Study team
EXECUTIVE SUMMARY

Background
Pneumonia is a leading cause of hospitalisation and death in children and adults in South Africa. HIV-infected individuals of all age groups have a substantially elevated risk of hospitalisation and death due to pneumonia. Outbreaks of novel respiratory pathogens such as pandemic influenza can have substantial public health implications. For the above reasons, a sustainable, representative syndromic pneumonia surveillance programme is needed.

OBJECTIVES

Epidemiology
Estimate the burden of total and aetiology-specific pneumonia-related hospitalization and mortality in HIV-infected and -uninfected children and adults.
Monitor trends in the relative contribution of important respiratory pathogens to the syndrome of pneumonia.
Describe the demographic and epidemiologic characteristics (including severity) of pneumonia cases presenting to sentinel surveillance sites and describe how these vary by age group, HIV status, aetiology and temporally.
Determine the seasonality of influenza, respiratory syncytial virus (RSV) and other respiratory pathogens.
Determine the extent of specific co-infections and describe how these co-infections relate to patient morbidity and outcome.
Compare the prevalence of Bordetella pertussis, Haemophilus influenzae, Streptococcus pneumoniae and influenza in vaccinated and unvaccinated children and identify vaccine failures.
Assess the burden of disease associated with influenza as well as other respiratory pathogens among hospitalised infants aged <1 year presenting with respiratory and non-respiratory acute medical illnesses for the 2 year period from 2016-2017.
Describe the clinical diagnosis and management of respiratory pathogens among hospitalised infants.

Organism characterization 
Characterize annual circulating influenza strains to guide vaccine strain selection for the Southern Hemisphere influenza vaccine, identify drift from the annual vaccine and identify antiviral drug resistance.
Monitor changes in pneumococcal serotype distribution over time and the effect of the pneumococcal conjugate vaccine on the proportion of vaccine and non-vaccine serotypes including serotype replacement.

Pandemic preparedness and outbreak response.
Identify and characterise novel respiratory pathogens in humans.
Identify and respond to respiratory disease outbreaks.

Methodology
Prospective hospital-based sentinel syndromic surveillance will be performed. Sites will be situated in: KwaZulu-Natal Province (Edendale Hospital), Mpumalanga Province (Mapulaneng and Matikwana Hospitals), North-West Province (Klerksdorp-Tshepong Hospital Complex), Gauteng Province (Rahima Moosa Mother and Child Hospital and Helen Joseph Hospital) and the Western Cape (Red Cross Children’s Hospital and Mitchell’s Plain Hospital (MPH)). Surveillance Officers (registered or enrolled nurses) will review the clinical details of patients admitted to sentinel sites on weekdays to establish whether cases meet the surveillance case definitions. Identified cases will be approached for inclusion. Consenting patients will be asked to complete a questionnaire by interview on previous medical history and clinical data and to provide specimens. A naso/oropharyngeal specimen (oropharyngeal (OP) and nasopharyngeal (NP) swab in all age groups and for a limited period an additional or nasopharyngeal aspirate (NPA) in cases <5 years of age enrolled at KTHC and Edendale Hospital) will be collected for identification of respiratory viruses and bacteria. Blood, urine (from adults) and sputum (from adults and from children at selected sites) will be collected for testing for respiratory viruses and bacteria. NPAs and NP/OP swabs will be tested at National Institute for Communicable Diseases (NICD) for the presence of influenza virus and other respiratory viruses (including RSV, human metapneumovirus and parainfluenza) and bacteria (including H. influenzae S. pneumoniae, Staphylococcus aureus, Bordetella species, Legionella species, Chlamydophila pneumoniae and Mycoplasma pneumoniae) by viral/bacterial culture and polymerase chain reaction (PCR) as appropriate. Sputum will be tested for Mycobacterium tuberculosis by culture as well as for other bacterial pathogens (including S. pneumoniae, Bordetella species, Legionella species, C. pneumoniae, M. Pneumoniae, S. aureus and Moraxella catarrhalis) and Pneumocystis jirovecii by PCR. Blood will be tested by PCR for bacteria including S. pneumoniae, H. influenzae, S. aureus, M. catarrhalis, Klebsiella pneumoniae and Pseudomonas aeruginosa. In addition, at selected sites, among hospitalised infants aged <1 year, additional blood that is collected at baseline (at enrolment) and an additional blood specimen collected 3-5 weeks after admission will be tested for the presence of antibodies to influenza, RSV and Bordetella pertussis in order to assess for acute infection. Urine will be tested for S. pneumoniae and Legionella pneumophila. Results of routine patient investigations such as HIV testing, CD4 count, HIV viral load, cytomegalovirus (CMV) testing and blood culture will be collected. If enrolled patients do not have a documented HIV status and HIV-positive patients do not have a recent CD4 count, HIV and CD4 count will be tested according to the national protocols. Data will be entered on a centralised database at NICD. Surveillance indicators will be reported to stakeholders weekly in the influenza season and monthly out of season. 

Background and literature review
Pneumonia in Children
Lower respiratory tract infections (LRTI) are the leading cause of morbidity and mortality amongst children aged <5 years globally. In 2010, an estimated 1.4 million children died due to LRTI and an estimated 12 million were hospitalised [1, 2]. LRTI is the most common cause of hospitalisation in HIV-infected children in sub-Saharan Africa and HIV infection is associated with increased severity of LRTI and higher case fatality ratios [1, 3] . The relative contribution of viral and bacterial aetiologies to the syndrome of pneumonia also varies by HIV status. Data from the SARI surveillance system estimates the overall incidence of LRTI in South African children <5 years of age to be 2530-3173/100,000 with the highest incidence in children aged <1 year (annual range 8446-10532/100,000). The risk of hospitalization with LRTI was up to three times higher in HIV-infected compared to HIV-uninfected children [4]. 

Many pneumonia hospitalisations and deaths in children are preventable through public health interventions the most notable of which is vaccination; with vaccines available against several of the major causes of childhood pneumonia (pneumococcus, pertussis, Hib and influenza). The Hib conjugate vaccine was introduced into the South African immunisation programme in 1999 and the pneumococcal conjugate vaccine (PCV) in April 2009[5]. Pertussis vaccination for children was introduced in the South African vaccine schedule in 1995. Influenza vaccine is administered through the public and private sectors in South Africa but <1 million doses are distributed each year through the public sector. As bacterial aetiologies decline due to vaccination, respiratory viral causes of LRTI may gain greater prominence.

Pneumonia in adults
Community-acquired pneumonia (CAP) and influenza are the second commonest cause of death in South African adults aged 15-49 years accounting for 10% of all deaths in this age group in 2005[6]. The major causative pathogens of CAP in South African adults include common bacterial causes (S. pneumoniae, S. aureus and H. influenzae) and atypical bacterial causes (Mycoplasma pneumoniae, Chlamydophila pneumoniae, and Legionella spp.) [7, 8]. SARI data describes the annual incidence of SARI hospitalisation in South African adults to be 388-619/100,000 and highest in individuals aged 45-64 years (640-1285/100,000). HIV-infected adults have 13-19 times increased risk of hospitalisation with SARI as compared to HIV-uninfected adults and were more likely to be co-infected with pneumococcus, experience prolonged hospitalization and increased risk of death once hospitalised. The case-fatality ratio was 1.5 times greater amongst HIV-infected (8%, 351/4610) as compared to HIV-uninfected (5%, 85/1653) individuals [9]. HIV-infected patients are at increased risk of pneumonia due to bacterial pathogens such as S. pneumoniae; in addition pathogens such as P. jirovecii and TB play a major role in adults infected with HIV [10, 11]. The role of atypical bacteria is not well described.

METHODS AND PROCEDURES

Objectives
Epidemiology
Estimate the burden of total and aetiology-specific pneumonia-related hospitalization and mortality associated with pneumonia and by aetiology in HIV-infected and –uninfected children and adults.
Monitor trends in the relative contribution of important respiratory pathogens to the syndrome of pneumonia.
Describe the demographic and epidemiologic characteristics (including severity) of pneumonia cases presenting to sentinel surveillance sites and describe how these vary by age group, HIV status, aetiology and temporally.
Determine the seasonality of the influenza and respiratory syncytial virus (RSV) and other respiratory pathogens seasons annually.
Determine the extent of specific co-infections and describe how these co-infections relate to patient morbidity and outcome.
Compare the prevalence of B. pertussis, H. influenzae, S. pneumoniae and influenza in vaccinated and unvaccinated children and identify vaccine failures.
Assess the burden of disease associated with influenza as well as other respiratory pathogens among hospitalised infants aged <1 year presenting with respiratory and non-respiratory acute medical illnesses for the 2 year period from 2016-2017.
Describe the clinical diagnosis and management of respiratory pathogens among hospitalised infants.

Organism characterization
Characterize annual circulating influenza strains to guide vaccine strain selection for the Southern Hemisphere influenza vaccine, identify drift from the annual vaccine and identify antiviral drug resistance.
Monitor changes in pneumococcal serotype distribution over time and the effect of the pneumococcal conjugate vaccine on the proportion of vaccine and non-vaccine serotypes including serotype replacement.

Pandemic preparedness
Identify and characterise novel respiratory pathogens in humans
Identify and respond to respiratory disease outbreaks

Surveillance Programme Design
A prospective syndromic hospital-based sentinel surveillance programme will be introduced. Sites will be situated in five of the nine provinces. The programme will be conducted at hospitals with existing surveillance programmes: KwaZulu-Natal Province (Edendale Hospital), Mpumalunga Province (Mapulaneng and Matikwana Hospitals), North-West Province (Klerksdorp-Tshepong Hospital Complex). Additional sites will be started in Gauteng (Rahima Moosa Mother and Child Hospital (RMMCH) and Helen Joseph Hospital (HJH)) and Western Cape (Red Cross Children’s Hospital (RCCH) and Mitchell’s Plain Hospital (MPH)) during 2015-2016 period. Surveillance officers (registered/enrolled nurses) will review the clinical details of patients admitted to sentinel sites to establish whether cases meet the surveillance case definition/s (Table1). Identified cases will be approached for inclusion in the sentinel surveillance. Consenting patients will be asked to complete a questionnaire by interview on previous medical history and clinical data and to provide specimens as described in table 2 below. 

At RMMCH, surveillance for pertussis has been conducted by NICD investigators since 2013 (Protocol title: Sentinel Surveillance for pertussis in South Africa,  as an addition to the protocol ‘Essential Communicable Disease Surveillance Activities of the National Institute for Communicable Diseases of Public Health Importance in South Africa, M060449). Going forward, pertussis surveillance at RMMCH will be incorporated into the NICD pneumonia surveillance programme described in this protocol. To ensure comparability of data at this site case definitions and specimen collection procedures at RMMCH will be slightly modified as detailed below.

 

Study team