Frequency of Respiratory Syncytial Virus in Hospitalized Infants with Lower Acute Respiratory Tract Infection in Colombia

In developing countries acute respiratory infections (ARI) are the cause of death in >25% of children younger than 5 years of age. Lower tract respiratory infection, mostly pneumonia, causes 90% of these deaths.1, 2 Deaths from pneumonia can be reduced with antibiotic treatment; however, 40% of the pneumonias are caused by viruses3 for which antibiotics are useless. Respiratory syncytial virus (RSV) is the main causation agent of bronchiolitis and bronchopneumonia in nursing babies, especially between 2 and 6 months of age.3, 4 Indiscriminate antibiotic therapy in children with lower acute respiratory infections (ARI) in which no etiologic agent has been identified is ineffective and can lead to bacterial resistance, adverse reactions and increased cost of management because of longer hospitalization. In Colombia where most antibiotics can be purchased over the counter, there is a need to offer rapid diagnostic techniques to detect respiratory viruses, RSV among them, mainly in infants to decrease indiscriminate antibiotic use. There are few studies on the etiology of community-acquired pneumonia in developing countries. Immunofluorescence, a rapid diagnostic technique to detect viral particles, is inexpensive and has nearly the sensitivity of cell culture for detection of RSV infection in nasal wash specimens, with the advantage that it allows a diagnosis of RSV infection within hours.5, 6 The study presented here was conducted to determine the frequency of lower ARI caused by RSV in children younger than 1 year old who required hospitalization. Materials and methods. Patient selection. All patients younger than 1 year old, who had been hospitalized for lower acute respiratory infection at the General Hospital of Medellín, Colombia, from April, 1994, to April, 1995, were enrolled in the study. These patients were sent from different urban and rural medical centers from the south of the State of Antioquia. Clinical data recorded on admission included age, sex, duration of the illness before admission, presence of fever, cough, tachypnea (>50 respirations/min), use of respiratory accessory muscles, cyanosis in room air, prolonged expiration and expiratory wheezing. Other data collected were the duration of hospitalization and the presence and duration of antibiotic treatment. Erythrocyte sedimentation rate, white blood cell counts, C-reactive protein and chest radiographs were obtained during the first hours after admission. Diagnosis of viral infection was considered if the following findings were present: rhinorrhea; prolonged expiration; expiratory wheezing; interstitial infiltrates; and hyperinflation on chest radiographs, as well as negative tests for three or four acute phase reactants (50 respirations/minute), wheezing, use of respiratory accessory muscles, prolonged expiration, hyperinflation with or without atelectasis or infiltrates documented by chest radiographs. Finally the presence of barking cough, dysphonia and inspiratory stridor was indicative of laryngotracheobronchitis. Specimen collection. A nasopharyngeal wash was obtained from all patients within the first 36 h after admission. The samples were immediately transported to the virology laboratory of the Corporación para investigaciones Biológicas. They were centrifuged at 500 × g for 10 min. Supernatants from 12 samples were cultured in Hep-2 cells and isolated viruses were identified by fluorescein isothiocyanate-labeled monoclonal antibodies directed against RSV for detecting virus in infected tissue culture and prepared direct patient specimens (B 1029-80, Baxter Diagnostic, Inc., Isasaquah, WA). Cell pellet was centrifuged with phosphate-buffered saline and resuspended in 0.5 ml, a drop was placed on a slide and a direct immunofluorescence for RSV was carried out.6 All specimens had more than three epithelial cells per ×400 field. Statistics. We used statistically descriptive measures (mean and SD) for age, sex, duration of illness before hospitalization and mean duration in the hospital by using the statistical program EpiInfo 5.0. Results. Between April, 1994, and April, 1995, 103 patients with lower ARI were hospitalized at the General Hospital of Medellín. Forty-three patients (41.7%) were defined as having RSV infection by immunofluorescence. In 5 of 12 inoculated cell cultures a virus was isolated and RSV was identified. There was agreement among negative direct patient specimens immunofluorescence and no virus was isolated. Children with ARI had a mean age of 3.6 months (SD ± 3.2), similar to that found in RSV-positive patients, 3.2 months of age (SD ± 2.9); 37 (82.2%) of the latter patients were younger than 6 months old, whereas 27 (62.8%) were younger than 2 months of age. Sixty-five (63.1%) of the patients studied were males and 24 (55.8%) had RSV infection. The greatest number of admissions for both acute ARI and for RSV infection occurred in November, December and January (Fig. 1), which were particularly rainy that year. The peak incidence of cases was in January when 10 (23.3%) RSV-infected patients were identified. The mean duration of illness before hospitalization in the 103 patients was 6.4 days (SD ± 5.9); in RSV-positive patients it was 5.3 days (SD ± 4.3). The mean number of days of hospitalization of RSV-infected patients was 6.1 days (SD ± 3.4); in uninfected patients it was 9.4 days (SD ± 7.7). Antibiotic management of RSV-infected patients was as follows: 39.4% (15 of 43) patients with viral diagnosis strongly suggested by clinical findings did not receive antibiotics; 12 (27.9%) received antibiotics for 1 to 3 days based on findings for chest, erythrocyte sedimentation, white blood cell counts and C-reactive protein. Antibiotics were interrupted when RSV immunofluorescence yielded a positive result. All patients recovered without complications. Antibiotics were continued for >3 days in 16 (37.2%) RSV-infected patients when there was a clinical diagnosis of mixed pneumonia (viral and bacterial), when the patient was less than 2 months old, when there were two or more positive acute phase reactants or when chest radiographs were suspicious for bacterial infection. Discussion. The etiology of acute lower respiratory infection varies according to the diagnostic techniques used, the age of the population and the diagnostic techniques used, the age of the population and the geographic area. The Board on Science and Technology for International Development of the National Research Council reported in a comparison of findings of acute respiratory tract infection in young children from several developing countries that RSV was the predominant etiologic agent in Latin America, Africa and Asia and was more important that Streptococcus pneumoniae and Haemophilus influenzae.10 In a study by Trujillo et al.5 in Medellín, it was found that RSV occurred in 44% of the patients with lower ARI, 77% of whom were younger than 1 year old. These findings are in agreement with our current data showing that 41.7% cases of lower ARI in infants requiring hospitalization have RSV infection. We found pneumonia followed by bronchiolitis to be the principal reasons prompting hospitalization. Antibiotic therapy was withheld in children who either had lower ARI caused by RSV or had clinical and laboratory findings suggestive of viral infection. These patients did not have complications and no other diagnostic test were required, thus decreasing time of hospitalization. In Colombia respiratory viral infections are not routinely identified by culture or antigen detection. Consequently it was important to observe the impact of RSV diagnosis on the management of patients. Our results confirm that RSV is an important cause of serious lower respiratory tract illness in Colombian infants, especially those younger than 3 months of age. Diagnosis is useful if it lessens the indiscriminate use of antibiotics. Acknowledgments. We thank Dr. Frank Montoya, the nurses and all physicians of Hospital General de Medellín for collaboration in the surveillance of their patients; Dr. María Isabel Munera for helpful comments on the manuscript; and Drs. Angela Restrepo Moreno and Silvia Botero for reviewing the English language version of the manuscript. Victoria I. Bedoya, M,D.; Veronica Abad, M.D.; Hugo Trujillo, M.D. Corporación para Investigaciones Biológicas; Hospital General de Medellín Medellín, ColombiaFIG. 1: Monthly frequency distribution of lower respiratory tract infection during the study period.