- Email: [email protected]
The Yield of Flexible Fiberoptic Bronchoscopy in Pediatric Intensive Care Patients
The Yield of Flexible Fiberoptic Bronchoscopy in Pediatric Intensive Care Patients* Dan Bar-Zohar, MD; and Yakov Sivan, MD
Study objective: To evaluate the contribution of flexible fiberoptic bronchoscopy (FFB) and BAL to the clinical management of patients in a pediatric ICU (PICU). Setting and design: A retrospective study based on medical records in a six-bed pediatric ICU of a tertiary care children’s hospital serving as a referral center for airway surgery. Patients and participants: One hundred consecutive infants and children hospitalized in a PICU, who underwent FFB with or without BAL. Measurements and results: One hundred fifty-five procedures were performed, for the following causes: search for airways anatomic pathologies (114 of 155 procedures, 74%), including 55 procedures during the perioperative period of airway surgery; treatment of atelectasis (35 of 155 procedures, 22.5%); and BAL (30 of 155 procedures, 19%). Thirty-five percent of procedures had more than one cause. Airway pathology was observed in 79 of 114 procedures (69%). Management changed from conservative to surgical in 44 of 114 procedures (39%). In airway surgery cases, reoperation subsequent to postoperative FFB took place in 35%. BAL results changed antimicrobial treatment in 15 of 30 cases, with clinical improvement in 10 of 30 cases (33%). Treatment of atelectasis was successful in 26 of 35 cases (74.3%). No procedure-related mortality, life-threatening complications, or significant changes in patient status occurred. Conclusions: FFB is an important and safe procedure in very sick infants and children with a variety of respiratory diseases, and significantly contributes to their management. FFB should be considered to be a PICU staff expertise. (CHEST 2004; 126:1353–1359) Key words: bronchoscopy; children; infants; pediatric ICU Abbreviations: BALF ⫽ BAL fluid; ETT ⫽ endotracheal tube; FFB ⫽ flexible fiberoptic bronchoscopy; Fio2 ⫽ fraction of inspired oxygen; PICU ⫽ pediatric ICU
fiberoptic bronchoscopy (FFB) is freF lexible quently used in the assessment and treatment of
infants and children with a variety of pediatric respiratory diseases.1–3 FFB with or without BAL is particularly important in the diagnosis and treatment of specific respiratory problems in critically ill infants and children hospitalized in the pediatric ICU (PICU). These comprise situations such as segmental lung collapse, pulmonary infections including community-acquired and ventilator-associated pneumonia, pulmonary infections in immunocompromised hosts, and pulmonary bleeding. In addition,
*From the Pediatric Intensive Care Unit, Dana Children’s Hospital, Tel-Aviv Sourasky Medical Center, Tel-Aviv University Sackler Faculty of Medicine, Tel-Aviv, Israel. Manuscript received January 5, 2004; revision accepted May 26, 2004. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (e-mail: [email protected]). Correspondence to: Yakov Sivan, MD, Pediatric Intensive Care, Dana Children’s Hospital, Tel-Aviv Medical Center, 6, Weizman St, Tel-Aviv 64239, Israel; e-mail: [email protected] www.chestjournal.org
FFB has a special role in airway problems in very sick pediatric patients. It may be crucial in intubation of the difficult airways and in the assessment of airway anomalies. With the evolution of airway surgical techniques for treating various compromising and life-threatening congenital airway anomalies, bedside FFB has becomes an indispensable clinical tool in the immediate perioperative PICU course. At present, there is only little information on the yield of FFB and BAL in the PICU. Most previously reported studies that analyzed the contribution of pediatric FFB originated from cohorts of noncritically ill patients with a variety of airway and lung diseases who were treated under the pediatric pulmonary services outside the PICU. These included ambulatory patients admitted to the hospital only for the procedure and noncritical patients hospitalized in the general pediatric and pediatric surgery wards.4,5 The largest pediatric series5 published included 2,836 patients. Most of the procedures were performed in the PICU. Nevertheless, CHEST / 126 / 4 / OCTOBER, 2004
1353
from the list of indications it is apparent that a substantial number were not true “PICU patients” and were admitted to the PICU only for FFB. At present, the only study6 in the English-language literature that specifically studied the application of FFB in the PICU was published in 1988, before BAL became widely accepted and airway surgery cases were not included. Hence, because PICU patients differ from other patient groups in terms of systems stability and the underlying problems, they should be evaluated separately for clinical benefit, indications, contraindications, and complications. The purpose of this study was to evaluate the application and yield of FFB and BAL in a unique group of PICU patients.
introduction, anatomical/structural/dynamic findings, BAL fluid (BALF) results, any change in management (from conservative to surgical or change in antimicrobial treatment), and any change in clinical status or chest radiograph findings following the procedure. An informative FFB was defined as an investigation in which the findings explained the clinical situation. Arterial blood gases were obtained before and within 1 h after the procedure. The lowest systolic and diastolic BP values recorded during and up to 30 min after the procedure were compared to pre-FFB values. Hypotension was defined as either a decrease in systolic BP ⬎ 10 mm Hg or a systolic BP that was below the lower limit value for age.8 Data are presented as either percentages or mean ⫾ SD as appropriate. 2 test and two-tailed, paired t tests were used for comparisons. Data were processed by Excel 2002 for Windows (Microsoft; Redmond, WA) and analyzed by BMDP Statistical Software (Los Angeles, CA).
Materials and Methods
Results
The medical records of the last 100 consecutive infants and children admitted to the PICU between October 1999 and October 2001 who underwent FFB with or without BAL were reviewed. The institution serves as a referral center for pediatric airway surgery. Procedure, Instrumentation, and Monitoring The study was approved by the institutional review board of our hospital. Sedation and analgesia protocols for FFB included one of the following: (1) propofol, 2.5 mg/kg loading dose, and subsequent boluses of 0.5 to 1 mg/kg; (2) ketamine, 1 to 2 mg/kg loading dose, following atropine, 0.02 mg/kg, and midazolam, 0.1 to 0.2 mg/kg; or (3) supplementation of morphine and midazolam boluses in patients receiving mechanical ventilation who were already receiving continuous infusions of these drugs. The endoscope was introduced into the airway through one of the following routes: the nasopharynx, a tracheostomy, an endotracheal tube (ETT), or a laryngeal mask airway. In the latter three routes, a customized three-way adapter was used for simultaneous positive pressure ventilation. Patient oxygenation, anatomic evaluation of the airways, therapeutic bronchial lavage, and BAL were performed according to previously published guidelines.7 A Pentax FB10X (equivalent to the FB-10V) bronchoscope (Pentax; Tokyo, Japan) with a distal outside diameter of 3.5 mm was employed in all procedures. The F1–7P model (Pentax) with an outer diameter of 2.2 mm was used in intubated patients with ETT sizes ⱕ 4.5, for selective intubation of a small or narrowed left main bronchus, and for severely narrowed airways. The FB10X bronchoscopes have a 1.2-mm working channel. Two senior pediatric intensivists/pulmonologists performed all procedures. The staff involved in each procedure included one senior pediatric intensivists/pulmonologist, with the assistance of the rotating pediatric resident and a registered PICU nurse. Oxygen saturation, respirations, and ECG were continuously monitored in all procedures. BP was measured and recorded either continuously by an indwelling arterial line catheter, or noninvasively every 1 min, from 5 min before to 30 min after the procedure. A chest radiograph was obtained routinely after all procedures.
Patient Characteristics One hundred fifty-five procedures were performed in 100 patients (71 boys). The age distribution is presented in Figure 1. The youngest patient was a 2-day old boy, who was admitted to the PICU because of fixed stridor and respiratory distress caused by severe congenital subglottic stenosis. Forty-nine of 155 procedures (31.6%) were performed in hemodynamically unstable patients receiving vasopressors, and 46 of 155 procedures (29.6%) were performed in patients receiving mechanical ventilation. Patients receiving mechanical ventilation who were studied for lung pathologies (n ⫽ 30) had significantly lower pre-FFB Pao2/fraction of inspired oxygen (Fio2) ratios compared with patients receiving mechanical ventilation after airway surgery who were studied for anatomic investigation of the tracheobronchial tree (n ⫽ 10) [271 ⫾ 73 vs 337 ⫾ 49 (mean ⫾ SD), p ⫽ 0.01]. Patients were classified into three groups by their causes for admission to the PICU: group A, 47 of 100 patients (47%; 70 of 155 procedures) with nonsurgical conditions (ie, community-acquired pneumonia,
Data Collection and Analysis Information retrieved included the following: demographic data, indications for FFB, sedation and analgesia, route of scope 1354
Figure 1. Age distribution of the study cohort. Pts ⫽ patients; y ⫽ year; m ⫽ month. Bronchoscopy
near drowning, meningitis, metabolic diseases etc); group B, 22 of 100 patients (22%; 30 of 155 procedures) with nonairway/lung surgical procedures (ie, neurosurgical, orthopedic and abdominal); and group C, 28 of 100 patients (28%; 55 of 155 procedures) airway surgery cases (ie, laryngoplasty, tracheal reconstruction, tracheal and bronchial stenting, tracheostomy, cricoid split, and tracheal/bronchial anastomosis). The endoscopies that were associated with airway surgical interventions were performed during the perioperative period (2 weeks before to 30 days after the operation). The indications for FFB and their frequencies in the three groups are presented in Table 1. One hundred fourteen procedures were performed in search for an anatomic airway lesion. The search for pathology in the upper airways was more common than in the lower airways (86 of 114 procedures vs 25 of 114 procedures, respectively). Of these 114 procedures, extubation failure was the major indication (44 of 114 procedures; 38.6%). This was true for the entire study group as well (57 of 155 procedures; 36.7%).
Diagnostic Yield of Airway Anatomy Investigations
Sedation and Endoscope Introduction
BAL Findings
Adequate sedation and analgesia were accomplished by administration of propofol (85 of 155 cases; 54.8%), ketamine, atropine, and midazolam (19 of 155 cases; 12.2%), and supplementation of morphine and midazolam to patients receiving mechanical ventilation (50 of 155 cases; 32.2%). The flexible bronchoscope was introduced transnasally in 84 procedures (54.2%), via an ETT in 49 procedures (31.6%), through a tracheostomy in 8 procedures (5.1%), and using laryngeal mask airway in 7 procedures (4.5%). In five cases, extubation and transnasal endoscopy followed immediately after a bronchoscopy performed via the ETT.
Five BALs were performed in patients not receiving mechanical ventilation. Three BALs were performed for noninfectious causes; two thirds were highly positive for lipid-laden alveolar macrophages, supporting the diagnosis of recurrent aspiration pneumonia. Two procedures were done in patients with community-acquired pneumonia refractory to empiric treatment, yielding Staphylococcus aureus and a penicillin-resistant Streptococcus pneumoniae. The remaining 25 BAL investigations were performed in patients receiving mechanical ventilation. The yield is summarized in Table 4. New persistent pulmonary infiltrates accompanied by fever and/or
The yield of FFB for the various indications is presented in Table 2. FFB yielded positive anatomic information in 35 of 55 airway surgical cases (63.6%) [group C], compared to 44 of 59 nonairway surgical cases (74.5%) [groups A and B combined, p ⫽ 0.8]. Of the 31 postoperative FFBs performed in group C patients, airway reoperation took place in 11 FFBs (35.5%). Overall, anatomic airway pathologies were found in 65 of 89 cases (73%) in which an upper airway lesion was suspected, and in 14 of 25 cases (56%) in which lower airway pathology was pursued. A list of structural airway pathologies disclosed by FFB is presented in Table 3. Of 59 investigations that were performed for anatomic evaluation in nonairway surgery patients, 19 investigations (32.2%) resulted in a change in patient management or directed treatment toward surgical interventions. For example, in an infant admitted for severe stridor, FFB disclosed bilateral vocal cord paralysis, mandating CNS imaging that showed Arnold-Chiari malformation that led to posterior fossa decompression.
Table 1—Indications for FFB*
Patient Groups Suspected anatomic lesions Suspected upper airway lesions Suspected lower airway lesions Therapeutic bronchial lavage BAL Total†
Nonsurgical (70 Procedures)
Surgical Non-Airway (30 Procedures)
43 (61.4) [21]
Airway Surgery (n ⫽ 55 Procedures) Preoperative
Postoperative
Total, No. (Extubation Failures)
16 (53.2) [6]
24 (43.6)
31 (56.3) [17]
114 [44]
32 (45.7) [15]
8 (26.6) [4]
21 (38.2)
28 (50.1) [14]
89 [33]
11 (15.7) [6]
8 (26.6) [2]
3 (5.5)
3 (5.5) [3]
25 [11]
21 (30) [5]
8 (26.6) [3]
0
6 (11) [5]
35 [13]
19 (27.1) 83 [26]
5 (16.6) 29 [9]
0 24 [0]
6 (11) 43 [22]
30 179 [57]
*Data are presented as No. of indication (% of procedures) [extubation failures] or No. of indications (% of procedures) unless otherwise indicated. †Total of indications; 54 of 155 procedures (34.8%) were performed for more than one indication. www.chestjournal.org
CHEST / 126 / 4 / OCTOBER, 2004
1355
Table 2—Positive Findings on Observation for Anatomic Investigation as a Percentage of Each Indication* Patient Groups
Nonsurgical
Surgical Non-Airway
Airway Surgery
All Cases
Anatomic diagnosis Upper airways Lower airways Extubation failure† Possible foreign body, No.
32/43 (74.4) 25/32 (78.1) 7/11 (63.6) 14/21 (66.6) 0 of 3
12/16 (75) 7/8 (87.5) 5/8 (62.5) 3/6 (50) 0
35/55 (63.6) 33/49 (67.3) 2/6 (33.3) 11/17 (64.7) 0
79/114 (69.3) 65/89 (73) 14/25 (56) 28/44 (63.6) 0 of 3
*Data are presented as No. of positive findings/total FFBs (%) unless otherwise indicated. A positive finding implied either a distinct anatomic lesion, “clearance” towards a trial of extubation or an unsatisfactory surgical result that could explain respiratory noise. †Data are presented as No./total cases (%) as a manifestation of either upper or lower airway pathology.
leukocytosis after ⬎ 5 days of intubation (ventilatorassociated pneumonia) was the indication for BAL in 21 of 25 patients. The other four cases were investigated due to treatment failure of communityacquired pneumonia. Overall, offending pathogens were identified by BAL in 14 of 25 procedures (56%), while 11 BAL findings were negative. Bacteria were isolated in 9 of 14 procedures, viral antigens were detected in 3 of 14 procedures, and Pneumocystis carinii and Candida albicans in one procedure each. Blind tracheal aspirates prior to the BAL were positive in 22 of 25 cases (88%) cases, while BAL findings were negative in 11 of these 22 cases (50%) [p ⫽ 0.2]. Concordance of the infectious agents isolated by both techniques was found in only 4 of 11 cases (36%). Therapeutic Procedures In cases of segmental or lobar lung collapse, a successful therapeutic bronchial lavage was accom-
Table 3—Airway Pathologies Found by FFB* Diagnosis Normal airway Vocal cord paralysis, bilateral Subglottic stenosis Laryngomalacia† Postintubation subglottic edema Tracheomalacia Bronchial stenosis Bronchomalacia Tracheostomy granuloma Tracheal stenosis Vascular ring Tracheal fistula Vocal cord paralysis, unilateral Glottic stenosis Foreign body
Data 35 (30.7) 16 (14) 13 (11.4) 12 (10.5) 10 (8.7) 7 (6.1) 5 (4.4) 4 (3.5) 4 (3.5) 3 2 2 1 1 0
*Data are presented as No. of cases (% of 114 diagnostic FFBs) or No. †Severe cases, defined as infants with respiratory distress requiring intensive monitoring or additional oxygen therapy. 1356
plished in 26 of 35 cases (74.3%) [re-expansion of collapsed lung areas by chest radiography]. Weaning and extubation failure (one or more trial) was the major manifestation and indication for the endoscopy in 13 of 35 cases (37.1%). This was effectively treated by FFB in 11 of 13 cases, ie, successful weaning and extubation. Foreign bodies were not identified in any of the patients. Complications There were no procedure-related deaths. Significant nasal/airway trauma, stridor, hemoptysis, pneumothorax, septicemia, or cardiac arrhythmias were not encountered after any of the procedures. FFB with or without BAL did not result in significant reductions in Pao2/Fio2 ratios in children receiving mechanical ventilation, nor in deleterious effects on blood pH or Paco2 (Table 5). In procedures performed under ventilatory support, no significant changes in ventilation pressures were observed fol-
Table 4 —The Yield of BAL and Blind Tracheal Aspirates in 25 Patients Receiving Mechanical Ventilation
Variables Positive tracheal aspirates prior to BAL Negative tracheal aspirates prior to BAL Change in antimicrobial treatment† Clinical improvement‡
Positive BALF Findings (n ⫽ 14), No.
Negative BALF Findings (n ⫽ 11), No.
Total (n ⫽ 25), No. (%)
11*
11
22 (88)
3
0
3 (12)
9
6
15 (60)
8
4
12 (48)
*In only four cases, the organisms cultured from tracheal aspirates coincided with the positive BAL isolates. †Including downgrading or cessation of antimicrobial treatment. ‡Resolution of fever accompanied by resolution of infiltrates, following any change (including cessation or downgrading) in antimicrobial treatment. Bronchoscopy
Table 5—Respiratory Variables in Patients Receiving Mechanical Ventilation Before and After FFBa* Pao2/Fio2 Ratio Variables
Total† (n ⫽ 40)
BAL‡ (n ⫽ 14)
Blood pH (n ⫽ 39)
Paco2, torr (n ⫽ 39)
PEEP, cm H2O
PIP, cm H2O
Before FFB§ After FFB㛳 p Value
286.3 ⫾ 73.4 302.7 ⫾ 74 0.12
287.9 ⫾ 88 282.3 ⫾ 77.8 0.37
7.36 ⫾ 0.02 7.37 ⫾ 0.03 0.17
44.38 ⫾ 4.6 43.56 ⫾ 5.9 0.33
6.0 ⫾ 1.3 5.9 ⫾ 1.2 0.18
27.1 ⫾ 4.3 27 ⫾ 4.5 0.53
*Data are presented as mean ⫾ SD. PEEP ⫽ positive end-expiratory pressure; PIP ⫽ peak inspiratory pressure. †Regardless of the indication for FFB; the Pao2/Fio2 ratio was available only in 40 of 46 patients receiving mechanical ventilation. ‡Only in patients receiving mechanical ventilation who underwent BAL with no atelectasis on chest radiography. §Recorded 60 min before the procedure. 㛳Recorded 60 min after the procedure.
lowing FFB. In 2 cases, intubation was performed immediately after FFB due to respiratory depression and inability to clear pulmonary secretions. Both patients were extubated successfully within the following 12 h. Average systolic and diastolic BP declined significantly during FFB (99.8 ⫾ 14.7 mm Hg vs 97.7 ⫾ 14.4 mm Hg [p ⫽ 0.001] and 58.6 ⫾ 9.1 mm Hg vs 56.3 ⫾ 8.8 mm Hg [p ⬍ 0.001], respectively). Systolic hypotension was observed during 30 of 155 procedures (19.3%). Twenty of these cases (66%) were stabilized by one to two boluses of 10 mL/kg NaCl 0.9% during the procedure or within 30 min afterwards. In the rest, BP returned to preprocedure values without any intervention. None required initiation or augmentation of vasopressors in order to stabilize BP. In all cases, the BP returned to preprocedure levels within ⬍ 1 h after discontinuation of IV sedation. Discussion The present report analyzes the application and value of FFB in a group of very sick infants and children admitted to the PICU. Previous studies4,5 were not confined to this unique group, and included mainly outpatients and inpatients hospitalized in the general wards who were admitted to the PICU only for the procedure as a part of the evaluation of their respiratory problem. Except for one study6 in 1988 that evaluated FFB in PICU patients, this report is the first to address these issues. Obviously, the characteristics of the specific medical center from where the information is retrieved will affect the distribution of indications for FFB. Our hospital is a referral center for airway surgery. This, apparently, influences the present results, for example, by including FFB for post-airway surgery evaluation. Some of the indications for FFB in the PICU are not different from those in non-PICU patients.4,5 Nevertheless, the medical status is significantly www.chestjournal.org
worse, manifested by the relatively low Pao2/Fio2 values. A higher risk for complications is, therefore, anticipated. The evaluation of airway abnormalities constitutes the major indication for FFB in the present study (74% of all procedures). This is in accordance with previous reports.4,5 An earlier report6 of 104 FFBs in PICU patients found a higher rate: 92%. However, BAL was not included and pulse oximetry was not universally used. All three studies4 – 6 showed a substantial contribution of FFB performed for anatomic evaluation of the upper and lower airways (76 to 95%). In the present study, the rate of positive findings was lower: 69%. Godfrey et al4 and Nussbaum5 studied heterogeneous groups of patients and included a substantial number of noncritical and ambulatory outpatients.4,5 Their reports included patients with noisy breathing, congenital stridor, and adenoid hypertrophy in whom the expected rate of positive findings is almost 100%. Our data are supported by a publication in Spanish that reported a 65% rate of airway abnormalities in 51 FFBs performed in PICU patients.9 The definition of positive contribution may be different in PICU patients. In the setting of the pulmonary services, patient/parent reassurance (negative findings) was included under the definition of a “clinically meaningful contribution.”5 This will hardly be the case in the PICU environment. The diagnosis of “normal airways” carries an important clinical and practical contribution to patient management and decision making regarding extubation, particularly in the postoperative airway surgery cases. For example, the observation of patent, noncollapsing airways following specific tracheobronchial procedures such as reconstruction surgeries and stenting may direct toward extubation. Hence, the yield of FFB should include any contributive information that cannot be obtained by other means. The contribution of FFB is evident also by the fact that in almost one third of the CHEST / 126 / 4 / OCTOBER, 2004
1357
cases, patient management changed from conservative to surgical as a result of the FFB findings. Whereas cytologic examination of alveolar fluid is of great interest in ambulatory patients, BALs in a PICU are employed mainly for the diagnosis of lung infections.10,11 Most of BAL procedures (90%) were performed to isolate organisms causing pulmonary infections. The advantage of BALF microbiologic studies over the regular tracheal aspirates was clearly demonstrated by Fagon et al,12 in terms of both decreased usage of antibiotics and lower mortality rates. BAL lowers the number of cases treated incorrectly due to tracheal colonization, thus allowing a more judicious use of antimicrobials.13 Although the present study was not designated to compare BAL to tracheal aspirates, the findings correspond with these observations. In the present series, negative and positive microbiologic BALF results changed treatment in 6 of 11 cases and 9 of 14 cases, followed by a marked clinical improvement in 4 of 6 cases and 8 of 9 cases, respectively. Negative microbiologic BALF findings are important as they may accelerate the search for other foci of infection and lead to cessation or to downgrading of antibiotic treatment. Therapeutic procedures accounted for 25% of 200 ambulatory and PICU bronchoscopies reported by Godfrey et al,4 and for 27% of 51 FFBs performed in PICU patients in the series of Garcia et al.9 This is similar to the 24% rate of interventional procedures performed in the present cohort, most of which were due to a collapsed lung segment refractory to chest physiotherapy and tracheal suction. A 74% success rate of treating lung atelectasis by FFB and lavage was found in the present series. This rate is similar to the findings in non-PICU infants and children.4,14 The 80% rate of successful extubation following FFB and BAL in patients who had failed earlier because of atelectasis highlights the importance of targeted, under-vision removal of airway plugs in critically ill infants and children. With the recent introduction of the 2.8-mm bronchoscope that has a suction channel, this has become a therapeutic option also in infants receiving ventilation via a small ETT. We recommend that the procedure should be considered early, especially in patients receiving ventilation. FFB may have particular contribution in children with dysfunctional cough reflexes, due to postoperative pain, neuromuscular diseases7 and CNS depression. The role of FFB in the perioperative management of airway surgery patients has not been established yet. In one third of the postoperative investigations, most of which were performed just prior to a planned extubation, FFB provided information that changed the course toward airway reintervention. Such cases in1358
cluded, for example, obstructing granulation tissue in tracheal anastomoses and airway stent dislodgment. Generally, FFB is regarded as a very safe procedure. In a report5 of 2,836 procedures performed in both PICU and ambulatory patients, no related mortality was recorded. Our results show that it can be safely performed also in critically ill infants and children. Severe complications of FFB are rare; nevertheless, side effects are common. Most often, these are short lived. Examples are transient hypoxemia or hypercarbia due to partial airway blockage or to inadvertent clearance of surfactant during BAL. Many of the FFBs were performed in hypoxemic children with decreased Pao2/Fio2 ratios. These patients belonged mainly to the nonairway surgery groups (ie, groups A and B). Nevertheless, significant hypoxemia did not complicate any of the procedures in this series and no increase of ventilatory variables was required. This may be attributed to the protective assisted positive pressure ventilatory support and the application of Fio2 of 1.0 during the procedure that may have prevented deterioration in respiratory status in critical patients. Also, the assisted ventilation has possibly prevented hypoxemia and hypercarbia due to hypoventilation that may occur secondary to the effects of sedative agents.15 Overall, there was a statistically significant decrease in BP during the procedure. This may be attributed to the vasodilator and myocardial depression properties of the anesthetic agents employed.16,17 Unintentional rises in positive end-expiratory pressure during the procedure may aggravate BP drops.15 However, the incidence of true hypotension was relatively low, and the drop in BP was mild and either short lived or responded promptly to fluid management and did not require additional vasopressor treatment. Hence, the clinical significance of the decrease in BP is minor. This series shows that FFB is used in a myriad of PICU scenarios and contributes substantially to the management of the critically ill child. Therefore, we suggest that due to its frequent use in sick infants and children and due to the possible need for urgent endoscopy also at times when pulmonary services may not be immediately available, FFB should be a PICU expertise employed by intensivists without relying exclusively on pulmonary service support. This recommendation may not be universal and may be applicable especially in centers where FFB is performed frequently in the PICU. It is advisable that in order to be qualified to perform FFBs independently, “hands-on” courses should be attended. Such courses presently take place in the United States and Europe. It has been suggested that each trainee assist in 50 procedures and perform 25 supervised procedures.7 Bronchoscopy
Conclusion Pediatric FFB is a safe diagnostic and interventional tool in the management of very sick infants and children hospitalized in a PICU. FFB in this unique population may be indicated in situations that are different from those encountered in outpatient clinics and “regular” pulmonary practice. Owing to the demand for such a powerful bedside tool, PICU professionals are encouraged to acquire the appropriate skills and be accredited for performing FFB. ACKNOWLEDGMENT: In memory of Gabriel Hugo BarZohar.
References 1 Nussbaum E. Flexible fiberoptic bronchoscopy and laryngoscopy in children under 2 years of age: diagnostic and therapeutic applications of a new pediatric flexible fiberoptic bronchoscope. Crit Care Med 1982; 10:770 –772 2 Fan LL, Sparks LM, Dulinski JP. Applications of an ultrathin flexible bronchoscope for neonatal and pediatric airway problems. Chest 1986; 89:673– 676 3 Wood RE. Pitfalls in the use of the flexible bronchoscope in pediatric patients. Chest 1990; 97:199 –203 4 Godfrey S, Avital A, Maayan C, et al. Yield from flexible bronchoscopy in children. Pediatr Pulmonol 1997; 23:261– 269 5 Nussbaum E. Pediatric fiberoptic bronchoscopy: clinical experience with 2,836 bronchoscopies. Pediatr Crit Care Med 2002; 3:171–176 6 Fan LL, Sparks LM, Fix FJ. Flexible fiberoptic endoscopy for
www.chestjournal.org
7 8 9 10
11 12
13
14 15 16 17
airway problems in a pediatric intensive care unit. Chest 1988; 93:556 –560 Balfour-Lynn IM, Spencer H. Bronchoscopy: how and when? Paediatr Respir Rev 2002; 3:255–264 Guidelines 2000 for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Currents in Emergency Cardiovascular Care, Fall 2000. Vol. 11, No. 3 Garcia GE, Perez RE, Quesada RL, et al. Usefulness of fiberoptic bronchoscopy in critical pediatric care. Arch Bronconeumol 1999; 35:525–528 Leigh MW, Henshaw NG, Wood RE. Diagnosis of Pneumocystis carinii pneumonia in pediatric patients using bronchoscopic bronchoalveolar lavage. Pediatr Infect Dis J 1985; 4:408 – 410 Stokes DC, Shenep JL, Parham D, et al. Role of flexible bronchoscopy in the diagnosis of pulmonary infiltrates in pediatric patients with cancer. J Pediatr 1989; 115:561–567 Fagon JY, Chastre J, Wolff M, et al. Invasive and non-invasive strategies for the management of suspected ventilator-associated pneumonia: a randomized trial. Ann Intern Med 2000; 132:621– 630 Ben-Ari J, Yaniv I, Nahum E, et al. Yield of bronchoalveolar lavage in ventilated and non-ventilated children after bone marrow transplantation. Bone Marrow Transplant 2001; 27: 191–194 Holmgren NL, Cordova M, Ortuzar P, et al. Role of flexible bronchoscopy in the re-expansion of persistent atelectasis in children. Arch Bronconeumol 2002; 38:367–371 Masters IB, Cooper P. Position paper: paediatric flexible bronchoscopy. J Paediatr Child Health 2002; 38:555–559 Tobias JD. Sedation and anesthesia for pediatric bronchoscopy. In: Prakash UBS, ed. Bronchoscope. New York, NY: Raven Press, 1994; 345–356 Pue CA, Pacht ER. Complications of fiberoptic bronchoscopy at a university hospital. Chest 1995; 107:430 – 432
CHEST / 126 / 4 / OCTOBER, 2004
1359
Study objective: To evaluate the contribution of flexible fiberoptic bronchoscopy (FFB) and BAL to the clinical management of patients in a pediatric ICU (PICU). Setting and design: A retrospective study based on medical records in a six-bed pediatric ICU of a tertiary care children’s hospital serving as a referral center for airway surgery. Patients and participants: One hundred consecutive infants and children hospitalized in a PICU, who underwent FFB with or without BAL. Measurements and results: One hundred fifty-five procedures were performed, for the following causes: search for airways anatomic pathologies (114 of 155 procedures, 74%), including 55 procedures during the perioperative period of airway surgery; treatment of atelectasis (35 of 155 procedures, 22.5%); and BAL (30 of 155 procedures, 19%). Thirty-five percent of procedures had more than one cause. Airway pathology was observed in 79 of 114 procedures (69%). Management changed from conservative to surgical in 44 of 114 procedures (39%). In airway surgery cases, reoperation subsequent to postoperative FFB took place in 35%. BAL results changed antimicrobial treatment in 15 of 30 cases, with clinical improvement in 10 of 30 cases (33%). Treatment of atelectasis was successful in 26 of 35 cases (74.3%). No procedure-related mortality, life-threatening complications, or significant changes in patient status occurred. Conclusions: FFB is an important and safe procedure in very sick infants and children with a variety of respiratory diseases, and significantly contributes to their management. FFB should be considered to be a PICU staff expertise. (CHEST 2004; 126:1353–1359) Key words: bronchoscopy; children; infants; pediatric ICU Abbreviations: BALF ⫽ BAL fluid; ETT ⫽ endotracheal tube; FFB ⫽ flexible fiberoptic bronchoscopy; Fio2 ⫽ fraction of inspired oxygen; PICU ⫽ pediatric ICU
fiberoptic bronchoscopy (FFB) is freF lexible quently used in the assessment and treatment of
infants and children with a variety of pediatric respiratory diseases.1–3 FFB with or without BAL is particularly important in the diagnosis and treatment of specific respiratory problems in critically ill infants and children hospitalized in the pediatric ICU (PICU). These comprise situations such as segmental lung collapse, pulmonary infections including community-acquired and ventilator-associated pneumonia, pulmonary infections in immunocompromised hosts, and pulmonary bleeding. In addition,
*From the Pediatric Intensive Care Unit, Dana Children’s Hospital, Tel-Aviv Sourasky Medical Center, Tel-Aviv University Sackler Faculty of Medicine, Tel-Aviv, Israel. Manuscript received January 5, 2004; revision accepted May 26, 2004. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (e-mail: [email protected]). Correspondence to: Yakov Sivan, MD, Pediatric Intensive Care, Dana Children’s Hospital, Tel-Aviv Medical Center, 6, Weizman St, Tel-Aviv 64239, Israel; e-mail: [email protected] www.chestjournal.org
FFB has a special role in airway problems in very sick pediatric patients. It may be crucial in intubation of the difficult airways and in the assessment of airway anomalies. With the evolution of airway surgical techniques for treating various compromising and life-threatening congenital airway anomalies, bedside FFB has becomes an indispensable clinical tool in the immediate perioperative PICU course. At present, there is only little information on the yield of FFB and BAL in the PICU. Most previously reported studies that analyzed the contribution of pediatric FFB originated from cohorts of noncritically ill patients with a variety of airway and lung diseases who were treated under the pediatric pulmonary services outside the PICU. These included ambulatory patients admitted to the hospital only for the procedure and noncritical patients hospitalized in the general pediatric and pediatric surgery wards.4,5 The largest pediatric series5 published included 2,836 patients. Most of the procedures were performed in the PICU. Nevertheless, CHEST / 126 / 4 / OCTOBER, 2004
1353
from the list of indications it is apparent that a substantial number were not true “PICU patients” and were admitted to the PICU only for FFB. At present, the only study6 in the English-language literature that specifically studied the application of FFB in the PICU was published in 1988, before BAL became widely accepted and airway surgery cases were not included. Hence, because PICU patients differ from other patient groups in terms of systems stability and the underlying problems, they should be evaluated separately for clinical benefit, indications, contraindications, and complications. The purpose of this study was to evaluate the application and yield of FFB and BAL in a unique group of PICU patients.
introduction, anatomical/structural/dynamic findings, BAL fluid (BALF) results, any change in management (from conservative to surgical or change in antimicrobial treatment), and any change in clinical status or chest radiograph findings following the procedure. An informative FFB was defined as an investigation in which the findings explained the clinical situation. Arterial blood gases were obtained before and within 1 h after the procedure. The lowest systolic and diastolic BP values recorded during and up to 30 min after the procedure were compared to pre-FFB values. Hypotension was defined as either a decrease in systolic BP ⬎ 10 mm Hg or a systolic BP that was below the lower limit value for age.8 Data are presented as either percentages or mean ⫾ SD as appropriate. 2 test and two-tailed, paired t tests were used for comparisons. Data were processed by Excel 2002 for Windows (Microsoft; Redmond, WA) and analyzed by BMDP Statistical Software (Los Angeles, CA).
Materials and Methods
Results
The medical records of the last 100 consecutive infants and children admitted to the PICU between October 1999 and October 2001 who underwent FFB with or without BAL were reviewed. The institution serves as a referral center for pediatric airway surgery. Procedure, Instrumentation, and Monitoring The study was approved by the institutional review board of our hospital. Sedation and analgesia protocols for FFB included one of the following: (1) propofol, 2.5 mg/kg loading dose, and subsequent boluses of 0.5 to 1 mg/kg; (2) ketamine, 1 to 2 mg/kg loading dose, following atropine, 0.02 mg/kg, and midazolam, 0.1 to 0.2 mg/kg; or (3) supplementation of morphine and midazolam boluses in patients receiving mechanical ventilation who were already receiving continuous infusions of these drugs. The endoscope was introduced into the airway through one of the following routes: the nasopharynx, a tracheostomy, an endotracheal tube (ETT), or a laryngeal mask airway. In the latter three routes, a customized three-way adapter was used for simultaneous positive pressure ventilation. Patient oxygenation, anatomic evaluation of the airways, therapeutic bronchial lavage, and BAL were performed according to previously published guidelines.7 A Pentax FB10X (equivalent to the FB-10V) bronchoscope (Pentax; Tokyo, Japan) with a distal outside diameter of 3.5 mm was employed in all procedures. The F1–7P model (Pentax) with an outer diameter of 2.2 mm was used in intubated patients with ETT sizes ⱕ 4.5, for selective intubation of a small or narrowed left main bronchus, and for severely narrowed airways. The FB10X bronchoscopes have a 1.2-mm working channel. Two senior pediatric intensivists/pulmonologists performed all procedures. The staff involved in each procedure included one senior pediatric intensivists/pulmonologist, with the assistance of the rotating pediatric resident and a registered PICU nurse. Oxygen saturation, respirations, and ECG were continuously monitored in all procedures. BP was measured and recorded either continuously by an indwelling arterial line catheter, or noninvasively every 1 min, from 5 min before to 30 min after the procedure. A chest radiograph was obtained routinely after all procedures.
Patient Characteristics One hundred fifty-five procedures were performed in 100 patients (71 boys). The age distribution is presented in Figure 1. The youngest patient was a 2-day old boy, who was admitted to the PICU because of fixed stridor and respiratory distress caused by severe congenital subglottic stenosis. Forty-nine of 155 procedures (31.6%) were performed in hemodynamically unstable patients receiving vasopressors, and 46 of 155 procedures (29.6%) were performed in patients receiving mechanical ventilation. Patients receiving mechanical ventilation who were studied for lung pathologies (n ⫽ 30) had significantly lower pre-FFB Pao2/fraction of inspired oxygen (Fio2) ratios compared with patients receiving mechanical ventilation after airway surgery who were studied for anatomic investigation of the tracheobronchial tree (n ⫽ 10) [271 ⫾ 73 vs 337 ⫾ 49 (mean ⫾ SD), p ⫽ 0.01]. Patients were classified into three groups by their causes for admission to the PICU: group A, 47 of 100 patients (47%; 70 of 155 procedures) with nonsurgical conditions (ie, community-acquired pneumonia,
Data Collection and Analysis Information retrieved included the following: demographic data, indications for FFB, sedation and analgesia, route of scope 1354
Figure 1. Age distribution of the study cohort. Pts ⫽ patients; y ⫽ year; m ⫽ month. Bronchoscopy
near drowning, meningitis, metabolic diseases etc); group B, 22 of 100 patients (22%; 30 of 155 procedures) with nonairway/lung surgical procedures (ie, neurosurgical, orthopedic and abdominal); and group C, 28 of 100 patients (28%; 55 of 155 procedures) airway surgery cases (ie, laryngoplasty, tracheal reconstruction, tracheal and bronchial stenting, tracheostomy, cricoid split, and tracheal/bronchial anastomosis). The endoscopies that were associated with airway surgical interventions were performed during the perioperative period (2 weeks before to 30 days after the operation). The indications for FFB and their frequencies in the three groups are presented in Table 1. One hundred fourteen procedures were performed in search for an anatomic airway lesion. The search for pathology in the upper airways was more common than in the lower airways (86 of 114 procedures vs 25 of 114 procedures, respectively). Of these 114 procedures, extubation failure was the major indication (44 of 114 procedures; 38.6%). This was true for the entire study group as well (57 of 155 procedures; 36.7%).
Diagnostic Yield of Airway Anatomy Investigations
Sedation and Endoscope Introduction
BAL Findings
Adequate sedation and analgesia were accomplished by administration of propofol (85 of 155 cases; 54.8%), ketamine, atropine, and midazolam (19 of 155 cases; 12.2%), and supplementation of morphine and midazolam to patients receiving mechanical ventilation (50 of 155 cases; 32.2%). The flexible bronchoscope was introduced transnasally in 84 procedures (54.2%), via an ETT in 49 procedures (31.6%), through a tracheostomy in 8 procedures (5.1%), and using laryngeal mask airway in 7 procedures (4.5%). In five cases, extubation and transnasal endoscopy followed immediately after a bronchoscopy performed via the ETT.
Five BALs were performed in patients not receiving mechanical ventilation. Three BALs were performed for noninfectious causes; two thirds were highly positive for lipid-laden alveolar macrophages, supporting the diagnosis of recurrent aspiration pneumonia. Two procedures were done in patients with community-acquired pneumonia refractory to empiric treatment, yielding Staphylococcus aureus and a penicillin-resistant Streptococcus pneumoniae. The remaining 25 BAL investigations were performed in patients receiving mechanical ventilation. The yield is summarized in Table 4. New persistent pulmonary infiltrates accompanied by fever and/or
The yield of FFB for the various indications is presented in Table 2. FFB yielded positive anatomic information in 35 of 55 airway surgical cases (63.6%) [group C], compared to 44 of 59 nonairway surgical cases (74.5%) [groups A and B combined, p ⫽ 0.8]. Of the 31 postoperative FFBs performed in group C patients, airway reoperation took place in 11 FFBs (35.5%). Overall, anatomic airway pathologies were found in 65 of 89 cases (73%) in which an upper airway lesion was suspected, and in 14 of 25 cases (56%) in which lower airway pathology was pursued. A list of structural airway pathologies disclosed by FFB is presented in Table 3. Of 59 investigations that were performed for anatomic evaluation in nonairway surgery patients, 19 investigations (32.2%) resulted in a change in patient management or directed treatment toward surgical interventions. For example, in an infant admitted for severe stridor, FFB disclosed bilateral vocal cord paralysis, mandating CNS imaging that showed Arnold-Chiari malformation that led to posterior fossa decompression.
Table 1—Indications for FFB*
Patient Groups Suspected anatomic lesions Suspected upper airway lesions Suspected lower airway lesions Therapeutic bronchial lavage BAL Total†
Nonsurgical (70 Procedures)
Surgical Non-Airway (30 Procedures)
43 (61.4) [21]
Airway Surgery (n ⫽ 55 Procedures) Preoperative
Postoperative
Total, No. (Extubation Failures)
16 (53.2) [6]
24 (43.6)
31 (56.3) [17]
114 [44]
32 (45.7) [15]
8 (26.6) [4]
21 (38.2)
28 (50.1) [14]
89 [33]
11 (15.7) [6]
8 (26.6) [2]
3 (5.5)
3 (5.5) [3]
25 [11]
21 (30) [5]
8 (26.6) [3]
0
6 (11) [5]
35 [13]
19 (27.1) 83 [26]
5 (16.6) 29 [9]
0 24 [0]
6 (11) 43 [22]
30 179 [57]
*Data are presented as No. of indication (% of procedures) [extubation failures] or No. of indications (% of procedures) unless otherwise indicated. †Total of indications; 54 of 155 procedures (34.8%) were performed for more than one indication. www.chestjournal.org
CHEST / 126 / 4 / OCTOBER, 2004
1355
Table 2—Positive Findings on Observation for Anatomic Investigation as a Percentage of Each Indication* Patient Groups
Nonsurgical
Surgical Non-Airway
Airway Surgery
All Cases
Anatomic diagnosis Upper airways Lower airways Extubation failure† Possible foreign body, No.
32/43 (74.4) 25/32 (78.1) 7/11 (63.6) 14/21 (66.6) 0 of 3
12/16 (75) 7/8 (87.5) 5/8 (62.5) 3/6 (50) 0
35/55 (63.6) 33/49 (67.3) 2/6 (33.3) 11/17 (64.7) 0
79/114 (69.3) 65/89 (73) 14/25 (56) 28/44 (63.6) 0 of 3
*Data are presented as No. of positive findings/total FFBs (%) unless otherwise indicated. A positive finding implied either a distinct anatomic lesion, “clearance” towards a trial of extubation or an unsatisfactory surgical result that could explain respiratory noise. †Data are presented as No./total cases (%) as a manifestation of either upper or lower airway pathology.
leukocytosis after ⬎ 5 days of intubation (ventilatorassociated pneumonia) was the indication for BAL in 21 of 25 patients. The other four cases were investigated due to treatment failure of communityacquired pneumonia. Overall, offending pathogens were identified by BAL in 14 of 25 procedures (56%), while 11 BAL findings were negative. Bacteria were isolated in 9 of 14 procedures, viral antigens were detected in 3 of 14 procedures, and Pneumocystis carinii and Candida albicans in one procedure each. Blind tracheal aspirates prior to the BAL were positive in 22 of 25 cases (88%) cases, while BAL findings were negative in 11 of these 22 cases (50%) [p ⫽ 0.2]. Concordance of the infectious agents isolated by both techniques was found in only 4 of 11 cases (36%). Therapeutic Procedures In cases of segmental or lobar lung collapse, a successful therapeutic bronchial lavage was accom-
Table 3—Airway Pathologies Found by FFB* Diagnosis Normal airway Vocal cord paralysis, bilateral Subglottic stenosis Laryngomalacia† Postintubation subglottic edema Tracheomalacia Bronchial stenosis Bronchomalacia Tracheostomy granuloma Tracheal stenosis Vascular ring Tracheal fistula Vocal cord paralysis, unilateral Glottic stenosis Foreign body
Data 35 (30.7) 16 (14) 13 (11.4) 12 (10.5) 10 (8.7) 7 (6.1) 5 (4.4) 4 (3.5) 4 (3.5) 3 2 2 1 1 0
*Data are presented as No. of cases (% of 114 diagnostic FFBs) or No. †Severe cases, defined as infants with respiratory distress requiring intensive monitoring or additional oxygen therapy. 1356
plished in 26 of 35 cases (74.3%) [re-expansion of collapsed lung areas by chest radiography]. Weaning and extubation failure (one or more trial) was the major manifestation and indication for the endoscopy in 13 of 35 cases (37.1%). This was effectively treated by FFB in 11 of 13 cases, ie, successful weaning and extubation. Foreign bodies were not identified in any of the patients. Complications There were no procedure-related deaths. Significant nasal/airway trauma, stridor, hemoptysis, pneumothorax, septicemia, or cardiac arrhythmias were not encountered after any of the procedures. FFB with or without BAL did not result in significant reductions in Pao2/Fio2 ratios in children receiving mechanical ventilation, nor in deleterious effects on blood pH or Paco2 (Table 5). In procedures performed under ventilatory support, no significant changes in ventilation pressures were observed fol-
Table 4 —The Yield of BAL and Blind Tracheal Aspirates in 25 Patients Receiving Mechanical Ventilation
Variables Positive tracheal aspirates prior to BAL Negative tracheal aspirates prior to BAL Change in antimicrobial treatment† Clinical improvement‡
Positive BALF Findings (n ⫽ 14), No.
Negative BALF Findings (n ⫽ 11), No.
Total (n ⫽ 25), No. (%)
11*
11
22 (88)
3
0
3 (12)
9
6
15 (60)
8
4
12 (48)
*In only four cases, the organisms cultured from tracheal aspirates coincided with the positive BAL isolates. †Including downgrading or cessation of antimicrobial treatment. ‡Resolution of fever accompanied by resolution of infiltrates, following any change (including cessation or downgrading) in antimicrobial treatment. Bronchoscopy
Table 5—Respiratory Variables in Patients Receiving Mechanical Ventilation Before and After FFBa* Pao2/Fio2 Ratio Variables
Total† (n ⫽ 40)
BAL‡ (n ⫽ 14)
Blood pH (n ⫽ 39)
Paco2, torr (n ⫽ 39)
PEEP, cm H2O
PIP, cm H2O
Before FFB§ After FFB㛳 p Value
286.3 ⫾ 73.4 302.7 ⫾ 74 0.12
287.9 ⫾ 88 282.3 ⫾ 77.8 0.37
7.36 ⫾ 0.02 7.37 ⫾ 0.03 0.17
44.38 ⫾ 4.6 43.56 ⫾ 5.9 0.33
6.0 ⫾ 1.3 5.9 ⫾ 1.2 0.18
27.1 ⫾ 4.3 27 ⫾ 4.5 0.53
*Data are presented as mean ⫾ SD. PEEP ⫽ positive end-expiratory pressure; PIP ⫽ peak inspiratory pressure. †Regardless of the indication for FFB; the Pao2/Fio2 ratio was available only in 40 of 46 patients receiving mechanical ventilation. ‡Only in patients receiving mechanical ventilation who underwent BAL with no atelectasis on chest radiography. §Recorded 60 min before the procedure. 㛳Recorded 60 min after the procedure.
lowing FFB. In 2 cases, intubation was performed immediately after FFB due to respiratory depression and inability to clear pulmonary secretions. Both patients were extubated successfully within the following 12 h. Average systolic and diastolic BP declined significantly during FFB (99.8 ⫾ 14.7 mm Hg vs 97.7 ⫾ 14.4 mm Hg [p ⫽ 0.001] and 58.6 ⫾ 9.1 mm Hg vs 56.3 ⫾ 8.8 mm Hg [p ⬍ 0.001], respectively). Systolic hypotension was observed during 30 of 155 procedures (19.3%). Twenty of these cases (66%) were stabilized by one to two boluses of 10 mL/kg NaCl 0.9% during the procedure or within 30 min afterwards. In the rest, BP returned to preprocedure values without any intervention. None required initiation or augmentation of vasopressors in order to stabilize BP. In all cases, the BP returned to preprocedure levels within ⬍ 1 h after discontinuation of IV sedation. Discussion The present report analyzes the application and value of FFB in a group of very sick infants and children admitted to the PICU. Previous studies4,5 were not confined to this unique group, and included mainly outpatients and inpatients hospitalized in the general wards who were admitted to the PICU only for the procedure as a part of the evaluation of their respiratory problem. Except for one study6 in 1988 that evaluated FFB in PICU patients, this report is the first to address these issues. Obviously, the characteristics of the specific medical center from where the information is retrieved will affect the distribution of indications for FFB. Our hospital is a referral center for airway surgery. This, apparently, influences the present results, for example, by including FFB for post-airway surgery evaluation. Some of the indications for FFB in the PICU are not different from those in non-PICU patients.4,5 Nevertheless, the medical status is significantly www.chestjournal.org
worse, manifested by the relatively low Pao2/Fio2 values. A higher risk for complications is, therefore, anticipated. The evaluation of airway abnormalities constitutes the major indication for FFB in the present study (74% of all procedures). This is in accordance with previous reports.4,5 An earlier report6 of 104 FFBs in PICU patients found a higher rate: 92%. However, BAL was not included and pulse oximetry was not universally used. All three studies4 – 6 showed a substantial contribution of FFB performed for anatomic evaluation of the upper and lower airways (76 to 95%). In the present study, the rate of positive findings was lower: 69%. Godfrey et al4 and Nussbaum5 studied heterogeneous groups of patients and included a substantial number of noncritical and ambulatory outpatients.4,5 Their reports included patients with noisy breathing, congenital stridor, and adenoid hypertrophy in whom the expected rate of positive findings is almost 100%. Our data are supported by a publication in Spanish that reported a 65% rate of airway abnormalities in 51 FFBs performed in PICU patients.9 The definition of positive contribution may be different in PICU patients. In the setting of the pulmonary services, patient/parent reassurance (negative findings) was included under the definition of a “clinically meaningful contribution.”5 This will hardly be the case in the PICU environment. The diagnosis of “normal airways” carries an important clinical and practical contribution to patient management and decision making regarding extubation, particularly in the postoperative airway surgery cases. For example, the observation of patent, noncollapsing airways following specific tracheobronchial procedures such as reconstruction surgeries and stenting may direct toward extubation. Hence, the yield of FFB should include any contributive information that cannot be obtained by other means. The contribution of FFB is evident also by the fact that in almost one third of the CHEST / 126 / 4 / OCTOBER, 2004
1357
cases, patient management changed from conservative to surgical as a result of the FFB findings. Whereas cytologic examination of alveolar fluid is of great interest in ambulatory patients, BALs in a PICU are employed mainly for the diagnosis of lung infections.10,11 Most of BAL procedures (90%) were performed to isolate organisms causing pulmonary infections. The advantage of BALF microbiologic studies over the regular tracheal aspirates was clearly demonstrated by Fagon et al,12 in terms of both decreased usage of antibiotics and lower mortality rates. BAL lowers the number of cases treated incorrectly due to tracheal colonization, thus allowing a more judicious use of antimicrobials.13 Although the present study was not designated to compare BAL to tracheal aspirates, the findings correspond with these observations. In the present series, negative and positive microbiologic BALF results changed treatment in 6 of 11 cases and 9 of 14 cases, followed by a marked clinical improvement in 4 of 6 cases and 8 of 9 cases, respectively. Negative microbiologic BALF findings are important as they may accelerate the search for other foci of infection and lead to cessation or to downgrading of antibiotic treatment. Therapeutic procedures accounted for 25% of 200 ambulatory and PICU bronchoscopies reported by Godfrey et al,4 and for 27% of 51 FFBs performed in PICU patients in the series of Garcia et al.9 This is similar to the 24% rate of interventional procedures performed in the present cohort, most of which were due to a collapsed lung segment refractory to chest physiotherapy and tracheal suction. A 74% success rate of treating lung atelectasis by FFB and lavage was found in the present series. This rate is similar to the findings in non-PICU infants and children.4,14 The 80% rate of successful extubation following FFB and BAL in patients who had failed earlier because of atelectasis highlights the importance of targeted, under-vision removal of airway plugs in critically ill infants and children. With the recent introduction of the 2.8-mm bronchoscope that has a suction channel, this has become a therapeutic option also in infants receiving ventilation via a small ETT. We recommend that the procedure should be considered early, especially in patients receiving ventilation. FFB may have particular contribution in children with dysfunctional cough reflexes, due to postoperative pain, neuromuscular diseases7 and CNS depression. The role of FFB in the perioperative management of airway surgery patients has not been established yet. In one third of the postoperative investigations, most of which were performed just prior to a planned extubation, FFB provided information that changed the course toward airway reintervention. Such cases in1358
cluded, for example, obstructing granulation tissue in tracheal anastomoses and airway stent dislodgment. Generally, FFB is regarded as a very safe procedure. In a report5 of 2,836 procedures performed in both PICU and ambulatory patients, no related mortality was recorded. Our results show that it can be safely performed also in critically ill infants and children. Severe complications of FFB are rare; nevertheless, side effects are common. Most often, these are short lived. Examples are transient hypoxemia or hypercarbia due to partial airway blockage or to inadvertent clearance of surfactant during BAL. Many of the FFBs were performed in hypoxemic children with decreased Pao2/Fio2 ratios. These patients belonged mainly to the nonairway surgery groups (ie, groups A and B). Nevertheless, significant hypoxemia did not complicate any of the procedures in this series and no increase of ventilatory variables was required. This may be attributed to the protective assisted positive pressure ventilatory support and the application of Fio2 of 1.0 during the procedure that may have prevented deterioration in respiratory status in critical patients. Also, the assisted ventilation has possibly prevented hypoxemia and hypercarbia due to hypoventilation that may occur secondary to the effects of sedative agents.15 Overall, there was a statistically significant decrease in BP during the procedure. This may be attributed to the vasodilator and myocardial depression properties of the anesthetic agents employed.16,17 Unintentional rises in positive end-expiratory pressure during the procedure may aggravate BP drops.15 However, the incidence of true hypotension was relatively low, and the drop in BP was mild and either short lived or responded promptly to fluid management and did not require additional vasopressor treatment. Hence, the clinical significance of the decrease in BP is minor. This series shows that FFB is used in a myriad of PICU scenarios and contributes substantially to the management of the critically ill child. Therefore, we suggest that due to its frequent use in sick infants and children and due to the possible need for urgent endoscopy also at times when pulmonary services may not be immediately available, FFB should be a PICU expertise employed by intensivists without relying exclusively on pulmonary service support. This recommendation may not be universal and may be applicable especially in centers where FFB is performed frequently in the PICU. It is advisable that in order to be qualified to perform FFBs independently, “hands-on” courses should be attended. Such courses presently take place in the United States and Europe. It has been suggested that each trainee assist in 50 procedures and perform 25 supervised procedures.7 Bronchoscopy
Conclusion Pediatric FFB is a safe diagnostic and interventional tool in the management of very sick infants and children hospitalized in a PICU. FFB in this unique population may be indicated in situations that are different from those encountered in outpatient clinics and “regular” pulmonary practice. Owing to the demand for such a powerful bedside tool, PICU professionals are encouraged to acquire the appropriate skills and be accredited for performing FFB. ACKNOWLEDGMENT: In memory of Gabriel Hugo BarZohar.
References 1 Nussbaum E. Flexible fiberoptic bronchoscopy and laryngoscopy in children under 2 years of age: diagnostic and therapeutic applications of a new pediatric flexible fiberoptic bronchoscope. Crit Care Med 1982; 10:770 –772 2 Fan LL, Sparks LM, Dulinski JP. Applications of an ultrathin flexible bronchoscope for neonatal and pediatric airway problems. Chest 1986; 89:673– 676 3 Wood RE. Pitfalls in the use of the flexible bronchoscope in pediatric patients. Chest 1990; 97:199 –203 4 Godfrey S, Avital A, Maayan C, et al. Yield from flexible bronchoscopy in children. Pediatr Pulmonol 1997; 23:261– 269 5 Nussbaum E. Pediatric fiberoptic bronchoscopy: clinical experience with 2,836 bronchoscopies. Pediatr Crit Care Med 2002; 3:171–176 6 Fan LL, Sparks LM, Fix FJ. Flexible fiberoptic endoscopy for
www.chestjournal.org
7 8 9 10
11 12
13
14 15 16 17
airway problems in a pediatric intensive care unit. Chest 1988; 93:556 –560 Balfour-Lynn IM, Spencer H. Bronchoscopy: how and when? Paediatr Respir Rev 2002; 3:255–264 Guidelines 2000 for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Currents in Emergency Cardiovascular Care, Fall 2000. Vol. 11, No. 3 Garcia GE, Perez RE, Quesada RL, et al. Usefulness of fiberoptic bronchoscopy in critical pediatric care. Arch Bronconeumol 1999; 35:525–528 Leigh MW, Henshaw NG, Wood RE. Diagnosis of Pneumocystis carinii pneumonia in pediatric patients using bronchoscopic bronchoalveolar lavage. Pediatr Infect Dis J 1985; 4:408 – 410 Stokes DC, Shenep JL, Parham D, et al. Role of flexible bronchoscopy in the diagnosis of pulmonary infiltrates in pediatric patients with cancer. J Pediatr 1989; 115:561–567 Fagon JY, Chastre J, Wolff M, et al. Invasive and non-invasive strategies for the management of suspected ventilator-associated pneumonia: a randomized trial. Ann Intern Med 2000; 132:621– 630 Ben-Ari J, Yaniv I, Nahum E, et al. Yield of bronchoalveolar lavage in ventilated and non-ventilated children after bone marrow transplantation. Bone Marrow Transplant 2001; 27: 191–194 Holmgren NL, Cordova M, Ortuzar P, et al. Role of flexible bronchoscopy in the re-expansion of persistent atelectasis in children. Arch Bronconeumol 2002; 38:367–371 Masters IB, Cooper P. Position paper: paediatric flexible bronchoscopy. J Paediatr Child Health 2002; 38:555–559 Tobias JD. Sedation and anesthesia for pediatric bronchoscopy. In: Prakash UBS, ed. Bronchoscope. New York, NY: Raven Press, 1994; 345–356 Pue CA, Pacht ER. Complications of fiberoptic bronchoscopy at a university hospital. Chest 1995; 107:430 – 432
CHEST / 126 / 4 / OCTOBER, 2004
1359