The burden and etiology of lower respiratory tract infections in children under five years of age in Indonesia

Introduction: Lower respiratory tract infections (LRTI) are a substantial threat for children worldwide. Currently, there is a lack of knowledge about the burden and etiology of LRTI in children under five years of age in Indonesia. Methodology: We have systematically searched the available English and Indonesian scientific literature to review and summarize data on LRTI and LRTI-associated invasive disease, and bacterial carriage in the upper respiratory tract in children under five years of age in Indonesia. Results: Overall, data on the burden and etiology of LRTI in children under five years of age in Indonesia is very limited. The data are primarily collected in Java. Data from other parts of Indonesia, including Sumatra, Kalimantan, and Sulawesi, are scarce. The case fatality rate (CFR) of LRTI in children under five years of age in Indonesia was 0.11%. Influenza was the most commonly reported viral etiological agent of LRTI in children under five years of age in Indonesia. Klebsiella pneumoniae was the most frequently reported bacterial agent of LRTI. Streptococcus pneumoniae showed the highest carriage rates. Conclusions: Surveillance and diagnostic studies are urgently needed and should be conducted in different parts of Indonesia to improve insight in the burden and etiology of LRTI in Indonesia. These data are pivotal to increase the effectiveness of public health strategies, including vaccination and prevention of antimicrobial resistance.


Introduction
Lower respiratory tract infections (LRTI) are a substantial threat for children resulting in high morbidity and mortality rates worldwide. It is estimated that, globally, 15.5% of children mortality is due to LRTI [1]. Indonesia is one of the developing countries with a high contribution to global pneumonia cases [2]. Previous studies have been conducted to determine strategies to reduce the burden of LRTI in Indonesia through surveillance and vaccination [3][4][5][6][7]. In several epidemiological studies, carriage rates of potentially pathogenic bacteria, i.e. Streptococcus pneumoniae, Staphylococcus aureus, Haemophilus influenzae, Moraxella catarrhalis and also Klebsiella pneumoniae were measured in healthy children in Central Java, West Java, and West Sumatra provinces [8][9][10]. The serotype distribution of S. pneumoniae was studied in Jakarta and West Nusa Tenggara to determine the effectiveness of the pneumococcal vaccine [11][12][13]. Although the Ministry of Health of the Republic of Indonesia has tried to investigate the etiology of LRTI in children through Severe Acute Respiratory Infection (SARI) surveillance, currently, many findings have not been published in international journals and are thus not accessible through medical literature databases [14]. In order to obtain insight into what is currently known about the burden of LRTI, associated invasive disease and bacterial carriage in Indonesian children, we systematically searched all available English and Indonesian literature.

Methodology
The PRISMA checklist was followed for the development of our study protocol [15]. The systematic review protocol of this study was registered in PROSPERO under number CRD42020153212 (available at https://www.crd.york.ac.uk/prospero/display_record.p hp?RecordID=153212).

Eligibility criteria
Studies on LRTI in children under five years of age in Indonesia were eligible for inclusion. In this paper, the term LRTI is used in its broadest sense referring to all infections in the lungs or below the vocal cord, excluding tuberculosis. Additionally, we included studies that have stratified age data of children below the age of five. Exclusion criteria were studies without age-specific data, studies without any information on pathogen identification, review articles, either systematic or narrative, and posters.

Literature search
We searched the following electronic databases: PubMed, Embase, Web of Science, Scopus, and ProQuest for studies published between January 2009 and October 2019. The search was limited to publications written in English and Indonesian. We were using a combination of search strings for lower respiratory tract infection, children, and Indonesia. In addition, we used Google Scholar to search for articles written in Indonesian using the keywords pneumonia, and anak (children).

Data collection
Two reviewers independently screened the title and the abstract of all retrieved studies. Disagreements between the two reviewers were solved through discussion. Studies without relevant information were excluded. Later, the full texts of the selected articles were assessed for eligibility using the same procedure as the abstract screening.

Quality assessment
Two reviewers independently assessed the risk of bias by using the Newcastle-Ottawa Scale (NOS) [16].
The final global quality score was discussed by two reviewers.

Data extraction and analysis
All the included studies were extracted using the modified Cochrane data extraction form. Extracted data included study period, research method, sample size, age group, prevalence, incidence, case fatality rate, carrier and carriage rate, etiology, and laboratory method of identification.

Results
We archived 1,756 publications from our search strategy and included 36 publications for analysis ( Figure 1).

Etiology
Twelve studies described the etiology of LRTI in children under five years of age in Indonesia (Table 2) [18,19,33,[35][36][37][38][39][40][41][42][43]. The studies were conducted in different settings (hospital based and/ or primary health care based). In addition, different laboratory method of identification was used. Five studies focused on viral pathogens, while six studies focused on bacteria as the main pathogen [18,19,36,38,41]. Only one study described both bacterial and viral [35]. Figure 3 shows the location and indicates the sample size of the different studies reporting on etiology of lower respiratory tract infection in children under five years of age in Indonesia.

Viral
Based on the number of studies and the number of cases described in the separate studies, influenza virus was the most common cause of viral respiratory tract infections in children under five years of age in Indonesia. Other viruses prevalent in this age group and mentioned in the different studies were respiratory syncytial virus, coxsackievirus, enterovirus, coronavirus, adenovirus, parainfluenza virus, bocavirus, rhinovirus, and human metapneumovirus [35,41].

Influenza Virus
Data regarding influenza virus infections in Indonesian children were found in five different studies [19,35,36,38,41]. Influenza A virus was the most commonly reported type. The studies identified not only H1 and H3 subtypes but also H5 [19,35,38,41]. One study conducted in an infectious disease hospital in Jakarta analyzed the clinical, laboratory, and radiologic characteristics of confirmed avian influenza (H5N1) cases. There were three patients under five years of age from all thirteen children. Two out of three died. The main clinical conditions were fever, productive cough, and dyspnea. The cause of death was acute respiratory distress syndrome [38]. Another study, also conducted in Jakarta, found influenza B in twelve patients among 167 children age 0-14 years diagnosed with influenzalike-illness and pneumonia [36].

RSV
Two studies identified RSV as the causative agent of LRTI [18,41]. One large study in Bandung, West Java, consist of 2,014 participants under five years of age, revealed that in Indonesia, the incidence of LRTI due to RSV was relatively low. No RSV found as the etiology of LRTI in children under two months of age [18].

Bacterial
All studies that reported bacteria as the cause of LRTI in children under five years of age in Indonesia were hospital-based. All studies were conducted in Java Island, except two studies located in Bali and Sumatera [33,43].

S. pneumoniae
Only three out of twelve studies identified S. pneumoniae, 19 cases in total [35,40,43]. Two studies identified it from a blood culture sample (see Invasive Disease). The third study used multiplex PCR to detect S. pneumoniae among hospitalized-suspected H5N1 patients. In this study, the prevalence of S. pneumoniae co-occurring or co-infecting with H. influenzae and K. pneumoniae was higher than a single infection with S. pneumoniae [35].

B. pertussis
One study retrospectively identified B. pertussis associated with pertussis-like symptoms in sixty-one probable and two confirmed cases. All patients were in the paroxysmal phase when they came to the hospital; thus, only a limited number of B. pertussis could be detected. Seventy-nine percent of the case was at the age of 0-5 months. Only two patients had a history of pertussis vaccination. No fatality found. Unfortunately, household contacts could not be screened, and no other information related to a possible outbreak was reported [37].

Viral-bacterial co-infection
One study conducted in Jakarta aimed to find the etiology of lower respiratory tract infection in 108 children and 122 adults. The authors reported polymicrobial infections, both viral and bacterial pathogens, as measured in the nasal swab, throat swab, tracheal aspirates, and bronchoalveolar lavage. From this study, it was concluded that bacterial-viral coinfection, in particular K. pneumoniae and Influenza A, was the most common cause, followed by bacterial coinfection (S. pneumoniae and H. influenzae) [35].

Bacterial carriage
Studies focused on the bacterial carriage in the upper respiratory tract of children under five years of age were done in Java, Lombok, and Sumatera island ( Table 3) [8][9][10][11][12][44][45][46][47]. Figure 4 shows the location and indicates the sample size of the different studies reporting on bacterial carriage in children under five years of age in Indonesia.
Three of the five studies found 6 A/B serotype as the most prevalent serotype in Indonesian children [10,12,45]. Other serotypes that were frequently found in all five studies were vaccine types 19F and 23F and non-vaccine types 2 and 15B/C [9][10][11][12]45].

H. influenzae
H. influenzae carriage rates ranged from 12% up to 32.9% in two studies [9,47]. One of these studies serotyped H. influenzae isolate, and none of the 83 isolates was H. influenzae type B, which reflects the effect of the Hib vaccine on a carriage [9].

M. catarrhalis
One study in Semarang, Central Java, reports the carriage rate of M. catarrhalis from children under five years of age at 5.3%. The prevalence of M. catarrhalis from healthy children aged 12-24 months in Lombok, West Nusa Tenggara, was ten times higher than Semarang, approximately, making it the highest also compared to Bandung and Padang, West Sumatra [9,44,47].
S. aureus S. aureus prevalence was the lowest compared to S. pneumoniae, H. influenzae, and M. catarrhalis. The prevalence was between 6.7% until approximately 9.0%. S. aureus had the same prevalence in Bandung as in Central Lombok (around 8.0%) [9,47].

Other bacterial pathogens identified in carriage studies
Semarang was the only place where K. pneumoniae, Pseudomonas spp. and A. baumannii comp. were identified in asymptomatic carriers. The range of carriage rates of K. pneumoniae varied from 2.9% to 7.0% [8,46]. Pseudomonas spp. and A. baumannii comp. were found in 7% and 5% of the carriers, respectively [8].

Invasive Disease
Three studies located in Denpasar, Bandung, and Jakarta showed the burden of invasive pneumococcal disease (IPD) in Indonesia [33,40,43]. Two studies directly aimed to measure the burden of invasive pneumococcal disease [33,40]. No pneumococci were detected in the 736 IPD and pneumonia patients in Denpasar [33]. A study in Bandung found nine pneumococcal isolates in blood cultures from 486 patients with pneumonia [43]. A study in Jakarta found one IPD case from 205 patients diagnosed with pneumonia, meningitis, sepsis, and suspected occult bacteremia. The IPD patient was a three-month-old baby boy infected with S. pneumoniae serotype 7F, had bilateral lobar pneumonia and meningitis and died on the fifth day of hospitalization because of respiratory failure [40].

Discussion
The objective of this study was to collect all available data to have a complete overview of the current status of the burden and etiology of LRTI in children under five years of age in Indonesia. Furthermore, we obtained all available information about asymptomatic carriage of well-known LRTI pathogens and about invasive infections associated with respiratory complaints in this population, which are strongly related to LRTI and therefore important to include in this review.
We found that the influenza virus was the most commonly reported viral cause of LRTI in children under five years of age in Indonesia. Although this result differs from other studies performed in other populations, which found that RSV is most common in children under five years of age [50][51][52]. K. pneumoniae was the most commonly reported bacterial cause of LRTI. This has earlier been found in other populations, as reported in a study conducted in China [53].
In healthy asymptomatic Indonesian children under five years of age, not vaccinated with a pneumococcal vaccine, S. pneumoniae had the highest prevalence rate. The highest carrier rate reported was 68.4% [45]. These data are in line with a previous study in Thailand, showing a carriage rate of 67.6-83.6% [54]. Although we noticed substantial carriage rates with S. pneumoniae in healthy children under five years of age in Indonesia, it is not the most commonly identified bacterial cause of LRTI in children under five years of age in Indonesia. This is likely an underrepresentation due to the use of insensitive diagnostic methods. Interestingly K. pneumoniae, the most commonly reported cause of LRTI, is also found in asymptomatic carriers in Indonesia. It is speculated that this is due to the consumption of contaminated food and water [8]. The coverage of vaccine serotypes (based on PCV13) was between 45-60%. A similar conclusion was reached from a study reporting prevalence of S. pneumoniae in South East Asia, in which a coverage of 65% was found [55].
Unfortunately, there is a lack of information regarding the incidence of LRTI-related invasive disease in children under five years of age in Indonesia. Therefore no conclusions can be drawn based on the available literature. This strongly demands for more studies identifying the etiology of LRTI-related invasive disease.
Although we only included a total of 36 English and Indonesian articles, we are confident that this is all relevant literature available at this moment. Reviewing this literature gives a better understanding of what is known about the burden and etiology of LRTI in children in Indonesia. However, at the same time we have to conclude that there is an important lack of information regarding the burden and etiology of LRTI, especially because most studies are done in certain regions of Java, while there is hardly any information available from other regions. Based on this review we recommend to improve the diagnostics of respiratory tract infections in children in Indonesia and to enhance the capacity of infectious disease surveillance, in particular in other provinces than Jakarta, West and Central Java.

Conclusions
Surveillance and diagnostic studies are urgently needed and should be conducted in different parts of Indonesia to improve insight in the burden and etiology of LRTI in Indonesia. These data are pivotal to increase the effectiveness of public health strategies, including vaccination and prevention of antimicrobial resistance.