Introduction

Pneumonia is a leading cause of death in children worldwide and we need updated evidence about its current causes

Pneumonia sickens about 150 million children every year,1 Nair H, Simoes EAF, Rudan I, et al. Global and regional burden of hospital admissions for severe acute lower respiratory infections in young children in 2010: A systematic analysis. Lancet. 2013;381(9875):1380-1390. doi:10.1016/S0140-6736(12)61901-1 and hospitalizes at least 11 million.

About 900,000 children die of pneumonia annually—more children than HIV, TB, Zika, Ebola, and malaria combined.2 Liu L, Oza S, Hogan D, et al. Global, regional, and national causes of under-5 mortality in 2000-15: an updated systematic analysis with implications for the Sustainable Development Goals. Lancet 2016; 388: 3027–35.

The main causes of pneumonia today are not the same as they were in past decades, in part due to successful vaccines. H. influenzae and pneumococcal disease used to be the top causes of pneumonia but widespread use of effective vaccines to prevent these diseases is now common. We need updated evidence to understand which bacteria, viruses, and other germs are now causing the majority of severe pneumonia in children.

Because new vaccines take years to develop, today’s pneumonia research is necessary so children of the future can have an even greater opportunity to be healthy and thrive.

To collect this evidence, we conducted the Pneumonia Etiology Research for Child Health (PERCH) Study in settings representative of where severe childhood pneumonia is common.

When the PERCH study began, more than 12,000 Malian children died annually from pneumonia. (Photo: Jane Crawley, August 2012)
Thanks to scientific advocacy and support from Gavi, the Vaccine Alliance, pneumococcal conjugate vaccines are now being used in over 140 countries. (Photo: Adrian Brooks)
While millions of children’s lives have been saved in past decades thanks to advances in vaccines and access to antibiotic therapy, researchers are looking for ways to improve the prevention, diagnosis and treatment of pneumonia. (Photo: Jane Crawley)
Results

Causes of childhood pneumonia in PERCH

Over the course of two years in 7 countries, the PERCH study team enrolled and tested children hospitalized with pneumonia for the presence of over 30 viruses, bacteria and fungi.

Viruses caused most of the severe pneumonia cases (61%), and respiratory syncytial virus (RSV) was the leading pathogen (31%) at all sites.

Explore 10 of the top causes of severe pneumonia:

10 pathogens caused the vast majority (about 8 in 10 cases): RSV , RSV is a virus that can cause upper and lower respiratory tract illness. rhinovirus , Rhinovirus is the predominant cause of the “common cold.” human metapneumovirus (hMPV) , hMPV is a virus that can cause upper and lower respiratory tract illness. parainfluenza , Parainfluenza is a group of viruses that cause upper and lower respiratory tract illness and are unrelated to influenza (aka “the flu”). pneumococcus , Pneumococcus is a bacterium often found in the nose. It can cause serious illness besides pneumonia like meningitis, blood infections, and ear infections. Haemophilus influenzae , H. influenzae is a bacterium often found in the nose. It can cause serious illness besides pneumonia like meningitis, blood infections, and ear infections. tuberculosis (TB) , TB is a bacterium that mainly infects the lungs. Staphylococcus aureus (Staph) , Staph is a bacterium that is often found on skin and in the nose. Besides pneumonia, it can infect blood, skin, and soft tissue. influenza Influenza, aka “the flu,” is a group of viruses that can cause mild to severe respiratory illness. and Pneumocystis jirovecii . P. jirovecii is a yeast-like fungus that causes pneumonia.

Average in all 7 countries, by:
  • All ages and cases
  • Below age 1
  • Above age 1
  • Severe cases
  • Very severe cases
Average of all ages and severities, in:
  • The Gambia
  • Mali
  • Kenya
  • Zambia
  • South Africa
  • Bangladesh
  • Thailand
Note: Point estimates for the fraction of cases attributed to each pathogen include 95% credible intervals.

Average in all 7 countries, by:

Average of all ages and severities, in:

RSV is a virus that can cause upper and lower respiratory tract illness. Rhinovirus is the predominant cause of the “common cold.” hMPV is a virus that can cause upper and lower respiratory tract illness. Parainfluenza is a group of viruses that cause upper and lower respiratory tract illness and are unrelated to influenza (aka “the flu”). Pneumococcus is a bacterium often found in the nose. It can cause serious illness besides pneumonia like meningitis, blood infections, and ear infections. H. influenzae is a bacterium often found in the nose. It can cause serious illness besides pneumonia like meningitis, blood infections, and ear infections. TB is a bacterium that mainly infects the lungs. Staph is a bacterium that is often found on skin and in the nose. Besides pneumonia, it can infect blood, skin, and soft tissue. Influenza, aka “the flu,” is a group of viruses that can cause mild to severe respiratory illness. P. jirovecii is a yeast-like fungus that causes pneumonia.

Read the full results in The Lancet.
Methods

The PERCH study was the largest, most comprehensive pneumonia etiology study conducted in over three decades

Clinicians, governments, donors, and vaccine developers need new evidence on causes of pneumonia from countries where child mortality from pneumonia is still far too high in order to guide updates to treatment and prevention strategies.

The PERCH study used rigorously standardized clinical and laboratory methods and a novel data analysis method that allowed PERCH to identify the causes of pneumonia across the population and in each individual study child—this made PERCH different from any other study of pneumonia.

PERCH enrolled 4,232 children between 1 month and 5 years who were hospitalized with severe pneumonia. Researchers also included 5,325 children from the same communities who did not have severe pneumonia. The study was conducted in 7 countries in Africa and Asia, each country with differing characteristics that might influence the causes of pneumonia. H. influenzae type b vaccine was in routine use at all PERCH sites except Thailand, and pneumococcal conjugate vaccines were in routine use in The Gambia, Kenya, Mali and South Africa. See below to learn more about the study population and study sites.

Hundreds of clinicians and lab technicians developed cutting-edge skills through the PERCH study. (Photo: Jane Crawley, July 2011)
The PERCH team developed comprehensive quality indicators, data queries, and monitoring reports that were shared regularly during the study. (Photo: Jane Crawley, July 2011)

PERCH cases are a subset of all pneumonia cases

Children with pneumonia that resolved before progressing to severe pneumonia, or those who died before they reached the hospital were not included in the PERCH study.

Children enrolled into PERCH were hospitalized with severe or very severe pneumonia, with or without chest x-ray findings.

The proportion of children who died following enrollment varied by site and was up to 1 in 6 children in Zambia and Mali, despite receiving medical care.

Explore the 7 sites:

  • {country}, {years}

PERCH STUDY CASES

Number of cases

{value} out of 4,232

Age <1

0

Very severe pneumonia

13.6

Case fatality ratio

13.6
Site characteristics during PERCH study

Site location

Hib vaccine in use

PCV in use

HIV

Malaria

2012 U5 Mortality Rate*

*Per 1,000 live births. Applies to country. Source: United Nations Children’s Fund, State of the World’s Children 2014 (https://www.unicef.org/sowc2014/numbers/).
In Kenya, PERCH was nested within the Kilifi Health and Demographic Surveillance System, a 20-year-old surveillance system for invasive bacterial diseases and viral respiratory pathogens. (Photo: Jane Crawley, December 2011)
In The Gambia, PERCH was conducted at Basse Health Centre, the Upper River Region’s only major health center. (Photo: Jane Crawley, June 2011)
For those moving between PERCH’s sites in Dhaka and Matlab, Bangladesh, boat travel was often faster than vehicular travel over congested roads. (Photo: March 2011)

PERCH’s novel statistical package

It’s very hard to identify the causes of pneumonia in children because it is difficult to access the lung itself. Therefore, scientists and clinicians look for bacteria and viruses in samples from a child’s nose, blood, and other body fluids that may indicate what is infecting the lung.

Yet the utility of these samples alone is limited. Blood culture can only detect bacterial causes, and often yield nothing at all even when the child really does have bacterial pneumonia. Samples from the nose can be misleading since children tend to have many germs in the nose even when they are healthy, none of which may be causing disease in the lungs of children with pneumonia. Therefore, results must factor in findings from multiple samples.

PERCH devised a new way to combine evidence from multiple test results to estimate the causes of pneumonia. The Bayesian Analysis Kit for Etiology Research, or BAKER method, factors in the limitations and strengths of each laboratory test to estimate what’s causing pneumonia in each child. 3,  Wu Z, Deloria-Knoll M, Hammitt LL, Zeger SL. Partially latent class models for case-control studies of childhood pneumonia aetiology. J R Stat Soc Ser C Appl Stat 2016; 65: 97–114. 4,  Wu Z, Deloria-Knoll M, Zeger S. Nested Partially-Latent Class Models for Dependent Binary Data; Estimating Disease Etiology. Biostatistics 2017; 18: 200–13. 5,  Deloria Knoll M, Fu W, Shi Q, et al. Bayesian Estimation of Pneumonia Etiology: Epidemiologic Considerations and Applications to the Pneumonia Etiology Research for Child Health Study. Clin Infect Dis 2017; 64: S213–27. 6 Bayesian Analysis Kit for Etiology Research (BAKER), publicly available at https://github.com/zhenkewu/baker.

BAKER is now “open source” and being used to study other challenging questions like the causes of infection in newborns.

Chest x-rays can confirm a pneumonia diagnosis. In PERCH, chest x-rays were reviewed by a panel trained in WHO methodology. Lessons from PERCH's chest x-ray review process are helping others who study pneumonia. (Photo: Adrian Brooks)
The Bayesian Analysis Kit for Etiology Research (BAKER), an “open source” method developed by the PERCH team, overcame analytical barriers in etiology research.
Learning

The PERCH study will impact how we prevent, diagnose, and treat pneumonia in the 21st century

By focusing on the leading causes, governments, donors, researchers, and doctors can efficiently reduce child pneumonia illness and death.

(Photo: Jane Crawley, November 2011)

Prevent: Scaling up currently-available vaccines and developing new vaccines

PERCH showed that in the settings studied, viruses are more likely to cause severe pneumonia for a child in hospital than are bacteria. We need new vaccines and other products to address this viral burden, especially against the leading causes.

Vaccines and long-acting monoclonal antibodies against RSV are in the late stages of development. New vaccines against influenza, such as a universal flu vaccine that protects against all strains, are now in development. Scientists are also working on vaccines against parainfluenza virus and the Staph bacterium.

Diagnose: Developing new tools to identify the causes of pneumonia

New tools powered by findings from PERCH could help clinicians incorporate multiple test results to more accurately identify what has caused pneumonia for an individual child. Such individualized medicine, in turn, can enable better and smarter treatment of children with severe pneumonia.

Treat: Reducing our reliance on antibiotics

Antibiotics are the standard for treating severe pneumonia.

With PERCH results showing that viral causes predominate, a finding consistent in all settings evaluated by PERCH, and with the development of new tools to help identify the causes of pneumonia, treatment standards may be revised to tailor therapy and reduce antibiotic use.

Reducing the use of antibiotics is one strategy to lower the risk of antimicrobial resistance.

Preventing antimicrobial resistance by reducing our reliance on antibiotics would benefit children the world over. (Photo: Jane Crawley, January 2013)

Scientific, Human, and Technological Resources Forged

PERCH was committed to standardizing protocols, methods, testing algorithms, and equipment across all 7 countries.

Exchange visits for investigators and technicians—for example, between South Africa, The Gambia, and Kenya (shown here)—enabled learning and collaboration. (Photo: Maria Deloria Knoll, February 2013)
Chris Hani Baragwanath Hospital, PERCH’s South African base, is one of the world’s largest hospitals with over 3,000 beds. It serves the majority of Soweto’s population. (Photo: Dan Park, February 2013)
All PERCH testing was done at the sites except for quality assurance testing and a select subset of specialized assays performed at the study reference laboratory. (Photo: Adrian Brooks)

Hundreds of laboratory technicians, nurses, and other health workers developed cutting-edge skills.

All clinical standardization courses finished with an examination and presentation of certificates. Following training, investigators completed online follow up evaluations. (Photo: Jane Crawley, July 2013)
Pictured are PERCH clinical and laboratory staff at the PERCH South Africa site. (Photo: Jane Crawley, February 2012)
In Zambia, PERCH was carried out at University Teaching Hospital, an academic center and tertiary care facility located in the densely populated capital city of Lusaka. (Photo: Jane Crawley, June 2011)
PERCH clinicians and investigators pictured worked in Basse, the Gambia, which was one of the first low-income countries to begin using pneumococcal conjugate vaccines to prevent a common cause of severe child pneumonia. (Photo: Jane Crawley, April 2012)
In Dhaka, the Bangladeshi capital, PERCH tapped into an extensive demographic surveillance system run by icddr,b in a rapidly growing, densely populated urban community. (Photo: Jane Crawley, August 2012)
Pictured are PERCH clinical and laboratory staff at Hospital Gabriel Touré, Bamako, Mali, a teaching hospital in a district that suffers from high infant mortality. (Photo: Jane Crawley, August 2012)
PERCH clinicians and investigators in Sa Kaeo, a predominantly rural area in eastern Thailand. (Photo: Jane Crawley, June 2012)

11 institutions worked together to standardize protocols and train all involved in the research study:

  • BALTIMORE, MARYLAND Johns Hopkins Bloomberg School of Public Health, International Vaccine Access Center – central coordinating center
  • BASSE, THE GAMBIA Medical Research Council
  • KILIFI, KENYA KEMRI-Wellcome Trust Research Programme
  • BAMAKO, MALI Center for Vaccine Development, University of Maryland School of Medicine and Centre pour le Developpment des Vaccins (CVD-Mali)
  • DHAKA, BANGLADESH icddr,b
  • NONTHABURI, THAILAND Thailand Ministry of Public Health – U.S. CDC Collaboration
  • JOHANNESBURG, SOUTH AFRICA Medical Research Council: Respiratory and Meningeal Pathogens Research Unit; University of the Witwatersrand
  • LUSAKA, ZAMBIA Boston University School of Public Health and University Teaching Hospital
  • CHRISTCHURCH, NEW ZEALAND Canterbury Health Laboratories
    University of Otago
  • ROCKVILLE, MARYLAND The Emmes Company
  • OXFORD, UNITED KINGDOM Nuffield Department of Clinical Medicine, University of Oxford
FULL RESEARCH AND ADDITIONAL RESOURCES

Read the full results of PERCH in The Lancet.

Videos and training materials for clinicians on pneumonia diagnosis, measuring clinical signs, and collecting specimens are publicly available.

The study toolkit, including all protocols, data collection forms, and standard operating procedures created in PERCH, are also available to other researchers.

PERCH’s well-cataloged dataset and specimen bank, one of the largest available on childhood pneumonia, are being used by at least 15 external research groups to test, analyze, and pose their own questions. The dataset will soon be publicly available.

The statistical method (BAKER) used to analyze PERCH data is also available for researchers to analyze their data.

Thank you to the families and children who participated in PERCH.

PERCH was supported by a grant from The Bill & Melinda Gates Foundation.

© International Vaccine Access Center 2019. All are free to use facts and representation of findings with attribution.

Suggested Citation:
The PERCH Team. “Visualizing PERCH Results: Etiology of Pneumonia in Children Hospitalized in 7 Countries.” The International Vaccine Access Center: 2019. www.perchresults.org.

Design by House9