Yield from flexible bronchoscopy in children

Pediatric Pulmonology 23:261–269 (1997) Yield From Flexible Bronchoscopy in Children S. Godfrey, MD, PhD, FRCP,* A. Avital, MD, C. Maayan, MD, M. Rotschild, MD, and C. Springer, MD Summary. Flexible fiberoptic (FO) bronchoscopy can now be undertaken readily in children using topical anesthesia and light sedation and has largely supplanted rigid open tube (OT) bronchoscopy for diagnostic purposes. The present study examined the contribution of the FO bronchoscope to clinical management in children presenting with specific types of problems. We examined the first 200 consecutive flexible bronchoscopies performed in 1995 in children under 18 years of age (median age, 2.27 years). Indications for bronchoscopy were noisy breathing (26.5%), recurrent pneumonia (21.0%), suspected pneumonia in an immunocompromised patient (10.5%), atelectasis or bronchial toilet (12.5%), possible foreign body aspiration (13.0%), and miscellaneous other reasons (16.5%). Inspection of the airway was abnormal in 67.0% of all investigations and made a clinically meaningful contribution to management in 67.5%, especially in those with noisy breathing (98.1%), possible foreign body aspiration (100%), and atelectasis (76.0%). Bronchoalveolar lavage (BAL) cytology was abnormal in 80.4% of the 107 lavages, but contributed little to management except in those with recurrent pneumonia (73.8%). Bacteria were isolated in 26.6% of the 109 specimens cultured, but this finding rarely affected management. Fungi were isolated in 47.4% of the 19 lavages in the immunocomprised group. Together, inspection, BAL and microbiology contributed to management in a mean of 90.5% (range, 76.2–100%) of patients in the various groups. We concluded that a high yield of clinically meaningful information can be expected from FO bronchoscopy in children when coupled with BAL and microbiological studies of lavage fluid. Pediatr. Pulmonol. 1997; 23:261–269. © 1997 Wiley-Liss, Inc. Key words: bronchoscopy; children; bronchoalveolar lavage. INTRODUCTION It is now some 17 years since the first experimental version of an FO bronchoscope was developed that was small enough for routine use in children using sedation and with topical anesthesia.1 Prior to this all bronchoscopies in small children were undertaken with the rigid OT instrument under general anesthesia. Over this period there have also been a number of other important changes in pediatric respiratory medicine that have a bearing on bronchoscopic practice. The advent of computerized tomographic (CT) scanning of the lungs has virtually eliminated the need for bronchography, while BAL provides material for cytology and microbiological examination. The more aggressive treatment of hematological and other malignancies and HIV infection, as well as the introduction of bone marrow transplantation, has created a population of immunosuppressed children at risk of opportunistic pulmonary infections. Over a similar period our experience with and practice of pediatric bronchoscopy has to a large extent reflected these changes. Between January 1980 and December 1995 we undertook 2,270 bronchoscopies in children of which 705 were OT bronchoscopies, 1,432 were FO bronchoscopies, and 133 were for bronchography. However, whereas prior to 1987 virtually all were performed with the OT instru© 1997 Wiley-Liss, Inc. ment, in the period 1994–1995 FO bronchoscopies reached 95% (Fig. 1). The greater ease of performing FO bronchoscopy under topical anesthesia compared with OT bronchoscopy under general anesthesia has undoubtedly contributed to an overall increased use of endoscopy of the airway in children. The diagnostic effectiveness of pediatric FO bronchoscopy based on 1,000 investigations in young children was reviewed 10 years ago by Wood,2 who concluded that an endoscopic finding of direct relevance was obtained in 76% of cases. A more recent study of FO bronchoscopy in 50 older children by Raine and Warner3 concluded that the endoscopic findings were related to the indications for bronchoscopy in 86% of cases. Eight years ago we reviewed our experience with 364 OT bronchoscopies in children; at that time the removal of aspirated foreign bodies and bronchography were still Institute of Pulmonology, Hadassah University Hospital, Hebrew University-Hadassah Medical School, Jerusalem, Israel. *Correspondence to: Prof. S. Godfrey, Institute of Pulmonology, Hadassah University Hospital, POB 12000, Jerusalem 91120, Israel. Received 13 July 1996; accepted 19 December 1996. 262 Godfrey et al. Fig. 1. Numbers of bronchoscopies in children by type of procedure from 1980 to 1995. The first 200 consecutive fiberoptic bronchoscopies out of the total of 259 performed in 1995 were analyzed in the present study. Flex, flexible fiberoptic (FO) bronchoscopies under sedation and topical anesthesia; Rigid, rigid open tube (OT) bronchoscopies under general anaesthesia; Brongm, bronchoscopy (usually OT) performed for the purpose of bronchography. common indications for OT bronchoscopy.4 We noted that it might have been possible to use the FO instrument in the younger patients in whom congenital anomalies were often found. Given the changes that have occurred in techniques, in the patient mix, and in our familiarity with both FO and OT bronchoscopes, we decided to review our current practice to determine the contribution of bronchoscopy to management in children with different diagnostic problems. The present report is based on the first 200 consecutive FO bronchoscopies undertaken in children younger than 18 years of age in 1995. Since only 11 OT bronchoscopies were performed in the same period (5 for Abbreviations AT BACT BAL FB FO HIV IC MC NB OPPOR OT PCP RP Atelectasis/bronchial toilet Bacteriology Bronchoalveolar lavage Foreign body Flexible fiberoptic Human immunodeficiency virus Immunocompromised host Miscellaneous Noisy breathing Opportunistic infection Open tube Pneumocystis carinii Recurrent pneumonia foreign body aspiration, 4 for bronchial toilet, and 2 for suspected mass in the lung), these have not been included in the analysis. MATERIALS AND METHODS Data on all bronchoscopies performed by the pediatric staff of the Institute of Pulmonology, including demography, indications, endoscopic findings, BAL results, and complications, are stored in the departmental computer and were available for analysis. Data were analyzed for all the 200 consecutive FO bronchoscopies undertaken between January and November 1995 that were performed in 185 children younger than 18 years of age. Of the 185 children, 173 were investigated once, 9 were investigated twice, and 3 were investigated three times. The median age (interquartile range) was 2.27 years (range, 0.80–5.26) and of the 200 investigations, 29.0% were performed in children younger than 1 year of age, 26.5% in children between 1 and 3 years of age and 44.5% in children over 3 years of age. All bronchoscopies were performed for clinical indications. Bronchoscopy was performed in the bronchoscopy suite of the Institute or in the pediatric intensive care unit for patients admitted to that unit. The child was sedated with intravenous pethidine (up to a maximum of 2.0 mg/ kg) and midazolam (up to a maximum of 0.2 mg/kg) in Bronchoscopy in Children the presence of the parents who left the room after the child was drowsy. Topical anesthesia was obtained with lignocaine solution dripped into the nose and through the bronchoscope up to a maximum total dose of 5.0 mg/kg. The heart rate and arterial oxygen saturation were monitored continuously by pulse oximetry, and an experienced assistant was present throughout the procedure to monitor the cardiorespiratory status of the patient and administer medications. Our approach conforms to the guidelines for the management of pediatric patients under sedation for therapeutic procedures laid down by the American Academy of Pediatrics.5 Bronchoscopy was performed via the nasal route using either the Olympus BF3C10, BF3C20 or Pentax FB10X pediatric FO bronchoscopes with outside diameters of 3.5–3.6 mm, respectively. On a few occasions the ultrathin Olympus BFN20 was used when the other bronchoscopes could not be passed; an adult FO bronchoscope was used on 49 occasions in older children. Inspection of the airway included a search for any anatomical anomaly, abnormal mobility of the wall of the airway as in tracheomalacia, and evidence of localized or generalized inflammation as indicated by excessive secretions, edema, or erythema of the airway wall. All bronchoscopies were performed with continuous video recording, which enabled us to review the findings when entering the result into the computer database. BAL was performed when clinically indicated by wedging the bronchoscope in the relevant lung segment, lavaging with 3 aliquots of 1.0 ml/kg buffered isotonic saline and aspirating with gentle suction. In earlier investigations all three specimens were pooled, but in later investigations the second and third specimens were pooled and used for analysis. BAL fluid was gently mixed, and an aliquot was taken for total and differential cell counts. Total leukocyte numbers were counted in a standard hemocytometer. Slides for differential cell counts were prepared on a Shandon Cytospin 3 (Cheshire, England) using 100 ml BAL fluid stained with Diff-Quik (American Scientific, McGraw Park, IL). Two hundred cells were counted and scored as macrophages, neutrophils, lymphocytes, or eosinophils, expressed as percent of cells recovered. Aliquots of the fluid were gram stained and cultured for bacteria only if the squamous epithelial cell count was less than 10/low-power field (× 100). Criteria for defining a bacterial infection were a heavy growth of a single pathogen associated with a predominance of neutrophils in the BAL fluid.6 BAL samples from immunocompromised children were also examined microscopically for Pneumocystis carinii (Gomori-methenamine silver nitrate staining), Cytomegalovirus (hematoxylin and eosin and immunoperoxidase staining), and fungi (periodic acid-Schiff and silver staining). Analysis of the results was performed by dividing the 263 children into six groups according to the clinical indications for bronchoscopy as recorded before the results of the bronchoscopy were known. These six groups were as follows: Noisy breathing (NB): This group (53 investigations) was comprised of children in whom the primary indication was abnormal airway noise other than wheezing due to diffuse obstructive lung disease. In most cases the noise was suspected to originate from either the upper airway, and to be due to nasopharyngeal obstruction or laryngomalacia, or to originate from the lower airway due to congenital tracheal anomalies. Recurrent pneumonia (RP): This group (42 investigations) included children with unexplained persistent or recurrent pulmonary infiltrates consistent with pneumonia that had failed to resolve with appropriate antibiotic therapy. Children with known causes of pneumonia such as cystic fibrosis and those with immunodeficiency were not included in this group. Immunocompromised host (IC): This group (21 investigations) was made up of children receiving immunosuppressive chemotherapy for hematological or other malignancies, as well as children following bone marrow transplantation who developed unexplained pulmonary infiltrates, crepitations on auscultation, or hypoxia. One child was investigated following liver transplantation. Atelectasis/bronchial toilet (AT): This group (25 investigations) included children with radiological changes suggesting atelectasis or requiring removal of retained secretions. In five cases this followed cardiothoracic surgery, in two cases the child had cystic fibrosis, in two cases mucoid impaction due to bronchocentric granulomatosis was discovered, and for the rest there was no obvious etiology. Possible foreign body aspiration (FB): This group (26 investigations) comprised children in whom there was a possibility of foreign body aspiration, but this was by no means certain. Any child with clear radiological or clinical evidence of foreign body aspiration went straight to OT bronchoscopy. Miscellaneous causes (MC): This group (33 investigations) included those children who could not readily be classified in one of the other groups. There were seven children with problems related to a tracheostomy, seven with unexplained cough and a normal chest radiograph, six with unexplained generalized wheezing (possibly aspiration), four with possible hemoptysis, five with various congenital anomalies, and four with a variety of different problems. To determine whether or not the investigation contributed in a clinically meaningful manner to the management of the problem for which the bronchoscopy was performed, criteria were set up before the data were analyzed. The individual components of the investigation 264 Godfrey et al. could only be considered as contributing in terms of what was found, and there was no means of identifying falsenegative results (e.g., a negative BAL in a child with an opportunistic infection) with any certainty. The investigation was considered meaningful if it: 1. provided a diagnosis unobtainable by other reasonable means (e.g., the finding of tracheomalacia to explain noisy breathing), or 2. enabled treatment impossible by other means (e.g., removal of bronchial plugging causing atelectasis), or 3. prevented unnecessary further investigation (e.g., exclusion of the presence of a suspected aspirated foreign body), or 4. prevented unnecessary treatment (e.g., by excluding evidence of opportunistic infection in an immunocompromised host), or 5. suggested appropriate antibiotic therapy in suspected bacterial infection (e.g., in persistent or recurrent pneumonia with no structural cause). For the patients grouped by indication for bronchoscopy, the application of these criteria to inspection of the airway, BAL cytology, and microbiology were as follows: NB group: Inspection: meaningful if source of noise was located; BAL: no contribution to primary diagnosis of noisy breathing; microbiology: no contribution to primary diagnosis of noisy breathing. RP group: Inspection: meaningful if an obstructed bronchus was located or source of purulent secretion was found indicating infection; BAL: meaningful if cell count or percentage of neutrophil leukocytes was increased, suggesting infection; microbiology: meaningful if a pathogenic organism was cultured and served as a guide to antibiotic therapy, provided the BAL cell count or percentage of neutrophil leukocytes increased suggesting infection rather than contamination. IC group: Inspection: made no contribution to primary diagnosis in child with known immunodeficiency; BAL: meaningful if cell count or percentage of neutrophils was increased, suggesting infection; microbiology: meaningful if pathogenic organism was cultured as a guide to antibiotic therapy, provided the BAL cell count or percentage of neutrophils increased, suggesting bacterial infection rather than opportunistic infection, or if P. carinii, Cytomegalovirus, or a fungus was isolated. AT group: Inspection: meaningful if obstructed bronchus was located, source of purulent secretion was found, or atelectasis was relieved; BAL: meaningful if cell count or percentage of neutrophils was increased, suggesting infection; microbiology: meaningful if pathogenic organism was cultured and served as a guide to antibiotic therapy, provided the BAL cell count or percentage of neutrophils increased, suggesting infection rather than contamination. FB group: Inspection: was meaningfully positive if foreign body was present and meaningfully negative if foreign body was absent; BAL: was meaningfully negative if cell count or percentage of neutrophils was increased in absence of foreign body, indicating infection rather than foreign body as cause of problem; microbiology: was meaningful if a pathogenic organism was cultured and served as a guide to antibiotic therapy, provided the BAL cell count or percentage of neutrophils increased in absence of foreign body, indicating infection rather than foreign body as cause of problem. MC group: Inspection: was meaningful if explanation for problem was located; BAL: was meaningful if cell count or percentage of neutrophils was increased, suggesting infection; microbiology: was meaningful if pathogenic organisms were cultured and served as a guide to antibiotic therapy, provided the BAL cell count or percentage of neutrophils increased, suggesting infection and not contamination. When the bronchoscopy led to an additional diagnosis other than that for which it was performed, this finding was recorded separately. For analysis of the BAL results the upper limit of normal for the number of nucleated cells was taken as 250,000 cells/ml, the upper limit of the normal percentage of neutrophils as 6%, for lymphocytes 20%, and for eosinophils 0.5%. These values were based on our own observations and estimations from published results for normal children.7,8 RESULTS Findings on Inspection The age distribution and sex of the children in the various groups in the 200 investigations are given in Table 1 and the findings on inspection of the airway in Table 2. Overall, an abnormality on inspection was found in 67.0% of investigations, with the highest incidence of abnormality being in the NB (96.2%) and AT (76.0%) groups. In the other groups inspection alone was normal in some 40–58% of investigations. A congenital anomaly of the airway was found in 81.1% of investigations in the NB group but much less often in the other groups. The incidence of inflammatory changes was highest in the IC (42.9%) and AT (44.0%) groups. In the IC group inspection of the airway was almost always normal apart from inflammatory changes. Of the 26 investigations for possible foreign body aspiration, a foreign body was present in only 4; inflammatory changes were present in 5 instances and congenital anomalies in 2; inspection was totally normal in 15 instances. A coincidental finding of adenoid hypertrophy causing clinically meaningful airway obstruction was found in 7.5% of all investigations. Bronchoscopy in Children 265 TABLE 1—Details of Children Grouped by Indications for Bronchoscopy Group No. of investigations Boys/girls Median age (yr) No. < 1 year No. 1–3 years No. > 3 years NB RP IC AT FB MC Total 53 30:23 0.5 36 5 12 42 29:13 3.5 2 15 25 21 11:10 4.2 3 5 13 25 15:10 1.4 10 7 8 26 11:15 2.5 1 13 12 33 18:15 4.0 6 8 19 200 114:86 2.3 58 53 89 TABLE 2—Findings on Observation as a Percentage of Each Group and as a Percentage of the Total Number of Bronchoscopies Group NB RP No. of investigations Findings Any abnormality Inflammation Foreign body Congenital larynx Congenital trachea Congenital bronchi Other congenital Any congenital 53 42 96.2 3.8 0 35.8 43.4 3.8 7.5 81.1 52.4 23.8 0 0 4.8 11.9 0 21.4 IC AT FB MC Total 21 25 26 33 200 52.4 42.9 0 0 0 0 9.5 0 76.0 44.0 0 0 4.0 24.0 0 32.0 42.3 19.2 15.4 0 3.8 3.8 0 7.7 60.6 18.2 0 0 3.0 21.2 15.2 24.2 The total number of observations on inspection in each group shown in Table 2 ranged from 100 to 121% since more than one finding was noted in some patients. Findings From Bronchoalveolar Lavage The results of cytological studies of BAL fluid grouped by indications are given in Table 3. BAL was not performed in all investigations, and the numbers of lavages in each group are given in the table. Overall, an abnormality in the BAL cytology was found in 80.4% of the 107 lavages performed, and abnormalities were common in all groups. An increase in the total BAL cell count and in the polymorphonuclear leukocyte percentage was very common in the AT group (88.9 and 77.8%, respectively), and the polymorphonuclear leukocyte percentage was also high in the lavages performed in the FB group (88.9%). The total cell count was lowest in the NB group. An excess of lymphocytes was found in one-third of the lavages in the RP group, but was uncommon in the other groups. In the AT group an excess of eosinophils was found in three of the nine lavages, one of which was from a child with bronchocentric granulomatosis and mucoid impaction. Findings From Microbiology of BAL Fluid The results of the microbiological investigations and the numbers of these investigations in each group are given in Table 4. Overall, bacteria were cultured in 26.6% of the 109 lavages, the greatest incidence being from the FB group (60.0%), with positive cultures being 67.0 21.5 2.5 9.5 14.0 10.5 5.5 35.0 much less common in all the other groups. Many of the children were receiving antibiotics at the time of the bronchoscopy, but the details were not recorded on the report and were not available for analysis. Viral, P. carinii, and fungal studies were undertaken in all 19 investigations of children from the IC group but not from the other groups. Neither viruses nor P. carinii were isolated from any of these studies, but fungi were identified by special staining in nine of the samples (47.4%) and by culture from three of these specimens. Aspergillus fumigatus was identified in eight samples and Candida in one. Clinical Significance of Inspection, BAL, and Microbiology As described under Materials and Methods, criteria were established for each group to establish whether or not the findings on inspection of the airway, BAL cytology, or microbiology contributed to management in a clinically meaningful way. The results of these analyses for each group are shown in Table 5. Overall, the findings on observation were clinically meaningful in 67.5% of all investigations, BAL cytology in 36.5% of all bronchoscopies (68.2% of those in which BAL was performed), bacterial culture in 13.0% of all bronchoscopies (23.9% of those in which bacterial culture was performed), and fungal isolation in 4.5% of all bronchoscopies (47.4% of those in which opportunistic infection was sought). When all aspects of the study were considered together, i.e., observation, cytology, and bacteriology, the overall incidence of a clinically meaningful bronchoscopy was 90.5% of all bronchoscopies. 266 Godfrey et al. TABLE 3—Findings in BAL Fluid as a Percentage of Each Group and as a Percentage of all BAL Studies Group NB RP No. BAL studies BAL Any abnormality Excess nucleated cells Excess neutrophils Excess lymphocytes Excess eosinophils 14 39 17 IC 57.1 14.3 35.7 7.1 0 89.7 48.7 59.0 33.3 23.1 58.8 35.3 41.2 5.9 0 AT FB MC Total 9 9 19 107 100 88.9 77.8 0 33.3 88.9 44.4 88.9 11.1 44.4 84.2 52.6 63.2 5.3 5.3 80.4 45.8 57.9 15.9 15.9 TABLE 4—Findings From Microbiological Studies of BAL Fluid as a Percentage of Each Group and as a Percentage of the Total Number of Microbiological Studies Group Bacteriology No. of studies Culture—any positive Opportunistic infections No. of studies Viruses—any positive P. carinii—any positive Fungi—any positive NB RP IC AT FB MC Total 8 25.0 38 23.7 20 25.0 14 21.4 10 60.0 19 21.1 109 26.6 — — — — — — — — 19 0 0 47.4 — — — — — — — — — — — — 19 0 0 47.4 TABLE 5—Clinically Meaningful Contribution of Findings, BAL and Microbiology as a Percentage of Number of Investigations in Each Group and as a Percentage of the Total Number of Bronchoscopies in the Group and of all Bronchoscopies Group No. of investigations Meaningful inspection No. of BAL studies Meaningful BAL No. of BACT studies Meaningful BACT No. of OPPOR studies Meaningful virus Meaningful PCP Meaningful fungus Any meaningful result NB 53 98.1 14 0 8 0 — — — — 98.1 RP 42 45.2 39 73.8 38 21.4 — — — — 78.6 There were marked differences between the groups as far as clinical significance was concerned. The findings on inspection of the airway were clinically meaningful in 98.1% of the NB and 100% of the FB group, but in none of those from the IC group, and in only 45.2% of those from the RP group. BAL cytology was most often contributory in the RP group (73.8% of the total studies in the group), but made little or no contribution in the NB and FB groups. The clinical significance of bacterial isolation was low in all groups, the greatest incidence being 21–24% of the total studies in each of the RP, IC, and FB groups where the isolation established the diagnosis and resulted in appropriate antibiotic therapy. Viruses and P. carinii were not isolated from any of the 19 studies in the IC group but most of these patients were receiving prophylaxis against P. carinii infection. Fungi were isolated and contributed meaningfully to management in 42.9% of the total studies in the IC group. Using all modalities of the investigation, i.e., observation, cytology, and mi- IC 21 0 17 42.9 20 23.8 19 0 0 42.9 76.2 AT 25 76.0 9 36.0 14 12.0 — — — — 96.0 FB 26 100 9 23.1 10 23.1 — — — — 100 MC 33 57.6 19 54.5 19 9.1 — — — — 90.9 Total 200 67.5 — 36.5 — 13.0 — 0 0 4.5 90.5 crobiology, the incidence of a clinically meaningful contribution from the bronchoscopy as a percentage of the total studies in the group was highest in the NB and FB groups (98.1 and 100%, respectively), lowest in the IC group (76.2%), and between 78% and 96% in the other groups. Complications A record was kept of all complications, both during the bronchoscopy and during the postoperative period until the child was discharged from the recovery area— usually 3–4 hours after the procedure. Hypoxia requiring an increase in the concentration of inspired oxygen or a temporary pause in the bronchoscopy was encountered on 27 occasions (13.5%). Patient distress despite medication was encountered on five occasions during bronchoscopy, (2.5%) and significant bleeding (not requiring active intervention other than gentle suctioning) Bronchoscopy in Children on two occasions (1.0%). After bronchoscopy, transient hypoxia was encountered on two occasions (1.0%) and stridor that subsided spontaneously on four occasions (2.0%); one child who also underwent a transbronchial biopsy during the bronchoscopy developed a transient pneumothorax that subsided spontaneously. Transient fever was reported occasionally after BAL, but as the children were usually at home, we had no way of verifying the incidence of this complication. None of these complications resulted in any long-term adverse effects. One child died within 24 hours of the procedure, but she was terminally ill with severe cardiopulmonary insufficiency following cardiac surgery and her death was unrelated to the bronchoscopy. DISCUSSION This study has shown that the overall clinically meaningful contribution from observation, BAL, and microbiological studies at fiberoptic bronchoscopy in children was 90.5%. The present audit was undertaken because of the changes that have occurred in the practice of pediatric bronchoscopy and respiratory medicine since our review of OT rigid bronchoscopy.4 In addition, the mix of patients referred to our department has changed, with fewer patients suspected of aspirating foreign bodies and more immunocompromised children with suspected opportunistic infections. In our previous review of 364 OT bronchoscopies, we concluded that only some 30–40% of investigations could have been performed with an FO flexible instrument.4 At the present time we are performing over 250 bronchoscopies a year, of which 96% are with the FO instrument. Although the frequency of bronchoscopy for the diagnosis of foreign body aspiration fell from 31 to 13% between our previous analysis and the present analysis, this cannot account for the difference in preference for the FO bronchoscope.4 Considering that our staff has extensive experience with both OT and FO bronchoscopy in children, this finding suggests that there has indeed been a learning process that has resulted in an elective shift to the use of the FO bronchoscope. It is difficult to compare the present review of our experience with the FO bronchoscope in children with other published series because of the difference in mix of patients and ages, as well as the different ways in which the results have been presented. For example, in the series of both Fitzpatrick et al.9 and Wood,2 bronchoscopy was performed for the purposes of inspecting a tracheostomy site in 10–15% of cases, while this indication occurred only 7 times (3.5%) in our series and once (2.0%) in the series of Raine and Warner,3 which clearly reflects hospital practice. On the other hand, there are some consistent observations between the series. Thus, the frequency of stridor (noisy breathing) as an indication 267 for FO bronchoscopy was 32 and 25% in the published series and 26.5% in the present series.2,9 In fact, in all series that include a high proportion of young children, noisy breathing or stridor is a common indication for bronchoscopy, even when OT bronchoscopy is used,10 while this is uncommon in older children.3 Taking all indications related to an abnormal chest radiograph together (the RP and AT groups in the present study) the frequency of this indication in published series was 30 and 25%, compared with 33.5% in the present series.2,9 Bronchoscopy for pulmonary complications in immunocompromised children is more common now than in earlier series and obviously depends on the interests of the particular hospital or department. Thus, while this indication does not even appear in the very large series published by Wood2 10 years ago, it comprised 22% of indications from a tertiary referral center in London3 and 10.5% in our series from a university teaching hospital. These patients tend to be older (Table 1) and hence are more common in series including a high proportion of older children.3 Bronchoscopy for suspected foreign body aspiration is a particularly common and important problem in the management of pediatric respiratory disease, especially in boys, since the peak age for this problem is between 1 and 3 years.4,11 Rarely can a foreign body be removed safely with an FO bronchoscope at this age; therefore, whenever the clinical or radiological picture strongly suggests that a foreign body has been aspirated, an OT bronchoscopy under general anesthesia should be performed. However, there are many situations in which the picture is far from clear and if such patients are carefully selected, a negative FO bronchoscopy under sedation and topical anesthesia can provide a definitive diagnosis and avoid the need for the OT bronchoscopy when there is no foreign body present. This was the case in 22 of the 26 children in the present series, and in only 4 was a foreign body present that required subsequent OT bronchoscopy. This is almost identical to the observation of Wood,2 who found nine foreign bodies in 48 bronchoscopies for possible aspiration. However, clinical judgment must be applied to select only those patients in whom the probability of foreign body aspiration is low. In a separate analysis of 121 of our OT bronchoscopies in children in whom a foreign body was found and removed, 66% of the children were boys and 69% were younger than 3 years old. In the present series 42% were boys and 54% were under 3 years old, suggesting a clinical bias in favor of FO bronchoscopy in those children who did not fit the typical pattern of those likely to have aspirated a foreign body. The use of BAL as an additional diagnostic tool during pediatric bronchoscopy is a relatively new development, and even normal data are few.7,8 There appear to be no previous reports of the abnormalities in BAL cytology 268 Godfrey et al. found at bronchoscopy in children presenting with different problems. The results in Table 3 show that an increase in total nucleated cell count and in the percentage of neutrophils was a common finding but did not really distinguish between the groups. Undoubtedly, part of the explanation for this finding and the low incidence of positive bacterial cultures in most groups is the fact that the children were often receiving or had recently received antibiotics. We routinely sought fat-laden macrophages in children with suspected aspiration, but neither in these patients, nor in a subsequent formal study, did we find this to be helpful. The detection of opportunistic pulmonary infections by the examination of BAL fluid has been documented in children. Raine and Warner3 identified seven specific organisms in their 11 patients with presumed opportunistic infection (P. carinii twice, Candida albicans twice, cytomegalovirus twice, and measles once). McCubbin et al.12 reported the results of BAL studies in 27 children who were bone marrow transplant recipients. They isolated an opportunistic pathogen in 15 (52%) of the investigations, with cytomegalovirus being the most common pathogen.5 P. carinii was isolated twice and fungi on three occasions. Stokes et al.13 bronchoscoped and performed BAL in older children (median age, 17 years) with malignant diseases who developed new pulmonary infiltrates and obtained opportunistic pathogens in 28% of the lavages, the most common being fungi (10% of lavages). They felt their low yield was due to the use of antibiotics and antifungal agents. Riedler et al.14 investigated cytology and microbiology in lavages from 41 procedures in children after heart-lung transplantation, in other immunocompromised patients, and in those with suspected infection or interstitial disease. They had a higher yield of positive microbiological results but did not say whether their patients were receiving antibiotics. Higher total or neutrophil cell counts were found in those with evidence of infection. Interestingly, they obtained evidence of infection in 55% of their 11 studies in immunocompromised patients that is very similar to the incidence of 63% (either bacterial or fungal) in the 19 studies in the present series. In the present series we did not isolate P. carinii or viruses, but we did isolate a fungus in nearly half of the immunocompromised patients. This reflects the management policy of the treating physicians who referred their patients for bronchoscopy concerning the prevention of cytomegalovirus and P. carinii infections. The major objective of the present study was to try to determine whether bronchoscopy, BAL, and microbiology contributed in a meaningful way to the clinical management of the patients. In many cases it is likely that bronchoscopy is performed in children in whom the presumed diagnosis is fairly obvious, but bronchoscopy is needed to be sure that an alternative diagnosis requiring different management is not being missed. For example, a child with stridor from early infancy is likely to have laryngomalacia that will subside spontaneously over the first year or so of life, but it could also be due to a much more serious problem such as subglottic stenosis or an hemangioma. While other investigators have given figures for the yield from bronchoscopy in children, they did not explicitly state the criteria used to judge the usefulness of the procedure, and this makes comparisons with the present series difficult. In his very extensive series of 1,000 FO bronchoscopies in children, Wood2 suggested that a relevant endoscopic diagnosis was obtained in 76% of investigations, but this was apparently by inspection alone, as BAL results were not reported. Despite the somewhat different mix of patients, this figure of Wood2 is quite similar to the 68% of clinically meaningful results obtained by inspection in the present study. Raine and Warner3 claimed that their bronchoscopies, which included BAL in some patients, yielded a relevant diagnosis in 86% of patients. The overall incidence of clinically meaningful information obtained in the present study was 90.5%, using the information provided by inspection, BAL, and microbiology. The relative contributions of inspection, BAL cytology, and microbiology cannot be estimated because this was not a controlled trial and BAL was only undertaken in those patients in whom we were seeking evidence of possible infection. At the time of the bronchoscopy BAL was performed in some children with noisy breathing in whom we subsequently considered that this procedure did not contribute to the primary objective of the investigation (see criteria in Materials and Methods). However, these were patients in whom infection was considered as a possible complication of the airway anomaly causing the noise, and indeed evidence of infection was found in some of them (Tables 3 and 4). Since this was essentially a secondary diagnosis, it was not considered to contribute meaningfully to the elucidation of the source of the noisy breathing, but we believe the procedure was justified. The similarity in the incidence of obtaining clinically meaningful information between the groups (Table 5) suggests that our decision to perform bronchoscopy for these indications was reasonable. It is true that the yield from inspection was low in recurrent pneumonia and immunocompromised groups, but the yield was improved by adding BAL and microbiological studies, and these could not have been undertaken without bronchoscopy. In conclusion, the present study included children in whom the indications for bronchoscopy were within the guidelines for pediatric bronchoscopy laid down by the American Thoracic Society and by our predefined criteria.15 The procedures contributed to management in a clinically meaningful fashion in 90.5% of investigations. Bronchoscopy in Children REFERENCES 1. Wood RE, Sherman JM. Pediatric flexible bronchoscopy. Ann Ontol Rhinol Laryngol. 1980; 89:414–416. 2. Wood RE. The diagnostic effectiveness of the flexible bronchoscope in children. Pediatr Pulmonol. 1985; 1:188–192. 3. Raine J, Warner JO. Fibreoptic bronchoscopy without general anaesthetic. Arch Dis Child. 1991; 66:481–484. 4. Godfrey S, Springer C, Maayan C, Avital A, Vatashky E, Beilin B. Is there a place for rigid bronchoscopy in the management of pediatric lung disease? Pediatr Pulmonol. 1987; 3:179–184. 5. Committee on Drugs. Guidelines for monitoring and management of pediatric patients during and after sedation for diagnostic and therapeutic procedures. Pediatrics. 1992; 89:1110–1115. 6. Springer C, Avital A, Noviski N, Maayan C, Ariel I, Mogle P, Godfrey S. Role of infection in the middle lobe syndrome. Arch Dis Child. 1992; 67:592–594. 7. Riedler J, Grigg J, Stone C, Tauro G, Robertson CF. Bronchoalveolar lavage cellularity in healthy children. Am J Respir Crit Care Med. 1995; 152:163–168. 8. Ratjen F, Bredendiek M, Zheng L, Brendel M, Costabel U. Lymphocyte subsets in bronchoalveolar lavage fluid of children without bronchopulmonary disease. Am J Respir Crit Care Med. 1995; 152:174–178. 269 9. Fitzpatrick SB, Marsh B, Stokes D, Ko-Pen W. Indications for flexible fiberoptic bronchoscopy in pediatric patients. Am J Dis Child. 1983; 137:595–597. 10. Puhakka H, Kero P, Erkinjuntti M. Pediatric bronchoscopy during a 17-year period. Int J Pediatr Otorhinolaryngol. 1987; 13:171– 180. 11. Banerjee A, Rao KS, Khanna SK, Narayanan PS, Gupta BK, Sekar JC, Retnam CR, Nachiappan M. Laryngo-tracheo-bronchial foreign bodies in children. J Laryngol Otol. 1988; 102:1029–1032. 12. McCubbin MM, Trigg ME, Hendricker CM, Wagener JS. Bronchoscopy with bronchoalveolar lavage in the evaluation of pulmonary complications of bone marrow transplantation in children. Pediatr Pulmonol. 1992; 12:43–47. 13. Stokes DC, Shener JL, Parham D, Bozeman PM, Marienchek W, Mackert PW. Role of flexible bronchoscopy in the diagnosis of pulmonary infiltrates in pediatric patients with cancer. J Pediatr. 1989; 115:561–567. 14. Riedler J, Grigg J, Robertson CF. Role of bronchoalveolar lavage in children with lung disease. Eur Respir J. 1995; 8:1725–1730. 15. Green GG, Eisenberg J, Leong A, Nathanson I, Schnapf BM, Wood RE. Flexible endoscopy of the pediatric airway. Am Rev Respir Dis. 1992; 145:233–235.