Skip Navigation
Johns Hopkins Bloomberg School of Public Health
Bookmark and Share


PERCH logo


PERCH Documents

Study Details

Publications and Presentations

Clinical Standardization

Latest Updates



PERCH: Pneumonia Etiology Research for Child Health

The PERCH (Pneumonia Etiology Research for Child Health) project is a multi-country, case-control study to determine the etiology of and risk factors for severe and very severe pneumonia in children 1-59 months of age. Laboratory techniques that have remained largely unchanged for many decades and analytic methods that are not able to address the challenges inherent in the data have limited our understanding of pneumonia etiology. By applying modern diagnostics with standardized methods across all sites, PERCH will contribute new, detailed information to inform the development of new vaccines and treatment approaches.

Enrollment of cases and controls ended in January 2014. Publications on the methods and on aspects of the results are now published. The main findings will be published in 2017. Stay tuned!

Why was PERCH needed?

Many organisms can cause pneumonia. Prior to PERCH, bacteria such as Haemophilus influenzae type b (Hib) and pneumococcus were estimated to cause more than 50% of pneumonia deaths globally in children under 5 years of age (1). Viruses and fungi are also known to be causes of pneumonia infections (2).

Three major epidemiological changes have taken place since the formative pneumonia etiology studies of the 1980s making a new examination of pneumonia etiology essential (Scott 2008; Scott and English 2008). First, pneumococcal and Hib conjugate vaccination are now routinely used in 80% and 100% respectively of all low-income countries causing the underlying evidence base on etiology which was foundational for the existing treatment algorithms to be considered obsolete. Second, HIV-infection, which was largely not yet present in the populations of studies in the 1980s, has been influencing both the frequency of pneumonia and the distribution of pathogens causing it. Third, substantial changes in nutrition, urbanization, and access to health care have also likely modified the agents and the natural history of infection. Without a contemporary  pneumonia etiology evidence base representative of the current and future epidemiologic setting, our treatment algorithms will be ineffective and we will miss opportunities for prevention with new vaccines, biologics, or other strategies.

The strategy for achieving the goal of sufficient, appropriate information on pneumonia etiology to direct prevention and treatment efforts is complicated by various epidemiologic and microbiologic factors. First, our ability to definitely diagnose pneumonia, an infection of the lung tissue, is limited because the signs and symptoms are not specific to the diagnosis.  Even when we have relatively high confidence in the diagnosis, the condition ranges in severity and outcome. Second, the pathogens causing pneumonia and the distributions of those pathogens vary according to severity of the episode. The distribution of pathogens causing mild cases of pneumonia is not the same as that of pathogens causing fatal episodes as evidenced by Hib and PCV vaccine probe studies, as well as studies described above of the etiologic distribution of lung aspirate studies. This suggests either that some pneumonia pathogens are inherently more virulent or that different pathogens contribute to the pathogenesis of pneumonia at different stages of severity. Third, obtaining biologic samples for etiologic testing from the site of infection (the lung) is generally not possible. Fourth, the pathogens causing pneumonia are also commonly observed in the respiratory tracts of humans who do not have pneumonia, making the mere presence of the organism in a child with pneumonia difficult to interpret as causally related. Finally, cases of pneumonia are commonly the consequence of infection by more than one infectious agent either in parallel or in series.

A consequence of these features is that identification of a pathogen in a case of pneumonia does not necessarily imply that it is the cause of the illness, and, conversely, failure to identify a pathogen does not necessarily imply it is unrelated (Murdoch, O'Brien et al. 2009). To establish an etiologic diagnosis it is necessary to consider multiple possible etiologies (exposures) as the cause of any one pneumonia episode. Case-control studies are the design of choice when studying an outcome and its relationship to multiple exposures or risk factors, but this methodology has limits that include among others an inability to distinguish the causal chain of multiple pathogen infections. As such, our approach to estimating the etiologies of hospitalized pneumonia was to conduct a case-control study using modern diagnostics, testing multiple specimens, and determining the proportionate contribution of each using appropriate and innovative analytic approaches and statistical methods designed to address wherever possible the fundamental needs of pneumonia etiology data.

PERCH Study Sites and Collaborators

The PERCH project is a large, multi-country, case-control study of severe and very severe pneumonia in hospitalized children under five years of age. Nine PERCH sites in seven countries were selected for the study on the basis that they represented areas where most of the severe pneumonia cases in children were likely to occur in 2015 and where Hib and PCV vaccines were already implemented in most sites.

PERCH Sample Size: Number of Cases


Updated: January 2014

Collaborating institutions include the University of Witswatersrand in Johannesburg, South Africa, KEMRI Wellcome Trust Research Programme in Kilifi, Kenya, the University of Maryland’s Center for Vaccine Development in Bamako, Mali, the Thailand Ministry of Public Health – United States Centers for Disease Control and Prevention Collaboration’s at two sites in rural Thailand, Boston University at the University Teaching Hospital of Lusaka, Zambia, the Medical Research Council Unit in Basse, The Gambia and the International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDR, B) at two sites in Bangaldesh. Laboratory methods design and standardization as well as testing oversight was led by the University of Otago and Canterbury Health Laboratories in Christchurch, New Zealand with clinical field sites conducting the majority of lab testing on site.


  1. O'Brien K, Wolfoson L, Watt J, et al., Burden of Disease caused by Streptococcus pneumoniae in children younger than 5 years: global estimates. Lancet. 2009; 374:893-902.
  2. Mandell Lionel A, Wunderink Richard. Pneumonia. In: Fauci AS, Braunwald E, Kasper DL, Hauser SL, Longo DL, Jameson JL, Loscalzo J. Harrison's Principles of Internal Medicine (17th ed.) New York: McGraw-Hill, 2008. Accessed September 7, 2009.