Bronchiolitis in Thomas Cherian,
Mercy John,
Tropical South
MD; Eric A. F. Simoes, MD; Mark C. Steinhoff, MD; K. Chitra, MBBS; Raghupathy, MD; T. Jacob John, PhD, FRCP
MD; P.
\s=b\ In a prospective hospital-based study of 328 children under 5 years of age with acute lower respiratory infections, 114 (35%) were diagnosed to have acute bronchiolitis. Of them, 87 (76%) were less than 1 year and 107 (94%) were less than 2 years of age. Signs of severe lower respi-
ratory infections, namely tachypnea (respiratory rate greater than 50/min) and subcostal retraction, were present in 95% and 93%, respectively. Of 88 children of whom roentgenographs were taken, 30 (34%) had evidence of pneumonia. No clinical
signs discriminated between those with and without pneumonia. By culture or immunofluorescence antigen detection, viruses were found in 81 (71%) children with bronchiolitis; respiratory syncytial virus the most common agent, found in 65 Parainfluenza viruses were the next most common, found in 12 (11%). Most cases of bronchiolitis occurred in outbreaks during the rainy months of Auwas
(57%).
North America and Europe, acute bronchiolitis is the most common low¬ er respiratory tract infection (LRI) in infants.1,z Its epidemiologie and etiologic features are well described. It occurs as outbreaks in winter and spring; respira¬ tory syncytial virus (RSV) is the usual cause.2"5 Although known to occur in de¬ veloping countries in the tropics,6"18 its epidemiologie and etiologic features have not been examined in detail. One study described its clinical features and seasonality, but not the causes.6 Respi¬ ratory syncytial virus infection, preva-
Accepted for publication November 27,1989. From the Departments of Virology (Drs T. J. John and M. John) and Child Health (Drs Cherian, Raghupathy, and Chitra), Christian Medical College and Hospital, Vellore, Tamilnadu, India; De-
partments of Pediatrics and International Health, Hopkins University, Baltimore, Md (Dr Steinhoff); and the Department of Pediatrics, University of Colorado Medical Center, Denver (Dr Simoes). Reprint requests to Department of Virology, CMC Hospital, Vellore 632004, Tamil Nadu, India (Dr T.J.John).
Johns
India
gust through November, coinciding with
respiratory syncytial virus outbreaks. Although bacterial culture of blood was done in 56 children, no respiratory pathogen was isolated. In one child with bronchiolitis and consolidation, postmortem
lung aspirate yielded Staphylococcus aurThus, bronchiolitis is primarily a viral syndrome in this tropical region, just as it is in temperate regions. Eight (7%) children died (all were infants); 5 had roentgenographic pneumonia and the remaining had other abnormalities contributing
eus.
to death; all had been treated with antibiotics. Since one third of lower respiratory infections are bronchiolitis, and among infants under 1 year of age bronchiolitis comprises 47% of all lower respiratory infection cases, criteria for antibiotic management must take into account the availability of roentgenographic investigation.
(AJDC. 1990;144:1026-1030)
lent July through October, was found in 20% to 60% of cases of bronchiolitis in other studies.7"11 We have recently in¬ vestigated acute respiratory infections (ARIs) in children attending our hospi¬ tal and have examined the clinical fea¬ tures and causes of acute bronchiolitis to understand how it might fit in the sche¬ ma of the ARI control program of the World Health Organization (WHO).14
PATIENTS AND METHODS This investigation was conducted in chil¬ dren less than 6 years old, in one of the two pediatrie units in our hospital, from Febru¬ ary 1985 through December 1987. The crite¬ ria for diagnosis of ARI included nasal dis¬
charge; cough; or noisy, labored, or rapid breathing, with or without fever. All chil¬ dren hospitalized with ARI of less than 7 days' duration, and the first 1 to 3 children with ARI ofless than 5 days in the alternateday outpatient clinics, were recruited with parental consent. The children were exam¬ ined by one of two study physicians (T.C. and E. A. F. S. ) who recorded the history and clin-
Downloaded From: http://archpedi.jamanetwork.com/ by a Princeton University User on 09/28/2013
ical data on standardized forms. Chest roentgenographs and blood culture were recom¬ mended for all children with clinical evidence of LRI (see below) but were taken at the discretion of the attending physicians who were also responsible for treatment, includ¬ ing antibiotics or oxygen. The study physi¬ cians recorded the treatment given and the outcome of illness in each child on the forms. Children with crepitations, bronchial breath sounds, wheeze, or radiologie abnor¬ malities in the lungs, as recorded by the study physicians, were diagnosed to have LRI. Among them, those with expiratory wheeze (either audible or auscultated) and with or without clinical (hyperresonance on percussion) or radiological evidence of air trapping in the lungs were diagnosed to have acute bronchiolitis.5'15 Children with history of wheezing were excluded from this
analysis. Nasopharyngeal secretion (NPS) was col¬ lected from every child using a sterile, dis¬ posable suction catheter with a mucus trap
(Delee, Mallinckrodt, NY) connected to a hand suction pump (Nalgene, Nalge Co, Rochester, NY). The mucus was washed into the trap with 3 mL of transport medium (Hank's balanced salt solution with 0.5% gel¬ atin and antibiotics). In addition, two throat swabs were collected into a sterile screw-cap tube containing 3 mL of transport medium. Within 30 minutes of collection these speci¬ mens were brought to the laboratory on ice. The throat swab extract and the diluted NPS were centrifuged at 1500 rpm for 10 minutes at 4°C. The cell platelet from NPS was washed in phosphate-buffered saline and smears prepared on glass slides. The smears were examined by the indirect fluorescent
antibody (IFA) staining technique,16 using an epifluorescent microscope and commercial reagents (Wellcome Diagnostics, Dartford, England) for RSV, influenza A and B, parainfluenza 1 and 3, adenovirus, and measles virus.
Samples with two or more cells with fluo¬ rescence
4+)
of 2 + or more (on a scale of 1 + to considered positive for virus
were
antigen.
The supernatant fluids from throat swab extract and NPS were inoculated into prima-
Table 2.—Symptoms and Signs in Children With Bronchiolitis
Table 1 .—Age Distribution of Children With Lower Respiratory Infections (LRIs) and With Bronchiolitis No. (%) of Children With
mo
No. of Children With LRI
0-5 6-11 12-17 18-23 24-29 30-35 >35 0-59
131 56 48 24 16 11 42 328
(49) (41) (31) (21) (19) (18) 2*(5) 114 (35)
Age,
Chest retraction Refusal to feed Fever Vomiting with cough
Bronchiolitis 64 23 15 5 3 2
Signs Respiratory rate
Fig
bronchiolitis.
ry bonnet monkey kidney and HEp-2-cell cul¬ ture tubes. They were observed daily for cytopathic effect (CPE) for 7 to 10 days. He-
madsorption test, using guinea pig red blood cells, was done on primary bonnet monkey kidney cells on the day that CPE was ob¬ served, or in the absence of CPE on day 5 and if negative, again on day 10. All cultures were passed a second time. Those not show¬ ing CPE or hemadsorption in both passages were considered as negative. In positive cul¬ tures, viral isolates were provisionally iden¬ tified on the basis of CPE pattern or hemad¬ sorption and confirmed by the IFA test on the cells as described earlier.16 Additional reagents for parainfluenza type 2 and adenovirus (National Bacteriological Institute, Stockholm, Sweden) and for herpes simplex virus (Syva Co, Palo Alto, Calif) were used for virus identification. For bacterial culture, blood was inoculated at the bedside into biphasic MacConkey me¬ dium, biphasic infusion broth, and brainheart infusion broth. The inoculated media were transported to the laboratory at room temperature and processed using previously described techniques.17 In one child with roentgenographic evidence of consolidation, postmortem lung puncture aspirate
cultured for bacteria. The student's t test and the used for statistical analysis.
(93.0)
106 98 21 89
(93.0) (86.0) (18.4) (78.1)
72 24 10
Hyperresonance on percussion Flaring of alae nasi Wheeze (audible
(63.2) (21.1) (8.8)
46 38
(40.4) (33.3)
stethoscope)* Grunting Cyanosis
29 21 6
(25.4) (18.4) (5.3)
Suprasternal Crepitations Temperature
1.—The age distribution of children with
was
2 test were
RESULTS
Among the 809 children with ARI in the study, 328 had LRI. Acute bron¬ chiolitis was diagnosed in 114 (35%) of them; 90 children were hospitalized, and 24 were treated as outpatients. The age distribution of children with LRI and the number of children with bronchiolitis among each age group is shown in Table 1. The age distribution
of
>37°C >38°C >38.5°C
with bronchiolitis is shown in Fig 1. All patients were less than 4 years old; 76% were infants and 17% were 1 year old. The mean (SD) age for the 90 hospitalized children was 6.2 (6.4) months and for the 24 outpatients, 14.5 (11.3) months. This difference was sta¬ cases
tistically significant (t 4.73, P50/min Retraction Subcostal Intercostal
*One was 40 months old and the other 42 months old.
a
No. (%) With Symptom/Sign
Symptom/Sign Symptoms Rapid breathing
without
*Wheeze, audible in all
on
auscultation,
was
present
cases.
=
toms in the 114 children with acute bronchiolitis are shown in Table 2. A history of rapid breathing, observed tachypnea, and subcostal retraction was present in over 90% of cases. Of the 8 children with respiratory rate of 50/min or less, 7 had subcostal retraction. Thus 113 of the 114 children with acute bron¬ chiolitis had either subcostal retraction or respiratory rate over 50/min, or both. Chest roentgenographs were ob¬ tained in 88 children; evidence of hyper¬ inflation was present in 69 (78%). Streaky opacities radiating from the hilum (interstitial infiltrate) was seen in 23 (26%), patchy opacities (alveolar in¬ filtrate) in 23 (26%), consolidation in 7 (8%), and atelectasis in 4 (5%). Of the 36 children in whom hyperinflation was di¬ agnosed clinically and in whom a chest roentgenograph was available, 31 (86%)
had
roentgenographic hyperinflation;
among the 69 children with
roentgeno¬
graphic hyperinflation, only 31 (45%) had a hyperresonant percussion note. The clinical symptoms and signs in
children with and without roentgeno¬ or alveolar infil¬ trate are shown in Table 3. Their frequencies were not significantly dif¬ ferent in the two groups. Of the 7 chil-
graphic consolidation
Downloaded From: http://archpedi.jamanetwork.com/ by a Princeton University User on 09/28/2013
dren with consolidation, 6 were infants and of the 23 with alveolar infiltrate, 19 were infants. Viruses were detected by culture or IFA test on NPS in 81 (71%) children with bronchiolitis as shown in Table 4. The most frequent was RSV, detected in 65 (57%) children. In 4 children, parainfluenza type 1, an adenovirus, an enterovirus, or an unidentified virus was present in addition to RSV. In one child, parainfluenza type 3 was isolated while parainfluenza type 1 was detected by IFA. Among the 23 children with alveo¬ lar infiltrate on roentgenography, RSV was detected in 9, parainfluenza type 3 in 2, parainfluenza type 1 and adenovi¬ rus in 1 each, and parainfluenza types 1 and 3 in 1. Of the 7 children with consoli¬ dation, RSV was detected in 4 children and parainfluenza virus type 3 in 1 child. In all 4 children with roentgenographic atelectasis, RSV was detected. The monthly distribution of bronchio¬ litis, the numbers positive for RSV, the mean maximum daily temperature, and the monthly rainfall are shown in Fig 2. Of the total of 114 cases, 83 (73%) oc¬ curred between August and November, coinciding with prevalence of RSV; all except 3 cases due to RSV occurred dur¬ ing these months. Bronchiolitis due to
Table 3.—Clinical Symptoms and Signs in Children with Bronchiolitis With and Without Roentgenographic Consolidation (RC) or Alveolar Infiltrate ( ) ( 88)* =
(%) of Children With RC or Al
No.
Symptom/Sign Symptoms Rapid breathing
Absent (n = 58)
Present(n 30) =
27
Chest retraction
28
Fever
19
Refusal to feed
18
Sign Respiratory rate >50/min
(90) (93) (63) (60)
55 54 43 45
(87) (97) 25 (83) Crepitations 9 (30) Temperature >38.5°C 11 (37) Flaring of alae nasi 10 (33) Grunting 2 (7) Cyanosis *The 2 test was applied to the differences in the proportions symptom/sign and none was statistically significant (P>.05). Subcostal retraction
26
55
29
54 48 12 17
9 4
Virus
RSV Parainfluenza 3 Parainfluenza 2 Parainfluenza 1
(95) (93) (83) (21) (29) (16) (7)
of children in the two groups with each
Table 4.—Viruses Detected in Children With Bronchiolitis No. of Children
(95) (93) (74) (78)
Testing
(n 114)* =
Positive for Virus
Culture
IFA of NPS
37
63
8 1
2
by Culture or IFA of NPS 65 10
1
0 1
1
Influenza A
0
1
Adenovirus
4
0
Enteroviruses
2
4 2
Unidentified virus
1
1
=
detected, Staphylococcus postmor¬ tem lung aspirate. RSV
(aerobic spore-formers, micrococci or Staphylococcus epidermidis). Two chil¬ dren had mixed gram negative bacilli in
two culture bottles that were considered to be contaminants; they re¬ covered from ARI without receiving any antibiotic. Blood cultures yielded no bacteria in the remaining 37 children. Febrile children were treated with antipyretics. Small frequent liquid oral feeds were offered to all; 45 (40%) chil¬ dren required intravenous fluids since they had difficulty in feeding. Oxygen was administered to 57 (50%) children and antibiotics to 60 (53%), including 39 (34%) who received both. One child who had a ventriculoseptal defect in addition to bronchiolitis was given ventilatory one or
support.
Downloaded From: http://archpedi.jamanetwork.com/ by a Princeton University User on 09/28/2013
was
aureus was
2 1
*Viruses were detected by cultures of throat swab or nasopharyngeal secretions (NPS), and indirect fluorescent antibody (IFA) staining of NPS. In 5 children more than one virus was detected. They included 4 children with respiratory syncytial virus (RSV) and parainfluenza type 1, adenovirus, enterovirus, or an unidentified agent. One child had parainfluenza virus types 1 and 3.
other viruses occurred sporadically throughout the year. The frequency of RSV infection and of bronchiolitis rose with the onset of the rainy season fol¬ lowing the very hot and dry months of May through July each year. The num¬ ber of cases declined from December through February, which are the cool¬ est months in this region. Blood cultures were obtained in 56 children with bronchiolitis, 37 of whom had received prior treatment from a physician, including 13 who were known to have received antibiotics. In one child with RSV detected by culture and by IFA of NPS, Salmonella typhi was iso¬ lated; this child was treated with chloramphenicol palmitate. The blood cul¬ ture of 16 children yielded skin flora
The mean (SD) duration of hospital¬ ization for the 90 hospitalized children with bronchiolitis was 4.3 (3.1) days; it was 4.5 (2.5) for those with RSV infec¬ tion, 3.2 (2.0) days for those with other virus infection, and 4.6 (4.6) days for those with no virus detected (P>.05). For those treated with antibiotics, the duration was 5.0 (3.6) days and for the rest, it was 3.3 (1.9) days (i 2.65, P= .01). Eight infants, 7 of whom were below 6 months of age and 4 of whom had RSV infection, died. All of them had been treated with antibiotics. Three children had additional complications that could have contributed to death. One was aged 1 month, had a very low birth weight and weighed only 1.9 kg when he developed bronchiolitis. An¬ other, aged 2 months, had diarrhea with severe dehydration in addition to bron¬ chiolitis. A third, aged 4 months, had a ventriculoseptal defect in addition to bronchiolitis. Of the remaining 5, 3 had roentgenographic consolidation and 2 had alveolar infiltrate; in 1 child aged 12 months with consolidation and in whom isolated from the
COMMENT
Bronchiolitis, diagnosed using
stan¬
dard clinical criteria,515 is common among infants with LRI attending our hospital, accounting for 49% of infants under age 6 months and 41% of infants aged 6 to 11 months. Most reports on severe LRI in developing countries have dealt with pneumonia.18,19 That bronchiolitis accounts for a large pro¬ portion of severe LRI even in the trop¬ ics has not been widely appreciated. Clinical signs such as tachypnea and chest retraction, which can be used by health workers to identify children with moderate and severe ARI, were pres¬ ent in most children with bronchiolitis. These signs are also present in children with pneumonia20,21; therefore, differen¬ tiation between pneumonia and bron¬ chiolitis would be difficult in the field. However, the syndrome of bronchiolitis can easily be recognized by physicians if they are taught its distinctive clinical presentation and predictable seasonal-
ity.
On roentgenography, alveolar infil¬ trate was seen in 23 and consolidation in
outbreaks.10 The rainy season in tropical
Asia, like the winter in the temperate climates, keeps people indoors and the resulting crowding may facilitate the spread of RSV infection. The associa¬ tion of crowding during festivals and
Fig 2.—The distribution of children with bronchiolitis by month, the number positive for respira¬ mean of maximum daily
tory syncytial virus (RSV), the monthly rainfall, and the monthly temperatures. 7 children with bronchiolitis. There¬ fore these children could also be diag¬ nosed as having pneumonia, illustrating the overlap between LRI syndromes. Among the 5 children with bronchiolitis who died and who did not have pre¬ existing disease, 3 had consolidation and 2 had alveolar infiltrate, signifying that these roentgenographic signs indi¬ cate the presence of severe disease. However, clinical signs and symptoms did not distinguish children who had roentgenographic consolidation or alve¬ olar infiltrate from those who did not. Demonstration of respiratory viruses in the upper respiratory tract is gener¬ ally accepted as evidence of their being the causative agents of lower respira¬ tory disease.15 Accordingly, viral caus¬ ative agents were found in 70% of chil¬ dren with bronchiolitis. The etiologic agents of bronchiolitis in this study were similar to those described in west¬ ern
countries.1"5 Respiratory syncytial
virus was the most common, followed by parainfluenza viruses. Although en-
teroviruses
were
detected in two chil¬
dren, one also had RSV detected by IFA, suggesting that bronchiolitis was
due to RSV. Enteroviruses may be iso¬ lated from the throats of healthy chil¬ dren and it is unlikely that they cause bronchiolitis. More than one virus was detected in four other children; in each child both the viruses detected are known to cause bronchiolitis. The occurrence of bronchiolitis and RSV infection showed a definite season¬ ally, coinciding with the rainy season. The association of bronchiolitis and of RSV infection with rainfall has been de¬ scribed from other tropical coun¬ tries. 10·11,22 A similar seasonal pattern has been described for the spread of varicella virus in this region.23 The in¬ crease in relative humidity and the fall in ambient temperatures during this season may be conducive to the survival of virus in the environment. On the oth¬ er hand, human behavior rather than environmental conditions have been suggested to be responsible for these
Downloaded From: http://archpedi.jamanetwork.com/ by a Princeton University User on 09/28/2013
outbreaks of RSV infection has also been reported from Calcutta, India.24 It is likely that both climatic factors and human behavior are responsible for the seasonal outbreaks. Evidence for bacterial infection was found in only two children, in both of whom RSV was also detected. One of them in whom S typhi was isolated in blood culture probably had two illness¬ es, RSV bronchiolitis and typhoid fe¬ ver. The other had dual infection with RSV and S aureus. Since blood culture has a low sensitivity in detecting bacte¬ rial LRI and it was not done in all the children, the frequency of bacterial in¬ fection in children with bronchiolitis could not be accurately estimated, but it appears to be low. Thus, bronchiolitis, even in tropical countries, is primarily due to viral infection. We believe that bacterial infection may occur secondari¬ ly in some children who will have roent¬ genographic evidence of pneumonia. Ideally, the use of antibiotics should be limited to these children. We suspect that all pneumonias in children with bronchiolitis may not be bacterial. Our study has not resolved this issue, except showing that RSV or parainfluenza virus infection was found in 19 of the 30 children with bronchiolitis and pneumonia. Antibiotic resistance is emerging as an alarming problem in many parts of the world, including de¬ veloping countries.25"27 Therefore, it is prudent to exercise caution in its wide¬ spread use, especially when rational ap¬ proaches may be available to curtail its use without detriment. The WHO ARI case management al¬ gorithm for village health workers sug¬ gests referral to hospital of children with cough, rapid breathing, and sub¬ costal retractions. Over 90% of our sub¬ jects with bronchiolitis had rapid respi¬ ration and subcostal retractions and would have been referred. This is ap¬ propriate, given the difficulty in dis¬ tinguishing children with bronchiolitis and roentgenographic pneumonia from those with bronchiolitis alone (Table 3). In settings with roentgenographic fa-
cilities, if pneumonia can be excluded, presumptive antibiotic therapy is not necessary in bronchiolitis. In settings without radiology, the WHO suggests
antibiotics for all children with bron¬ chiolitis and a respiratory rate over 50/min, because a proportion will have pneumonia.28 Thus, in our study, 93% would have qualified for antibiotic treatment (Table 2). We offer an alter¬ native approach and suggest the use of antibiotics only in infants under 1 year of age with bronchiolitis (76% in our study; Fig 1). This approach is based on our data that show that all deaths were in infants less than 1 year old, and 9% of infants with bronchiolitis did not survive. These conclusions are from our hospi¬ tal study. Obviously, community-based investigations of the epidemiologie and etiologic features and outcome of bron¬ chiolitis are urgently needed. This study was supported by grant 3-p-82-00T8 from the International Development Research Centre, Ottawa, Canada. We thank the regional center of the India Meteo¬ rological Department, Madras, for providing the meteorological data.
References 1. Wohl MEB. Bronchiolitis. In: Kendig EL, Chernick V, eds. Disorders of the Respiratory Tract in Children. Philadelphia, Pa: WB Saunders Co; 1983:283-294. 2. Bronchiolitis in infancy and childhood. Br Med J. 1980;280:428-429. 3. Wohl MEB, Chernick V. Bronchiolitis. Am Rev Respir Dis. 1978;118:759-781.
4. Henderson FW,
Clyde WA, Collier AM, Denetiologic and epidemiologic spectrum of bronchiolitis in pediatric practice. J Pediatr. 1979;95:183-190. 5. Denny FW, Collier AM, Henderson FW, Clyde WA. The epidemiology of bronchiolitis. Pe-
ny FW. The
diatr Res. 1977;11:234-236. 6. Khatua SP. Acute bronchiolitis: a study of 205 cases. Indian Pediatr. 1977;14:285-294. 7. Agarwal SC, Bardoloi JNS, Mehta S. Respiratory syncytial virus infection in infancy and childhood in a community in Chandigarh. Indian J Med Res. 1971;59:19-25. 8. Subramaniam S, Krishnan Prakash S, Sukumar S, Lakshminarayana CS. Respiratory syncytial virus infection in Madras. Indian J Med Res.
1980;72:317-326.
9. Vellayappan K, Teo J, Doraisingham S. Respiratory syncytial virus infections in children. J Singapore Paediatr Soc. 1982;24:69-77. 10. Sung RY, Murray HG, Chan RC, Davies DP, French GL. Seasonal patterns of respiratory syncytial virus infection in Hong Kong: a preliminary report. J Infect Dis. 1987;156:527-528. 11. Spence L, Barratt N. Respiratory syncytial virus associated with acute respiratory infections in Trinidadian patients. Am J Epidemiol. 1968;88:
257-266. 12. Berman S, Duenas A, Bedoya A, et al. Acute lower respiratory tract illness in Cali, Colombia: a two-year ambulatory study. Pediatrics. 1983;71: 210-218. 13. Hayes EB, Hurwitz ES, Schonberger LB, Anderson LJ. Respiratory syncytial virus outbreak on American Samoa. AJDC. 1989;143:316\x=req-\ 321. 14. World Health Organization. A programme for controlling acute respiratory infections in children: memorandum from a WHO meeting. Bull WHO. 1984;62:47-58. 15. Loda FA, Clyde WA, Glezen WP, Senior RJ, Sheaffer CI, Denny FW. Studies on the role of viruses, bacteria and M pneumoniae as causes of lower respiratory infections in children. J Pediatr.
1968;72:161-176.
16. Gardner PS, McQuillin J. Rapid Virus Diagnosis: Application of Immunofluorescence. 2nd ed. Stoneham, Mass: Butterworths; 1980:110-123. 17. Koshy G, Alex W, Menon T. An improved
Downloaded From: http://archpedi.jamanetwork.com/ by a Princeton University User on 09/28/2013
technique for blood culture. Indian J Med Res. 1981;70:733-738. 18. Pio A, Leowski J, TenDam HG. The magnitude of the problem of acute respiratory infections. In: Douglas RM, Kerby-Eaton E, eds. Acute Respiratory Infections in Childhood (Proceedings of an International Workshop, Sydney, August 1984). Adelaide, Australia: University of Adelaide; 1985:3-16. 19. Campbell H, Byass P, Lamont AC, et al. Assessment of clinical criteria of severe acute lower respiratory tract infection in children. Lancet.
1989;1:297-299. 20. Shann F, Hart K, Thomas D. Acute lower
respiratory tract infections in children: possible criteria for selection of patients for antibiotic therapy and hospital admission. Bull WHO. 1984;62:749\x=req-\
753. 21. Cherian T, John TJ, Simoes E, Steinhoff MC, John M. Evaluation of simple clinical signs for the diagnosis of acute lower respiratory infection. Lancet. 1988;2:125-128. 22. Monto AS, Johnson KM. A community study of respiratory infection in the tropics, I: description of the community and observations on the activity of certain respiratory agents. Am J Epidemiol.
1967;86:78-91.
23. Venkitaraman AR, John TJ. The epidemiology of varicella in staff and students of a hospital in the tropics. Int J Epidemiol. 1984;13:502-505. 24. Hillis WD, Cooper MR, Bang FB, Dey AK, Shah KV. Respiratory syncytial virus infection in children in West Bengal. Indian J Med Res.
1971;59:1354-1364. 25. Hansmann D, Devitt L, Miles H, Riley I.
Pneumococci relatively insensitive to penicillin in Australia and New Guinea. Med JAust. 1974;2:353\x=req-\ 356. 26. Jesudason MV, Lalitha MK, Koshi G. Changes in incidence of shigella subgroups and their antibiotic susceptibility pattern in Vellore, South India. J Trop Med Hyg. 1985;88:355-358. 27. Shann F, Gratten M, Montgomery J, Lupiwa T, Polume H. Haemophilus influenzae resistant to penicillin in Goroka. P NG Med J. 1982;25:23-25. 28. Respiratory Infections in Children: Management in Small Hospitals. Geneva, Switzerland: World Health Organization; 1988.
Mercy John,
Tropical South
MD; Eric A. F. Simoes, MD; Mark C. Steinhoff, MD; K. Chitra, MBBS; Raghupathy, MD; T. Jacob John, PhD, FRCP
MD; P.
\s=b\ In a prospective hospital-based study of 328 children under 5 years of age with acute lower respiratory infections, 114 (35%) were diagnosed to have acute bronchiolitis. Of them, 87 (76%) were less than 1 year and 107 (94%) were less than 2 years of age. Signs of severe lower respi-
ratory infections, namely tachypnea (respiratory rate greater than 50/min) and subcostal retraction, were present in 95% and 93%, respectively. Of 88 children of whom roentgenographs were taken, 30 (34%) had evidence of pneumonia. No clinical
signs discriminated between those with and without pneumonia. By culture or immunofluorescence antigen detection, viruses were found in 81 (71%) children with bronchiolitis; respiratory syncytial virus the most common agent, found in 65 Parainfluenza viruses were the next most common, found in 12 (11%). Most cases of bronchiolitis occurred in outbreaks during the rainy months of Auwas
(57%).
North America and Europe, acute bronchiolitis is the most common low¬ er respiratory tract infection (LRI) in infants.1,z Its epidemiologie and etiologic features are well described. It occurs as outbreaks in winter and spring; respira¬ tory syncytial virus (RSV) is the usual cause.2"5 Although known to occur in de¬ veloping countries in the tropics,6"18 its epidemiologie and etiologic features have not been examined in detail. One study described its clinical features and seasonality, but not the causes.6 Respi¬ ratory syncytial virus infection, preva-
Accepted for publication November 27,1989. From the Departments of Virology (Drs T. J. John and M. John) and Child Health (Drs Cherian, Raghupathy, and Chitra), Christian Medical College and Hospital, Vellore, Tamilnadu, India; De-
partments of Pediatrics and International Health, Hopkins University, Baltimore, Md (Dr Steinhoff); and the Department of Pediatrics, University of Colorado Medical Center, Denver (Dr Simoes). Reprint requests to Department of Virology, CMC Hospital, Vellore 632004, Tamil Nadu, India (Dr T.J.John).
Johns
India
gust through November, coinciding with
respiratory syncytial virus outbreaks. Although bacterial culture of blood was done in 56 children, no respiratory pathogen was isolated. In one child with bronchiolitis and consolidation, postmortem
lung aspirate yielded Staphylococcus aurThus, bronchiolitis is primarily a viral syndrome in this tropical region, just as it is in temperate regions. Eight (7%) children died (all were infants); 5 had roentgenographic pneumonia and the remaining had other abnormalities contributing
eus.
to death; all had been treated with antibiotics. Since one third of lower respiratory infections are bronchiolitis, and among infants under 1 year of age bronchiolitis comprises 47% of all lower respiratory infection cases, criteria for antibiotic management must take into account the availability of roentgenographic investigation.
(AJDC. 1990;144:1026-1030)
lent July through October, was found in 20% to 60% of cases of bronchiolitis in other studies.7"11 We have recently in¬ vestigated acute respiratory infections (ARIs) in children attending our hospi¬ tal and have examined the clinical fea¬ tures and causes of acute bronchiolitis to understand how it might fit in the sche¬ ma of the ARI control program of the World Health Organization (WHO).14
PATIENTS AND METHODS This investigation was conducted in chil¬ dren less than 6 years old, in one of the two pediatrie units in our hospital, from Febru¬ ary 1985 through December 1987. The crite¬ ria for diagnosis of ARI included nasal dis¬
charge; cough; or noisy, labored, or rapid breathing, with or without fever. All chil¬ dren hospitalized with ARI of less than 7 days' duration, and the first 1 to 3 children with ARI ofless than 5 days in the alternateday outpatient clinics, were recruited with parental consent. The children were exam¬ ined by one of two study physicians (T.C. and E. A. F. S. ) who recorded the history and clin-
Downloaded From: http://archpedi.jamanetwork.com/ by a Princeton University User on 09/28/2013
ical data on standardized forms. Chest roentgenographs and blood culture were recom¬ mended for all children with clinical evidence of LRI (see below) but were taken at the discretion of the attending physicians who were also responsible for treatment, includ¬ ing antibiotics or oxygen. The study physi¬ cians recorded the treatment given and the outcome of illness in each child on the forms. Children with crepitations, bronchial breath sounds, wheeze, or radiologie abnor¬ malities in the lungs, as recorded by the study physicians, were diagnosed to have LRI. Among them, those with expiratory wheeze (either audible or auscultated) and with or without clinical (hyperresonance on percussion) or radiological evidence of air trapping in the lungs were diagnosed to have acute bronchiolitis.5'15 Children with history of wheezing were excluded from this
analysis. Nasopharyngeal secretion (NPS) was col¬ lected from every child using a sterile, dis¬ posable suction catheter with a mucus trap
(Delee, Mallinckrodt, NY) connected to a hand suction pump (Nalgene, Nalge Co, Rochester, NY). The mucus was washed into the trap with 3 mL of transport medium (Hank's balanced salt solution with 0.5% gel¬ atin and antibiotics). In addition, two throat swabs were collected into a sterile screw-cap tube containing 3 mL of transport medium. Within 30 minutes of collection these speci¬ mens were brought to the laboratory on ice. The throat swab extract and the diluted NPS were centrifuged at 1500 rpm for 10 minutes at 4°C. The cell platelet from NPS was washed in phosphate-buffered saline and smears prepared on glass slides. The smears were examined by the indirect fluorescent
antibody (IFA) staining technique,16 using an epifluorescent microscope and commercial reagents (Wellcome Diagnostics, Dartford, England) for RSV, influenza A and B, parainfluenza 1 and 3, adenovirus, and measles virus.
Samples with two or more cells with fluo¬ rescence
4+)
of 2 + or more (on a scale of 1 + to considered positive for virus
were
antigen.
The supernatant fluids from throat swab extract and NPS were inoculated into prima-
Table 2.—Symptoms and Signs in Children With Bronchiolitis
Table 1 .—Age Distribution of Children With Lower Respiratory Infections (LRIs) and With Bronchiolitis No. (%) of Children With
mo
No. of Children With LRI
0-5 6-11 12-17 18-23 24-29 30-35 >35 0-59
131 56 48 24 16 11 42 328
(49) (41) (31) (21) (19) (18) 2*(5) 114 (35)
Age,
Chest retraction Refusal to feed Fever Vomiting with cough
Bronchiolitis 64 23 15 5 3 2
Signs Respiratory rate
Fig
bronchiolitis.
ry bonnet monkey kidney and HEp-2-cell cul¬ ture tubes. They were observed daily for cytopathic effect (CPE) for 7 to 10 days. He-
madsorption test, using guinea pig red blood cells, was done on primary bonnet monkey kidney cells on the day that CPE was ob¬ served, or in the absence of CPE on day 5 and if negative, again on day 10. All cultures were passed a second time. Those not show¬ ing CPE or hemadsorption in both passages were considered as negative. In positive cul¬ tures, viral isolates were provisionally iden¬ tified on the basis of CPE pattern or hemad¬ sorption and confirmed by the IFA test on the cells as described earlier.16 Additional reagents for parainfluenza type 2 and adenovirus (National Bacteriological Institute, Stockholm, Sweden) and for herpes simplex virus (Syva Co, Palo Alto, Calif) were used for virus identification. For bacterial culture, blood was inoculated at the bedside into biphasic MacConkey me¬ dium, biphasic infusion broth, and brainheart infusion broth. The inoculated media were transported to the laboratory at room temperature and processed using previously described techniques.17 In one child with roentgenographic evidence of consolidation, postmortem lung puncture aspirate
cultured for bacteria. The student's t test and the used for statistical analysis.
(93.0)
106 98 21 89
(93.0) (86.0) (18.4) (78.1)
72 24 10
Hyperresonance on percussion Flaring of alae nasi Wheeze (audible
(63.2) (21.1) (8.8)
46 38
(40.4) (33.3)
stethoscope)* Grunting Cyanosis
29 21 6
(25.4) (18.4) (5.3)
Suprasternal Crepitations Temperature
1.—The age distribution of children with
was
2 test were
RESULTS
Among the 809 children with ARI in the study, 328 had LRI. Acute bron¬ chiolitis was diagnosed in 114 (35%) of them; 90 children were hospitalized, and 24 were treated as outpatients. The age distribution of children with LRI and the number of children with bronchiolitis among each age group is shown in Table 1. The age distribution
of
>37°C >38°C >38.5°C
with bronchiolitis is shown in Fig 1. All patients were less than 4 years old; 76% were infants and 17% were 1 year old. The mean (SD) age for the 90 hospitalized children was 6.2 (6.4) months and for the 24 outpatients, 14.5 (11.3) months. This difference was sta¬ cases
tistically significant (t 4.73, P50/min Retraction Subcostal Intercostal
*One was 40 months old and the other 42 months old.
a
No. (%) With Symptom/Sign
Symptom/Sign Symptoms Rapid breathing
without
*Wheeze, audible in all
on
auscultation,
was
present
cases.
=
toms in the 114 children with acute bronchiolitis are shown in Table 2. A history of rapid breathing, observed tachypnea, and subcostal retraction was present in over 90% of cases. Of the 8 children with respiratory rate of 50/min or less, 7 had subcostal retraction. Thus 113 of the 114 children with acute bron¬ chiolitis had either subcostal retraction or respiratory rate over 50/min, or both. Chest roentgenographs were ob¬ tained in 88 children; evidence of hyper¬ inflation was present in 69 (78%). Streaky opacities radiating from the hilum (interstitial infiltrate) was seen in 23 (26%), patchy opacities (alveolar in¬ filtrate) in 23 (26%), consolidation in 7 (8%), and atelectasis in 4 (5%). Of the 36 children in whom hyperinflation was di¬ agnosed clinically and in whom a chest roentgenograph was available, 31 (86%)
had
roentgenographic hyperinflation;
among the 69 children with
roentgeno¬
graphic hyperinflation, only 31 (45%) had a hyperresonant percussion note. The clinical symptoms and signs in
children with and without roentgeno¬ or alveolar infil¬ trate are shown in Table 3. Their frequencies were not significantly dif¬ ferent in the two groups. Of the 7 chil-
graphic consolidation
Downloaded From: http://archpedi.jamanetwork.com/ by a Princeton University User on 09/28/2013
dren with consolidation, 6 were infants and of the 23 with alveolar infiltrate, 19 were infants. Viruses were detected by culture or IFA test on NPS in 81 (71%) children with bronchiolitis as shown in Table 4. The most frequent was RSV, detected in 65 (57%) children. In 4 children, parainfluenza type 1, an adenovirus, an enterovirus, or an unidentified virus was present in addition to RSV. In one child, parainfluenza type 3 was isolated while parainfluenza type 1 was detected by IFA. Among the 23 children with alveo¬ lar infiltrate on roentgenography, RSV was detected in 9, parainfluenza type 3 in 2, parainfluenza type 1 and adenovi¬ rus in 1 each, and parainfluenza types 1 and 3 in 1. Of the 7 children with consoli¬ dation, RSV was detected in 4 children and parainfluenza virus type 3 in 1 child. In all 4 children with roentgenographic atelectasis, RSV was detected. The monthly distribution of bronchio¬ litis, the numbers positive for RSV, the mean maximum daily temperature, and the monthly rainfall are shown in Fig 2. Of the total of 114 cases, 83 (73%) oc¬ curred between August and November, coinciding with prevalence of RSV; all except 3 cases due to RSV occurred dur¬ ing these months. Bronchiolitis due to
Table 3.—Clinical Symptoms and Signs in Children with Bronchiolitis With and Without Roentgenographic Consolidation (RC) or Alveolar Infiltrate ( ) ( 88)* =
(%) of Children With RC or Al
No.
Symptom/Sign Symptoms Rapid breathing
Absent (n = 58)
Present(n 30) =
27
Chest retraction
28
Fever
19
Refusal to feed
18
Sign Respiratory rate >50/min
(90) (93) (63) (60)
55 54 43 45
(87) (97) 25 (83) Crepitations 9 (30) Temperature >38.5°C 11 (37) Flaring of alae nasi 10 (33) Grunting 2 (7) Cyanosis *The 2 test was applied to the differences in the proportions symptom/sign and none was statistically significant (P>.05). Subcostal retraction
26
55
29
54 48 12 17
9 4
Virus
RSV Parainfluenza 3 Parainfluenza 2 Parainfluenza 1
(95) (93) (83) (21) (29) (16) (7)
of children in the two groups with each
Table 4.—Viruses Detected in Children With Bronchiolitis No. of Children
(95) (93) (74) (78)
Testing
(n 114)* =
Positive for Virus
Culture
IFA of NPS
37
63
8 1
2
by Culture or IFA of NPS 65 10
1
0 1
1
Influenza A
0
1
Adenovirus
4
0
Enteroviruses
2
4 2
Unidentified virus
1
1
=
detected, Staphylococcus postmor¬ tem lung aspirate. RSV
(aerobic spore-formers, micrococci or Staphylococcus epidermidis). Two chil¬ dren had mixed gram negative bacilli in
two culture bottles that were considered to be contaminants; they re¬ covered from ARI without receiving any antibiotic. Blood cultures yielded no bacteria in the remaining 37 children. Febrile children were treated with antipyretics. Small frequent liquid oral feeds were offered to all; 45 (40%) chil¬ dren required intravenous fluids since they had difficulty in feeding. Oxygen was administered to 57 (50%) children and antibiotics to 60 (53%), including 39 (34%) who received both. One child who had a ventriculoseptal defect in addition to bronchiolitis was given ventilatory one or
support.
Downloaded From: http://archpedi.jamanetwork.com/ by a Princeton University User on 09/28/2013
was
aureus was
2 1
*Viruses were detected by cultures of throat swab or nasopharyngeal secretions (NPS), and indirect fluorescent antibody (IFA) staining of NPS. In 5 children more than one virus was detected. They included 4 children with respiratory syncytial virus (RSV) and parainfluenza type 1, adenovirus, enterovirus, or an unidentified agent. One child had parainfluenza virus types 1 and 3.
other viruses occurred sporadically throughout the year. The frequency of RSV infection and of bronchiolitis rose with the onset of the rainy season fol¬ lowing the very hot and dry months of May through July each year. The num¬ ber of cases declined from December through February, which are the cool¬ est months in this region. Blood cultures were obtained in 56 children with bronchiolitis, 37 of whom had received prior treatment from a physician, including 13 who were known to have received antibiotics. In one child with RSV detected by culture and by IFA of NPS, Salmonella typhi was iso¬ lated; this child was treated with chloramphenicol palmitate. The blood cul¬ ture of 16 children yielded skin flora
The mean (SD) duration of hospital¬ ization for the 90 hospitalized children with bronchiolitis was 4.3 (3.1) days; it was 4.5 (2.5) for those with RSV infec¬ tion, 3.2 (2.0) days for those with other virus infection, and 4.6 (4.6) days for those with no virus detected (P>.05). For those treated with antibiotics, the duration was 5.0 (3.6) days and for the rest, it was 3.3 (1.9) days (i 2.65, P= .01). Eight infants, 7 of whom were below 6 months of age and 4 of whom had RSV infection, died. All of them had been treated with antibiotics. Three children had additional complications that could have contributed to death. One was aged 1 month, had a very low birth weight and weighed only 1.9 kg when he developed bronchiolitis. An¬ other, aged 2 months, had diarrhea with severe dehydration in addition to bron¬ chiolitis. A third, aged 4 months, had a ventriculoseptal defect in addition to bronchiolitis. Of the remaining 5, 3 had roentgenographic consolidation and 2 had alveolar infiltrate; in 1 child aged 12 months with consolidation and in whom isolated from the
COMMENT
Bronchiolitis, diagnosed using
stan¬
dard clinical criteria,515 is common among infants with LRI attending our hospital, accounting for 49% of infants under age 6 months and 41% of infants aged 6 to 11 months. Most reports on severe LRI in developing countries have dealt with pneumonia.18,19 That bronchiolitis accounts for a large pro¬ portion of severe LRI even in the trop¬ ics has not been widely appreciated. Clinical signs such as tachypnea and chest retraction, which can be used by health workers to identify children with moderate and severe ARI, were pres¬ ent in most children with bronchiolitis. These signs are also present in children with pneumonia20,21; therefore, differen¬ tiation between pneumonia and bron¬ chiolitis would be difficult in the field. However, the syndrome of bronchiolitis can easily be recognized by physicians if they are taught its distinctive clinical presentation and predictable seasonal-
ity.
On roentgenography, alveolar infil¬ trate was seen in 23 and consolidation in
outbreaks.10 The rainy season in tropical
Asia, like the winter in the temperate climates, keeps people indoors and the resulting crowding may facilitate the spread of RSV infection. The associa¬ tion of crowding during festivals and
Fig 2.—The distribution of children with bronchiolitis by month, the number positive for respira¬ mean of maximum daily
tory syncytial virus (RSV), the monthly rainfall, and the monthly temperatures. 7 children with bronchiolitis. There¬ fore these children could also be diag¬ nosed as having pneumonia, illustrating the overlap between LRI syndromes. Among the 5 children with bronchiolitis who died and who did not have pre¬ existing disease, 3 had consolidation and 2 had alveolar infiltrate, signifying that these roentgenographic signs indi¬ cate the presence of severe disease. However, clinical signs and symptoms did not distinguish children who had roentgenographic consolidation or alve¬ olar infiltrate from those who did not. Demonstration of respiratory viruses in the upper respiratory tract is gener¬ ally accepted as evidence of their being the causative agents of lower respira¬ tory disease.15 Accordingly, viral caus¬ ative agents were found in 70% of chil¬ dren with bronchiolitis. The etiologic agents of bronchiolitis in this study were similar to those described in west¬ ern
countries.1"5 Respiratory syncytial
virus was the most common, followed by parainfluenza viruses. Although en-
teroviruses
were
detected in two chil¬
dren, one also had RSV detected by IFA, suggesting that bronchiolitis was
due to RSV. Enteroviruses may be iso¬ lated from the throats of healthy chil¬ dren and it is unlikely that they cause bronchiolitis. More than one virus was detected in four other children; in each child both the viruses detected are known to cause bronchiolitis. The occurrence of bronchiolitis and RSV infection showed a definite season¬ ally, coinciding with the rainy season. The association of bronchiolitis and of RSV infection with rainfall has been de¬ scribed from other tropical coun¬ tries. 10·11,22 A similar seasonal pattern has been described for the spread of varicella virus in this region.23 The in¬ crease in relative humidity and the fall in ambient temperatures during this season may be conducive to the survival of virus in the environment. On the oth¬ er hand, human behavior rather than environmental conditions have been suggested to be responsible for these
Downloaded From: http://archpedi.jamanetwork.com/ by a Princeton University User on 09/28/2013
outbreaks of RSV infection has also been reported from Calcutta, India.24 It is likely that both climatic factors and human behavior are responsible for the seasonal outbreaks. Evidence for bacterial infection was found in only two children, in both of whom RSV was also detected. One of them in whom S typhi was isolated in blood culture probably had two illness¬ es, RSV bronchiolitis and typhoid fe¬ ver. The other had dual infection with RSV and S aureus. Since blood culture has a low sensitivity in detecting bacte¬ rial LRI and it was not done in all the children, the frequency of bacterial in¬ fection in children with bronchiolitis could not be accurately estimated, but it appears to be low. Thus, bronchiolitis, even in tropical countries, is primarily due to viral infection. We believe that bacterial infection may occur secondari¬ ly in some children who will have roent¬ genographic evidence of pneumonia. Ideally, the use of antibiotics should be limited to these children. We suspect that all pneumonias in children with bronchiolitis may not be bacterial. Our study has not resolved this issue, except showing that RSV or parainfluenza virus infection was found in 19 of the 30 children with bronchiolitis and pneumonia. Antibiotic resistance is emerging as an alarming problem in many parts of the world, including de¬ veloping countries.25"27 Therefore, it is prudent to exercise caution in its wide¬ spread use, especially when rational ap¬ proaches may be available to curtail its use without detriment. The WHO ARI case management al¬ gorithm for village health workers sug¬ gests referral to hospital of children with cough, rapid breathing, and sub¬ costal retractions. Over 90% of our sub¬ jects with bronchiolitis had rapid respi¬ ration and subcostal retractions and would have been referred. This is ap¬ propriate, given the difficulty in dis¬ tinguishing children with bronchiolitis and roentgenographic pneumonia from those with bronchiolitis alone (Table 3). In settings with roentgenographic fa-
cilities, if pneumonia can be excluded, presumptive antibiotic therapy is not necessary in bronchiolitis. In settings without radiology, the WHO suggests
antibiotics for all children with bron¬ chiolitis and a respiratory rate over 50/min, because a proportion will have pneumonia.28 Thus, in our study, 93% would have qualified for antibiotic treatment (Table 2). We offer an alter¬ native approach and suggest the use of antibiotics only in infants under 1 year of age with bronchiolitis (76% in our study; Fig 1). This approach is based on our data that show that all deaths were in infants less than 1 year old, and 9% of infants with bronchiolitis did not survive. These conclusions are from our hospi¬ tal study. Obviously, community-based investigations of the epidemiologie and etiologic features and outcome of bron¬ chiolitis are urgently needed. This study was supported by grant 3-p-82-00T8 from the International Development Research Centre, Ottawa, Canada. We thank the regional center of the India Meteo¬ rological Department, Madras, for providing the meteorological data.
References 1. Wohl MEB. Bronchiolitis. In: Kendig EL, Chernick V, eds. Disorders of the Respiratory Tract in Children. Philadelphia, Pa: WB Saunders Co; 1983:283-294. 2. Bronchiolitis in infancy and childhood. Br Med J. 1980;280:428-429. 3. Wohl MEB, Chernick V. Bronchiolitis. Am Rev Respir Dis. 1978;118:759-781.
4. Henderson FW,
Clyde WA, Collier AM, Denetiologic and epidemiologic spectrum of bronchiolitis in pediatric practice. J Pediatr. 1979;95:183-190. 5. Denny FW, Collier AM, Henderson FW, Clyde WA. The epidemiology of bronchiolitis. Pe-
ny FW. The
diatr Res. 1977;11:234-236. 6. Khatua SP. Acute bronchiolitis: a study of 205 cases. Indian Pediatr. 1977;14:285-294. 7. Agarwal SC, Bardoloi JNS, Mehta S. Respiratory syncytial virus infection in infancy and childhood in a community in Chandigarh. Indian J Med Res. 1971;59:19-25. 8. Subramaniam S, Krishnan Prakash S, Sukumar S, Lakshminarayana CS. Respiratory syncytial virus infection in Madras. Indian J Med Res.
1980;72:317-326.
9. Vellayappan K, Teo J, Doraisingham S. Respiratory syncytial virus infections in children. J Singapore Paediatr Soc. 1982;24:69-77. 10. Sung RY, Murray HG, Chan RC, Davies DP, French GL. Seasonal patterns of respiratory syncytial virus infection in Hong Kong: a preliminary report. J Infect Dis. 1987;156:527-528. 11. Spence L, Barratt N. Respiratory syncytial virus associated with acute respiratory infections in Trinidadian patients. Am J Epidemiol. 1968;88:
257-266. 12. Berman S, Duenas A, Bedoya A, et al. Acute lower respiratory tract illness in Cali, Colombia: a two-year ambulatory study. Pediatrics. 1983;71: 210-218. 13. Hayes EB, Hurwitz ES, Schonberger LB, Anderson LJ. Respiratory syncytial virus outbreak on American Samoa. AJDC. 1989;143:316\x=req-\ 321. 14. World Health Organization. A programme for controlling acute respiratory infections in children: memorandum from a WHO meeting. Bull WHO. 1984;62:47-58. 15. Loda FA, Clyde WA, Glezen WP, Senior RJ, Sheaffer CI, Denny FW. Studies on the role of viruses, bacteria and M pneumoniae as causes of lower respiratory infections in children. J Pediatr.
1968;72:161-176.
16. Gardner PS, McQuillin J. Rapid Virus Diagnosis: Application of Immunofluorescence. 2nd ed. Stoneham, Mass: Butterworths; 1980:110-123. 17. Koshy G, Alex W, Menon T. An improved
Downloaded From: http://archpedi.jamanetwork.com/ by a Princeton University User on 09/28/2013
technique for blood culture. Indian J Med Res. 1981;70:733-738. 18. Pio A, Leowski J, TenDam HG. The magnitude of the problem of acute respiratory infections. In: Douglas RM, Kerby-Eaton E, eds. Acute Respiratory Infections in Childhood (Proceedings of an International Workshop, Sydney, August 1984). Adelaide, Australia: University of Adelaide; 1985:3-16. 19. Campbell H, Byass P, Lamont AC, et al. Assessment of clinical criteria of severe acute lower respiratory tract infection in children. Lancet.
1989;1:297-299. 20. Shann F, Hart K, Thomas D. Acute lower
respiratory tract infections in children: possible criteria for selection of patients for antibiotic therapy and hospital admission. Bull WHO. 1984;62:749\x=req-\
753. 21. Cherian T, John TJ, Simoes E, Steinhoff MC, John M. Evaluation of simple clinical signs for the diagnosis of acute lower respiratory infection. Lancet. 1988;2:125-128. 22. Monto AS, Johnson KM. A community study of respiratory infection in the tropics, I: description of the community and observations on the activity of certain respiratory agents. Am J Epidemiol.
1967;86:78-91.
23. Venkitaraman AR, John TJ. The epidemiology of varicella in staff and students of a hospital in the tropics. Int J Epidemiol. 1984;13:502-505. 24. Hillis WD, Cooper MR, Bang FB, Dey AK, Shah KV. Respiratory syncytial virus infection in children in West Bengal. Indian J Med Res.
1971;59:1354-1364. 25. Hansmann D, Devitt L, Miles H, Riley I.
Pneumococci relatively insensitive to penicillin in Australia and New Guinea. Med JAust. 1974;2:353\x=req-\ 356. 26. Jesudason MV, Lalitha MK, Koshi G. Changes in incidence of shigella subgroups and their antibiotic susceptibility pattern in Vellore, South India. J Trop Med Hyg. 1985;88:355-358. 27. Shann F, Gratten M, Montgomery J, Lupiwa T, Polume H. Haemophilus influenzae resistant to penicillin in Goroka. P NG Med J. 1982;25:23-25. 28. Respiratory Infections in Children: Management in Small Hospitals. Geneva, Switzerland: World Health Organization; 1988.
