Pediatric respiratory tract infection is one of the common reasons for physician's visits and hospitalization resulting in significant morbidity and mortality (Nair et al., 2010). The incidence of acute respiratory infections (ARIs) in children aged less than 5 years is estimated to be 0.29 and 0.05 episodes per child-year in developing and industrialized countries, respectively; 151 and 5 million new episodes each year, respectively. Pneumonia is responsible for about 21% of all deaths in children aged less than 5 years, leading to an estimate of 12 to 20 deaths per 1,000 live birth (Nair et al., 2010; Duke, 2005).

ARIs in children are classified as upper respiratory tract infections (URIs) and lower respiratory tract infections (LRIs). Most URIs resolve spontaneously within 3 to 10 days with symptomatic therapy alone. Frequent hand washing remains the most important preventive measure for most URIs (Killingley et al., 2011). On the other hand, about 20% of all deaths in children under 5 years are due to Acute lower respiratory tract infections (ALRIs) (pneumonia, bronchiolitis and bronchitis); 90% of these deaths are due to pneumonia. Every year, pneumonia contributes to 750,000 to 1.2 million neonatal deaths worldwide (Duke, 2005; World Health Organisation (WHO), 2010). Early detection and prompt treatment of pneumonia is lifesaving. 

In infants aged up to three months, pneumonia is mostly of a bacteriological etiology and Streptococcus pneumoniae is the most common isolated pathogen. In infants older than four months and preschool-aged children, viruses are the most contributing agents. Bacterial infections can occur at any time throughout the year in preschool/school-aged children and adolescents (Ostapchuk et al., 2004). Generally, S. pneumoniae is the most common bacterial cause of pneumonia; less common bacterial etiologies include Haemophilus influenzae serotype b, Moraxella catarrhalis, and Staphylococcus aureus. While Mycoplasma pneumoniae and Chlamydia pneumonia are more associated with pneumonia in pre-school-aged children. Mycobacterium tuberculosis may cause community acquired pneumonia (CAP) in exposed children. Co-infection of two or more microbial agents is not uncommon with a rate up to 41% in hospitalized patients (Bradley, 2002; Ostapchuk et al., 2004).                                           

In the same regards, the most important viral groups that cause pneumonia as a primary manifestation of a disease include influenza virus types A and B, respiratory syncytial virus (RSV), adenovirus, parainfluenza virus, rhinovirus, Hantavirus, severe acute respiratory syndrome (SARS) and cytomegalovirus (CMV). While the ones that cause pneumonia as part of a multisystem syndrome include Paramyxovirus species (measles), varicella-zoster virus, and Epstein-Barr virus, cytomegalo virus and herpes simplex virus (Mackie, 2003). According to the WHO, Sudan is one of the fifteen countries that bears the highest burden of child ALRI in the East Mediterranean Region (EMR). The estimated new cases of pneumonia is around two million and 0.48 episode per child per year (Rudan et al., 2008).

This study aims to identify causative bacteriological and virological agents of ALRI among children attending Khartoum Pediatrics Emergency Hospital and to assess association between selected factors and ALRI.

Ethical approval was obtained from the ethical review committee in Sudan National Board of Medical Specialization and Sudan Federal Ministry of Health (FMOH); informed consent was obtained from parents of all study participants.

Study design

The study design was a cross sectional, descriptive hospital based study. The study was carried out in Children Emergency Hospital in Khartoum which is a referral pediatrics hospital in Sudan and serving almost the whole population of Khartoum state. Sample collection was done from the period of November, 2011 to February, 2012.

Study population     

All children younger than 5 years old who were diagnosed with acute lower respiratory tract infection were considered to be the target study population. The presence of respiratory symptoms for ≤5 days and parental written consent to participate in the study were the study inclusion criteria. Children with contraindications to respiratory specimens sampling were excluded.

Sample size                                                                                                             

A total convenient sample of 123 participants who agreed to undergo blood and nasopharyngeal laboratory sampling were included and tested for bacteriological tests. While only 73 of them were successfully tested for both bacteriological and virological examinations. Shortage of the testing reagents and samples shipment restriction were the major constraint that prevented full samples analysis.

Data collection and processing

Data was collected through a structured questionnaire. Blood samples and nasopharyngeal samples were collected applying the recommended infection control precautions. Specimens were processed at the National Public Health Laboratory and private virological laboratory in Khartoum. Statistical Package for the Social Sciences (SPSS-19) was used to analyze the data. Descriptive statistics was done in addition to Chi square test to assess association between selected variables and the outcome.

Outcome determination and quality assurance

Children with ALRTI were diagnosed clinically by senior pediatricians based on presenting symptoms and signs. Severity of pneumonia was classified depending on the WHO classification of pneumonia. Direct immunofluorescence techniques and culture methods were used to identify etiological agents for respiratory infections (Murray et al., 1990). For virology, the study used IMAGENTM reagent from Oxoid (UK); its sensitivity ranges between 87 and 100% and specificity ranges between 98 and 100 % (according to the manufacturer). These reagents include monoclonal antibodies conjugated to fluorescein isothiocyanate (FITC) and Evans blue dye as a counter stain. The monoclonal antibodies are targeted against matrix protein of influenza A, nucleoprotein of influenza B, Fusion protein of RSV and parainfluenza 1, heamaglutinin protein of parainfluenza 2 and 3 and against the hexon protein of the adenovirus.   

For bacteriology, chocolate agar, blood agar and MacConkey agar were used for primary isolation.  Cultures were incubated at 35°C, for 24 to 48 h. Blood and chocolate plates were incubated in 5 to 7% Co₂.  Colonial morphology was observed and gram stain and relevant chemical tests were done for identification (Duke, 2005). For mixed growth, purification  was  attempted,  identification methods were chosen according to the colonial morphology and gram stain results.To identify gram positive cocci, catalase positive, bacitracin, coagulase, and mannitol fermentation tests were used. While for catalase negative, optochin susceptibility and pattern of blood hemolysis were used. To identify gram negative cocci/coccobacilli and short bacilli, ability to grow on MacConkey agar, oxidase, catalase, requirement for X and V factors, penicillin disc test, acid production from glucose and lactose and DNAse tests were used.

Overall

This study involved 123 children, 69 (56%) of them were males. Median participant’s age is 9 months (range: 1 to 60 months), with 60% of them ≤1 year of age. Sixteen children had low birth weight (13.7%) and 9 were born prematurely (7%). Sixty six fathers were smokers, while all mothers have denied a history of tobacco smoking. 14% of the mothers were found to be illiterate, 33% had primary education, 36% completed high school and 17% had university education. All 123 patients were tested for bacterial pathogens and seventy three of them were tested for viral etiology as well. One or more viral pathogens were detected in 49 (67%) patients. Bacterial pathogens were detected in 49 (40%) patients. Co-infection by both viral and bacterial agents was detected in 20 (27.5%) patients. Table 1 summarizes the study participant’s characteristics while Table 2 describes the results of microbiological testing.

Virology

Among 73 tested samples for virology, RSV was found to be the most prevalent virus (21 patients, 29%), followed by Influenza A (9 patients, 12%), then Parainfluenza 3 (7 patients, 10%), Influenza B (6 patients, 8%) and Adenovirus (3 patients, 4%). Parainfluenza 1 and 2 viruses were not detected in any of these samples. Concomitant infection and presence of more than one virus at a time was detected in three patients (4%); in all of them RSV was present along with either Influenza B or Influenza A.

Bronchiolitis was the commonest clinical diagnosis among study participants (49 patients, 40%), followed by pneumonia (47 patients, 38%)  and  the  rest  patients (27 patients, 22%) were diagnosed with severe pneumonia and viral agents were found predominantly in all of these.

In patients with a documented respiratory virus infection, Bronchiolitis was caused predominantly by RSV (31%), and it was also the most frequently detected virus in pneumonic cases (34.5%). RSV is found to be a causative agent responsible of around 13% of severe pneumonia along with Adenovirus and Influenza B virus.

Bacteriology

Among 123 tested samples for bacteriology, bacterial pathogens were detected in 39.8% of the children. Streptococcus pneumoniae was detected in 14 patients (11.4%). M. catarrhalis was isolated from 11 patients (8.9%). Klebsiella pneumoniae was isolated from 9 patients (7.3%).  Pseudomonas aeruginosa was isolated from seven patients (5.7%). H. influenzae was isolated in pure culture from 2 patients (1.6%). S. aureus was also isolated from five patients (4.1%).  Escherichia coli was isolated as a sole organism from one patient (0.8%). Combined bacterial and viral infections were found in 20 patients (27.4%). RSV infection was accompanied by S. pneumonia in 7 patients, with M. catarrhalis in 4 patients and with P. aeruginosa in 3 patients. Six Influenza A cases were found mixed with S. pneumoniae. Table 2 summarizes the common isolated organisms in the study.

Associated factors

There was statistically significant difference between males and females in susceptibility for acquiring viral respiratory infections in c² test (p = 0.03). Males showed higher prevalence of viral infections (77.5%). There was no significant difference in gender regarding bacterial pathogen (p = 0.2).

Children age group between 2 and 12 months showed a higher prevalence of viral agents (44%) of the overall positive viral agents. RSV virus was the most prevalent virus among all age groups except the oldest group (37 to 60 months). Older age group has a similar incidence of RSV, Influenza B and Parainfluenza 3.  Age had no statistical significance association with bacterial infections in this study (p = 0.5). Congenital anomalies were observed in 12 participants and bacterial pathogens (K. pneumoniae and P. aeruginosa) were detected in 4 (33%) of them. A chest deformity patient was found to be positive for both viral and bacterial pathogens (S. pneumoniae + Influenza B virus). History of low birth weight (<2 kg) was detected in 16 patients (13%) including 9 premature cases (7%), 10 of them were positive for bacterial pathogens. In addition they were found to be predominantly infected by RSV; while only one patient had a co-infection of RSV and Influenza B (p = 0.01). Four malnourished infants (3%) were encountered during the study; all of them were co-infected by viral and bacterial agents. Parental smoking was found to be strongly correlated with viral infection (p = 0.004). Variables that did not show any statistical association with child ALRI includes maternal education level, high mean leukocyte or lymphocytes counts, family history of asthma and exclusive child breast feeding in the first six months of life.

Proper diagnosis, recognition of complications, and appropriate management of patients with ALRIs is a crucial (Michelow et al., 2004). In this study, ALRI was found to be caused mostly by viruses, this findingis in a coherence with most of the studies worldwide (Michelow et al., 2004; Padilla et al., 2010; Debiaggi et al., 2012; Berkley et al., 2010; Niang et al., 2010). On the other hand, this finding contradicts another Sudanese study that concluded the reverse (viral agents were found to be less prevalent) (Salih et al., 1994). Contradiction between the mentioned study and current one could be due to difference of participants’ age groups.

RSV proved to be the most prevalent respiratory virus among children in Sudan; also concluded from previously mentioned Sudanese study in 1994, matches with most similar studies in developing and developed countries (Chen et al., 2014; Zurawski et al., 2011; Salih et al., 1994; Berkley et al., 2010). Importantly, Rhinovirus and human Bocavirus were detected as the most common pathogens in some other studies and unfortunately these tests could not be performed in the current study due to unavailability of their reagents in Sudan during the study period. This study confirmed the existence of dual infections by bacterial and viral agents at a time, this fact established from several similar studies (Cevey et al., 2009; da Silva et al., 2013; Mameli and Zuccotti, 2013).

Bacteriologically, this study proved that S. pneumoniae is the common circulating bacterial pathogen in child respiratory infections in Sudan. This finding was concluded as well by similar studies around the world (Cevey et al., 2009; Honkinen et al., 2012; Zurawski et al., 2011). Only two cases of H. influenzae were detected in the study participants and this would intellectually be explained by the introduction of H. influenzae type b (Hib) vaccine to the routine immunization program scheduled in Sudan with sustained coverage above 90% since 2009.

Regarding the associated factors, that prevalence of viral infection was found to be higher among infants (2 to 12 months) which is consistent with the previous study in Sudan (Salih et al., 1994). Male children showed higher susceptibility to viral infection. Although this correlation was found in other studies (Sovero et al., 2011), but currentlycould be confounded by other predisposing factors (age, prematurity, low birth weight and congenital anomalies) since no statistical adjustments was done. A strong relationship was demonstrated between viral infection and congenital anomalies, low birth weight, prematurity and under-nutrition. This result matches with the result of a global study done by the WHO (Rudan et al., 2011). The same global study attributed higher incidence of pneumonia to the lack of exclusive breast feeding in the first few months of life, while this study failed to demonstrate such a link. In accordance with another study, white blood cell count and differential was not found to be beneficial in discriminating between viral and bacterial etiologies (Don et al., 2009). Household overcrowding seems to have a great effect on acquisition of viral respiratory infections. The risk increases when more than three persons share one room at bed time. Exposure to tobacco smoking showed strong correlation with viral infection among study participants. This agrees with the Sudanese study which found a direct link between parental smoking and susceptibility to pneumonia (Salih et al., 2011).

This study has some limitations that could be addressed in future similar investigations. The sampling technique compromised the study and made it more descriptive. It would have been better if all 123 samples were tested for both bacteriological and virological agents. In addition, there is a possibility of result distortion by a cofounding effect of some other variables regarding assessing association with outcomes. This could be avoided by proper sampling technique, adequate sample size and applying advance statistical models.

Viral pathogens are more common than bacterial ones in young children attending Khartoum National Pediatrics Emergency Hospital. Confirmation  of  a  viral  etiology for respiratory infection is important, since currently more antiviral therapies become available as well as prophylaxis for critical patients such as the premature and those with congenital anomalies. In addition, national ARI surveillance system should be strengthened to provide the evidence for EPI department to include further vaccines in the national EPI schedule.

Authors disclose that there are no financial or other relevant competing interests.

The authors would like to acknowledge the support received from all colleagues, particularly Dr Alnajeeb Sulaiman senior microbiologist, Dr Sara Ahmed Eltigani microbiologist in Sudan National Public Health Laboratory and Mr Ahmad Khairat for their intellectual discussions and technical advises provided throughout the study phases.