemiology of ARI in young children from 10 sites.1
It is a paper of truly astonishing ambition, combining 16 studies of upper and lower respiratory tract infections in both community- and hospital-based settings. It examines incidence, prevalence, duration, case-fatality and the effects of age, sex and season on the patterns of disease. It describes bacterial and viral etiology and interrogates the clinical signs of respiratory tract infections to define these diseases more accurately. It evaluates risk factors for respiratory tract infections across several sites, including mother's age and education, weight-for-age percentiles, and crowding and smoking in the household. The hospital-based studies alone reported nearly 4000 episodes of ARI and the eight community-based cohort studies each included reports of between 8000 to 93 000 home visits.
The key findings of the analysis were:
1. The incidence of LRTIs varied forty-fold across the sites but the incidence of all respiratory tract infections (upper and lower combined) was remarkably consistent.
2. The incidence and case-fatality of LRTIs were consistently higher among younger children aged < 18 months.
3. The prevalence of ARI symptoms, at any one time, was 2240%.
4. Viruses caused more episodes of ARI than did bacteria.
Respiratory syncytial virus was the commonest viral cause of LRTIs and Streptococcus pneumoniae (pneumococcus) and Haemophilus influenzae were the commonest bacterial causes. A significant fraction (one-third in one hospital) of all H. influenzae infections were nontypeable.
Not all of the insights of the BOSTID research group could be summarized in a single paper or even a supplement. The programme provided an intellectual forum in which many lines of enquiry were distilled.2 The complete lack of understanding of how or why children die from pneumonia was identified as a critical future research question. The pathogenesis of ARI, including the complex synergism between viral and bacterial pathogens, was a second significant area. The investigators believed that understanding the mechanisms that controlled the magnitude and selectivity of the human inflammatory response would offer practical opportunities to influence disease outcome. This insight would need to be augmented by knowledge of the modulating effects of nutritional status and immune deficiency. The role of access to health care, and the quality of that care, in the outcome of disease was a third significant area of research that was likely to be fruitful.2 The list is strikingly similar to an evaluation of the research required to tackle pneumonia today.4 With the exception of oxygen therapy for severe pneumonia and zinc supplementation to prevent disease, there has been little clinical amelioration of pneumonia through developments in clinical science in 20 years.
The reasons for this apparent neglect probably lie with subsequent advances in public health policy and vaccine development. In 1991, WHO formulated its case-management strategy for pneumonia. The strategy was driven by bacteriological studies, particularly those incorporating lung aspirates, which identified pneumococcus and H. influenzae type b (Hib) as the dominant causes of severe and fatal pneumonia. These infections were treatable with cheap and widely available antibiotics. The policy was designed to identify patients with the syndrome of pneumonia at an early stage, often without a doctor's examination, and to initiate treatment with life-saving antibiotics. Studies undertaken around this time indicated that the case-management approach was capable of reducing all-cause mortality in children aged < 5 years by 2024%.5 Access to antibiotics has improved generally over the past 15 years but the fact that two million children still die of pneumonia each year suggests that the potential of this policy was never fully realized.
Focus on case management was followed, in the second half of the 1990s, by enthusiasm for new vaccines. A protein-conjugated Hib vaccine was shown to prevent approximately 20% of radiologically confirmed pneumonia in a trial in children in the Gambia. This seeded the idea that Hib vaccine, and possibly the pneumococcal conjugate vaccines that were in development at that time, could be deployed in low-income countries to prevent a significant fraction of the burden of childhood pneumonia. The taxing questions were how to finance and distribute these relatively expensive products in countries with inadequate resources.
The task was taken up by the GAVI Alliance (formerly known as the Global Alliance of Vaccines and Immunization), which has funded Hib vaccine introduction since 2001 and will begin supporting the introduction of the pneumococcal vaccine in a few developing countries in 2008. Experience with Hib vaccine suggests that the introduction process is slow but the fact that it has started challenges us to consider the management of a spectrum of pneumonia pathogens that may no longer be dominated by the easy targets of pneumococcus and Hib.6
Over this period the epidemiology and etiology of pneumonia have also evolved, particularly as a consequence of HIV. Unusual pathogens such as Pneumocystis jiroveci and cytomegalovirus have a prominent place in the etiology of pneumonia among children with HIV. Mycobacterium tuberculosis is a common cause of presentation with pneumonia in areas of high HIV prevalence regardless of HIV status.7 With advances in molecular diagnostic tools, we have also identified novel pathogens, such as human metapneumovirus and new human coronaviruses, in immunocompetent children with respiratory disease.
The process initiated by the BOSTID studies has therefore developed a new relevance two decades later: interest in pneumonia research is currently being rekindled by both scientists and funders.8 In her introduction to the BOSTID supplement, Judith Bale reflects: "With all the complexities of ARI, it is unrealistic to search for a 'magic bullet'. Research must include a focus on basic understanding of ARI, particularly the factors leading to severe disease." Given the complexities of the problem, a comprehensive and accurate description of the epidemiology and etiology will once again become the foundation of pneumonia research. As we rebuild a global network of pneumonia research sites, we might ponder how we failed to sustain the investment of the BOSTID initiative. Childhood pneumonia has remained the dominant public health problem in the developing world but we have not cultivated local research capacity in pneumonia.
What can be gleaned from the BOSTID studies to optimize a new pneumonia research network? The paper by Selwyn et al.1 was prescient, courageous and comprehensive but it also revealed some of the difficulties in creating an integrated global description of respiratory tract infections. The inclusion of upper respiratory tract infection (URTI) affirmed its biological connection with LRTI but also undermined the public health impact of the studies, given the generally benign perception of URTI. Site selection gave preference to underprivileged populations but, because the sites had to be close to competent laboratories (which are rare in low-income countries), the representation of the developing world was uneven. For example, five out of the 12 sites were located in Central and South America. A standardized case-definition is essential for international comparisons but most of the BOSTID investigators amended the standardized definitions, thus producing, in some cases, exceptional incidence results. The failure to obtain lung aspirate material reduced the sensitivity of the study to bacterial causes of pneumonia.
These factors do not detract from what was a Herculean task performed in an era when global networks were uncommon and international communications were challenging. However, the scientific community of today needs to regenerate pneumonia research in developing countries and the first step will be to learn from the difficulties encountered by this pioneering programme. As we take up the task, we are indebted to the BOSTID group for this far-sighted publication. 
Competing interests: None declared.
References
1. Selwyn BJ on behalf of the coordinated data group of BOSTID researchers. The epidemiology of acute respiratory tract infection in young children: comparison of findings from several developing countries. Rev Infect Dis 1990;12 Suppl. 8;S870-88. PMID:2270410
2. Grant JP. The state of the world's children 1982-3. Oxford University Press: 1983. p. 141.
3. Bale JR. Creation of a research program to determine the etiology and epidemiology of acute respiratory tract infection among children in developing countries. Rev Infect Dis 1990;12 Suppl 8;S861-6. PMID:2270408
4. Scott JA, Brooks WA, Peiris JS, Holtzman D, Mulholland EK. Pneumonia research to reduce childhood mortality in the developing world. J Clin Invest 2008;118:1291-300. PMID:18382741 doi:10.1172/JCI33947
5. Sazawal S, Black RE. Effect of pneumonia case management on mortality in neonates, infants, and preschool children: a meta-analysis of community-based trials. Lancet Infect Dis 2003;3:547-56. PMID:12954560 doi:10.1016/S1473-3099(03)00737-0
6. Scott JA, English M. What are the implications for childhood pneumonia of successfully introducing Hib and pneumococcal vaccines in developing countries. PLoS Med 2008;5:e86. doi:10.1371/journal.pmed.0050086
7. McNally LM, Jeena PM, Gajee K, Thula SA, Sturm AW, Cassol S, et al. Effect of age, polymicrobial disease, and maternal HIV status on treatment response and cause of severe pneumonia in South African children: a prospective descriptive study. Lancet 2007;369:1440-51. PMID:17467514 doi:10.1016/S0140-6736(07)60670-9
8. Greenwood BM, Weber MW, Mulholland K. Childhood pneumonia preventing the world's biggest killer of children. Bull World Health Organ 2007;85:502-3. PMID:17768493
Letters
Please visit http://www.who.int/bulletin/volumes/86/6/en/index.html to read the following letters received in response to Bulletin papers:
A way of measuring poverty that could further a change for the better, by Hermann Feldmeier & Ingela Krantz,
responding to:
Fosu AK. Poverty and development. Bull World Health Organ 2007;85:734. PMID:18038047
Contraception is the best kept secret for prevention of mother-to-child HIV transmission, by Tricia Petruney, Elizabeth Robinson, Heidi Reynolds, Rose Wilcher & Willard Cates,
responding to:
Stringer EM, Chi BH, Chintu N, Creek TL, Ekouevi DK, Coetzee D, et al. Monitoring effectiveness of programmes to prevent mother-to-child HIV transmission in lower-income countries. Bull World Health Organ 2008;86:57-62. PMID:18235891 doi:10.2471/BLT.07.043117
Access to medication: key to achieving treatment goals, by Hevertton LBS Santos & Nelson Rosario,
responding to:
Mendis S, Fukino K, Cameron A, Laing R, Filipe A Jr, Khatib O, et al. The availability and affordability of selected essential medicines for chronic diseases in six low- and middle-income countries. Bull World Health Organ 2007;85:279-88. PMID:17546309 doi:10.2471/BLT.06.033647
(Submitted: 10 March 2008 Accepted: 10 March 2008)
This section looks back on a ground-breaking contribution to public health, reproduces an extract of the original and adds a commentary on its significance from a modern perspective. To complement the theme of this month's issue, Anthony Scott reviews a paper by BJ Selwyn that was published in 1990 in a supplement to Reviews of infectious diseases.1 Pages S870, S886 and S887 are reproduced with kind permission of Chicago University Press at: http://www.who.int/bulletin/volumes/86/6/08-052753/en/index.html
1 Correspondence to J Anthony G Scott (e-mail: ascott@ikilifi.net).
doi:10.2471/BLT.08.052753
