Tracheostomy decannulation is a process that can create

Transcription

1 Original Research Pediatric Otolaryngology Using Polysomnography and Airway Evaluation to Predict Successful Decannulation in Children Otolaryngology Head and Neck Surgery 215, Vol. 153(4) Ó American Academy of Otolaryngology Head and Neck Surgery Foundation 215 Reprints and permission: sagepub.com/journalspermissions.nav DOI: / Neepa Gurbani, DO 1, Ussana Promyothin, MD 2, Michael Rutter, FRACS 3, Matthew C. Fenchel, MS 1,4, Rhonda D. Szczesniak, PhD 1,4, and Narong Simakajornboon, MD 1 Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article. Abstract Objective. Decannulation readiness is approached in several ways and is tailored to the individual patient. While microlaryngoscopy and bronchoscopy evaluate airway patency at all levels, polysomnography assesses sleep-related upper airway physiology. There are limited data in the literature on the utility of these tools. Thus, the main objectives of this study were (1) to identify sleep and respiratory parameters associated with successful decannulation and (2) to determine the agreement between microlaryngoscopy and bronchoscopy and polysomnography. Study Design. Case series with chart review. Setting. Quaternary care pediatric hospital. Subjects and Methods. A retrospective review of medical records, microlaryngoscopy and bronchoscopy, and polysomnographs was performed on subjects preparing for decannulation from 25 to 21. Fifty-nine subjects who underwent overnight polysomnography with a tracheostomy capping trial and microlaryngoscopy and bronchoscopy were included. Prediction of successful decannulation from polysomnography parameters was determined using logistic regression analysis. Results. Fifty-nine subjects with a total of 78 polysomnographs were subdivided into 2 groups: 42 polysomnographs were classified as successfully decannulated, and 36 polysomnographs belonged to the group that did not decannulate. Logistic regression analysis determined that the Apnea Hypopnea Index (P =.423) and maximal end-tidal CO 2 (P =.46) were significant predictors for successful decannulation. Conclusion. Airway evaluation by microlaryngoscopy and bronchoscopy is an essential tool in the assessment of decannulation readiness. Polysomnography is an important additional tool for children with complex airway problems. Our results indicate that certain polysomnographic parameters such as Apnea Hypopnea Index and maximal endtidal CO 2 are valuable in predicting successful tracheostomy decannulation. Keywords tracheostomy decannulation, microlaryngoscopy and bronchoscopy, polysomnography Received January 23, 215; revised May 12, 215; accepted May 26, 215. Tracheostomy decannulation is a process that can create anxiety in the patient, parent, and physician, but at the same time, it is a critical step in recovery from a chronic condition. 1,2 In children, the observed tracheostomy decannulation success rates range from 35% to 98%. 3,4 The timing and process of decannulation are dependent on several factors. Clinical readiness for decannulation involves cessation of need for mechanical ventilation and supplemental oxygen for at least 3 to 6 months and resolution of the original indication for tracheostomy. Comorbidities affecting the need for tracheostomy, including cardiac, pulmonary, or neurologic conditions, should have improved or resolved. A likelihood of needing surgery that may affect the airway caliber in a child would support the maintenance of the tracheostomy. 5 At Cincinnati Children s Hospital Medical Center (CCHMC), pulmonary and ear, nose, and throat physicians implement several methods to assess for decannulation readiness including (1) clinical observation and tracheostomy size reduction; (2) successful capping trial at home 1 Department of Pediatrics, Division of Pulmonary and Sleep Medicine, Cincinnati Children s Hospital Medical Center, Cincinnati, Ohio, USA 2 Department of Otorhinolaryngology, Royal Army Hospital, Bangkok, Thailand 3 Department of Pediatrics, Division of Otolaryngology, Cincinnati Children s Hospital Medical, Center, Cincinnati, Ohio, USA 4 Division of Biostatistics and Epidemiology, Cincinnati Children s Hospital Medical Center, Cincinnati, Ohio, USA Part of this research was presented at the 211 and 212 American Thoracic Society meetings (May 13-18, 211, Denver, Colorado; May 18-23, 212, San Francisco) and at the 211 International Congress on Pediatric Pulmonology (CIPP); June 25-27; Versailles, France. Corresponding Author: Neepa Gurbani, DO, Cincinnati Children s Hospital Medical Center, 3333 Burnet Ave. MLC#221, Cincinnati, OH 45229, USA. Neepa.Gurbani@cchmc.org

2 65 Otolaryngology Head and Neck Surgery 153(4) during the day and night, followed by hospital admission for a 48-hour capping trial; (3) microlaryngoscopy and bronchoscopy (MLB) findings, and (4) polysomnography (PSG) with capping trial, which is performed in approximately 25% of patients who are decannulated in our center. Decannulation readiness is approached in several ways and is tailored to the individual patient. Certain assessments of airway form and function are important in all patients prior to decannulation. While MLB establishes airway patency at all levels, PSG assesses physiologic function of the upper airway during sleep when the pharyngeal muscle tone is decreased. However, there are limited data in the literature on the utility of specific parameters of PSG in the assessment of decannulation readiness. Thus, the 2 main objectives of this study were (1) to identify respiratory and sleep parameters that predict successful decannulation () and (2) to determine the agreement between MLB and PSG. Methods A retrospective chart review was performed at CCHMC from February 25 to March 21. Patients who had a diagnostic PSG with a capping trial and an MLB within 2 months of PSG were included. Patients with inadequate sleep study time of less than 4 hours were excluded from the study. A total of 59 patients met the criteria to be included in the study. The study was approved by the Institutional Review Board at CCHMC. PSGs with capping trial were performed using the Grass system (Grass Telefactor, West Warwick, Rhode Island, USA). The standard pediatric montage was used, and the following parameters were recorded simultaneously: bilateral electroocculogram, electroencephalography (C3A2, C4A1, O1A2, O2A1), chin electromyogram, anterior tibialis electromyogram, tracheal microphone, electrocardiography, pulse oximetry and pulse waveform, thoracic and abdominal inductance plethysmography, nasal thermistor, nasal pressure transducer (Protech, Mukilteo, Washington, USA), endtidal CO 2 (ETCO 2 ; BCI Capnochecks). The PSG was performed while the tracheostomy tube was capped. Criteria for uncapping the tracheostomy tube in our center included 1 obstructive events per hour, persistent oxygen desaturation below 9%, significant increase in the work of breathing, or increase in the ETCO 2 by 15 mm Hg above baseline. The PSGs were performed in accordance with the American Academy of Sleep Medicine guidelines. 6 They were scored by registered sleep technologists and were reviewed by board-certified pediatric sleep specialists. The following definitions were used: obstructive apnea was defined as cessation of air flow in the presence of continued respiratory effort. Central apnea was defined as the absence of both airflow and respiratory effort. Hypopnea was defined as 5% decrease in the amplitude of nasal pressure lasting for at least 2 missed breaths, associated with arousal or desaturation 3%, or both. Hypoventilation was defined as 25% of the total sleep time with ETCO 2.5 mm Hg. Apnea Hypopnea Index (AHI) was defined as the total number of respiratory events per hour of sleep. Obstructive Index (OI) was defined as the total number of obstructive apnea and hypopnea per hour of sleep. Obstructive sleep apnea (OSA) was defined as OI.1 event per hour. Mild OSA was defined as OI from 1 to 5. Moderate-to-severe OSA was defined as OI.5. MLB and PSG were divided into 2 categories: favorable and unfavorable. A favorable PSG was defined as OI 5 and AHI 1 and no evidence of hypoventilation. An unfavorable PSG was defined as OI.5 orahi.1 or evidence of hypoventilation. MLB outcomes were categorized as favorable based on the clinical decision to proceed with a capping trial after the procedure. At CCHMC, there were 4 surgeons who performed most of the MLB prior to decannulation during the reviewed period. A review of MLB operative reports revealed certain airway abnormalities that may have led to an unfavorable MLB or decision to not attempt a decannulation trial. These airway abnormalities included significant laryngomalacia, tracheomalacia, vocal cord paralysis, suprastomal collapse, glottic web, subglottic stenosis grade II and higher, laryngeal cysts, vallecular cysts, A-frame deformity, intra-arytenoid banding, suprastomal granulation tissue with significant obstruction, complete tracheal rings, tracheal mass, and scarring from previous airway reconstruction with airway obstruction. Successful decannulation () was defined as staying decannulated for at least 2 years based on chart review. Patients who did not decannulate () either were deemed not to be fit for decannulation based on the physician s evaluation or were decannulated and later required reinsertion of the tracheostomy tube. Descriptive statistics were calculated for key demographic and PSG variables. Means and standard deviations were reported for continuous variables and frequencies, and percentages were reported for categorical variables. To assess differences between the 2 groups ( vs ), the Wilcoxon rank-sum test was used for continuous variables, and the likelihood x 2 test was used for categorical variables. Logistic regression methods were used to model the probability (log-odds) of in relation to various continuous and categorical predictors, including age at the time of sleep study, sex, degree of OSA, favorable MLB, favorable PSG, as well as sleep parameters including AHI, OI, maximum ETCO 2, and stages of sleep. Individual logistic regression analysis for each independent variable was performed to identify predictors of. Multiple logistic regression analyses were conducted with significant predictors to determine which combination of factors was associated with. Receiver-operating curve (ROC) analyses, using area under the curve (AUC), were conducted to measure the accuracy of MLB along with sleep parameters in predicting. Statistical significance was set a priori at a =.5. SAS (Cary, North Carolina, USA) 9.3 software was used for all analyses. Results A total of 78 PSGs (in 59 patients) met the criteria for entry into analysis and were subdivided into 2 groups: 42 PSGs

3 Gurbani et al 651 Age at PSG (years) Figure 1. Demographic characteristics. This graph shows the age at polysomnography for both groups (successful decannulation and no decannulation). were classified as and 36 PSGs were classified as. Of 36 PSGs in the group, 6 had their tracheostomy tubes uncapped during the sleep study because of significant obstruction or respiratory compromise (only PSG data during the capping portion of the study was used for analysis), 2 had subsequent reinsertion of the tracheostomy tube, 1 was decannulated but required noninvasive positive pressure ventilation with bilevel positive airway pressure, and the remainder were clinically deemed not to be fit for decannulation. There were no significant differences in the clinical characteristics between the various subgroups of the group. The demographics of the included patients are shown in Figure 1 and Table 1. The average age at tracheostomy was 2.7 years (range, -16 years), and the average age at decannulation was 6.7 years (range, 3-19 years). The mean ages of the patients at the time of the PSG in the and groups were 6.8 years and 5.2 years, respectively, with a range of 1 to 17 years. With regard to diagnosis, 16 patients had craniofacial abnormalities (27.1%), 15 patients had airway abnormalities (25.4%), 1 patients had chronic lung disease (16.9%), 9 patients had genetic disorders (15.3%), 4 patients had disorders of respiratory control (6.8%), and 5 patients had neuromuscular disorders (8.4%). There were 17 patients who belonged to both the group and group. These patients initially were in the group and then had airway surgery that later allowed them to be decannulated successfully. Surgical interventions performed in the interim included adenotonsillectomy, genioglossus advancement, mandibular distraction, arytenoidectomy, laryngotracheal reconstruction, somnoplasty, removal of suprastomal granulation tissue, and ligation/excision of the salivary glands. The average time between the first assessment and was 18.4 months, with a range of 3 to 48 months. Analysis of the type of sleep-related breathing abnormalities revealed that subjects who were decannulated successfully had a significantly lower proportion of moderate-to-severe OSA (14% [] vs 37% [], P \.5) and alveolar hypoventilation (2% [] vs 4% [], P \.5), as shown in Figure 2. Analysis of the sleep and respiratory parameters revealed that children in the group had lower AHI ( /h [] vs /h [], P \.5), OI ( /h [] vs /h [], P \.5), and maximal ETCO 2 ( mm Hg [] vs mm Hg [], P \.5), as shown in Figure 3. No significant differences were found in central apnea index ( [] vs []), arousal index ( [] vs []), sleep efficiency ( [] vs []), or sleep stage distribution: NREM 1 ( [] vs []), NREM 2 ( [] vs []), NREM 3 ( [] vs []), and REM ( [] vs []), as shown in Figures 3, 4, and5. Analysis of the predictive value revealed that both MLB and PSG were sensitive tests, as shown in Figure 6. Although 93% of the patients in the group had a favorable MLB, 54% of the patients in the group also had a favorable MLB. Similarly, 74% of the patients in the group had a favorable PSG, but 54% of the patients in the group also had a favorable PSG. Using both favorable MLB and PSG to predict increased the sensitivity from 93% (favorable MLB alone) to 98%. In addition, requiring both favorable MLB and PSG increased the specificity significantly from 46% (unfavorable MLB alone) to 69% (unfavorable MLB and PSG), as shown in Figure 6. The Table 1. Underlying Patient Diagnosis. Diagnosis Number of Patients Who Did Not Decannulate Number of Patients with Successful Decannulation Total Number of Patients Airway abnormality Craniofacial abnormality Neuromuscular disease Disorder of respiratory control Chronic lung disease Genetic abnormality 3 6 9

4 652 Otolaryngology Head and Neck Surgery 153(4) Mild OSA (%) P<.5 Mod-severe OSA(%) P<.1 Hypoventilation% P<.5 Figure 2. Percentage of patients in each group (successful decannulation and no decannulation) who had mild obstructive sleep apnea (OSA), moderate to severe OSA, and hypoventilation. AHI (/hr) P<.5 OI (/hr) P<.5 Max. ETCO2 (mmhg) P<.5 Figure 3. Comparison of respiratory sleep parameters (Apnea Hypopnea Index, Obstructive Index, central index, and maximal end-tidal CO 2 ) between the 2 groups (successful decannulation and no decannulation). Central index (/hr) Arousal Index (/hr) Sleep Efficiency (%) Figure 4. Comparison of sleep parameters (sleep efficiency and arousal index) between the 2 groups (successful decannulation and no decannulation). agreement between favorable MLB and PSG was 71% in the group and 54% in the group. Results of the logistic regression analysis are shown in Table 2.TheAHI(P =.423) and maximal ETCO 2 (P =.46) were significant predictors for. The ROC curves from the model using AHI along with a favorable MLB, compared with a favorable MLB alone, showed that the combination of AHI and a favorable MLB (AUC,.7755) was significantly

5 Gurbani et al 653 NREM Stage 1 (%) NREM Stage 2 (%) NREM Stage 3 (%) Figure 5. Comparison of the sleep stage distribution (non rapid eye movement and rapid eye movement sleep) between the 2 groups (successful decannulation and no decannulation). REM (%) % 98% 8 74% 69% 6 54% 54% 46% 46% 4 26% 31% 2 7% 2% Table 2. Modeling Decannulation Success/Failure: Fixed Effects Estimates for Individual Regressions. Effect ProbF Sex.3666 MLB_fav.1 PSG_fav.586 MLBPSG fav.85 Age PSG_.2222 Age MLB.2494 Apnea Hypopnea Index.423 Maximum end-tidal CO 2.46 Obstructive Index.786 Figure 6. Percentage of patients in the 2 groups (successful decannulation and no decannulation) who had favorable microlaryngoscopy and bronchoscopy (MLB), favorable polysomnography (PSG), unfavorable MLB, unfavorable PSG, both favorable MLB and PSG, and both unfavorable MLB and PSG. better (P =.4) at predicting decannulation than using a favorable MLB alone (AUC,.6865), as shown in Figure 7. Discussion Our study sought to evaluate the role of PSG and MLB in predicting. Our results indicate that the PSG parameters such as AHI, OI, and hypoventilation are significant predictors of. Further analysis showed that favorable PSG parameters including AHI 1, OI 5, and absence of hypoventilation (ETCO 2.5 mm Hg less than 25% of total sleep time) are good predictors of. We also determined that a favorable MLB is an excellent predictor of. Using MLB in conjunction with the PSG parameters such as AHI, OI, and hypoventilation significantly increases the predictability of. More importantly, adding a favorable PSG decreases the likelihood of. The agreement between favorable MLB and PSG is 71% in the group and 54% in the group. This discrepancy can be explained by the fact that MLB evaluates anatomic abnormalities and PSG assesses physiologic function and dynamics during sleep. A favorable MLB is an important tool in assessing for decannulation readiness. In addition, a favorable PSG with tracheostomy capping complements endoscopic assessment in patients with complex airway problems. There are several methods used to approach decannulation. Ceriana et al 7 have developed a protocol for adult tracheostomy decannulation that includes 7 criteria: (1) absence of distress and stable arterial blood gases on prolonged mechanical ventilation for 5 days; (2) stable condition including hemodynamic stability, PaCO 2 \6 mm Hg, and absence of fever, sepsis, or active infection; (3) normal endoscopic examination or examination revealing stenotic lesions occupying \3% of the airway; (4) absence of

6 654 Otolaryngology Head and Neck Surgery 153(4) Figure 7. Model represents the Apnea Hypopnea Index and favorable microlaryngoscopy and bronchoscopy (MLB) as predictors of successful decannulation compared with favorable MLB only. delirium or psychiatric disorders; (5) adequate swallowing evaluated by gag reflex, blue dye, and video fluoroscopy; (6) ability to expectorate on request; and (7) maximum expiratory pressure 4 cm of water. Gao et al 8 have also described airway function evaluation prior to decannulation with the use of upper airway resistance measurements during shallow and deep breathing. They suggested that subjects with higher inspiratory and expiratory resistances are unsuitable for decannulation. In children, several methods were examined to evaluate airway patency and function prior to decannulation. Mallory et al 9 used oral and tracheostomal peak inspiratory airflow measurements during awake and tidal breathing, with a ratio of greater than 1:4 (oral: tracheostomal peak inspiratory flow) for predicting. Willis et al 2 approached decannulation readiness with flexible endoscopy to assess laryngeal dynamics and rigid endoscopy to evaluate airway patency. Tunkel et al 1 recommended the use of PSG along with endoscopic assessment and successfully decannulated 13 of 16 patients with a favorable PSG. Based on this study, a multistep protocol for decannulation in children is suggested as follows: (1) assess for clinical readiness; (2) perform flexible laryngoscopy; (3) downsize the tracheostomy tube over several weeks and start daytime plugging; (4) perform MLB to remove suspected granuloma; (5) perform PSG with occluded, downsized tracheostomy tube; and (6) decannulate in the intensive care unit and observe for 24 hours if the PSG is reassuring. Subjects with a favorable PSG in this study had an OI \1.7 events per hour, \2.7 central events per hour, mean arterial oxygen saturation.95%, and maximum ETCO 2 between 38 and 46 mm Hg. Mukherjee et al 11 also supported the use of PSG in pediatric patients prior to decannulation. Subjects with a favorable PSG in this study had an OI \2 events per hour, \2.1 central events per hour, mean arterial oxygen saturation.96% in 12 of 22 patients, and normal ETCO 2 measurements. In our experience, PSGs with a capped tracheostomy tube usually reveal a mild degree of OSA that is likely due to the presence of a small tracheostomy tube in the airway. The decannulation process described by Tunkel et al 1 is comparable to the decannulation process at CCHMC. Our study confers a favorable PSG status with an OI 5; however, Tunkel et al 1 and Mukherjee et al 11 described a favorable PSG as OI \1.7 and 2 events per hour, respectively. These authors also determined that normal ETCO 2 measurements during sleep are an important factor in a favorable PSG. In contrast to these studies, we found that central apnea is not a significant factor in a favorable PSG for. The discrepancy may reflect population differences between the studies. Our sample size is larger, with 59 subjects compared with 24 and 31 subjects in the previous studies. 1,11 In addition, the subjects in our study were older, and a significant proportion of them had complex airway abnormalities. The mean age of the patients at the time of the PSG in the and groups was 6.8 years and 5.2 years, respectively, in our study, while previous studies reported the median age of the patients at the time of the PSG as 37 months and 27 months. 1,11 While previous studies focused only on parameters of favorable and unfavorable PSG, our study examined the combination of PSG and MLB results in predicting decannulation. Our results are more likely to reflect how clinicians approach decannulation in complex cases. Previous studies reported that 13 of 16 patients and 21 of 22 patients with a favorable PSG were decannulated successfully. 1,11 In our study, 74% of the patients in the group and 54% of the patients in the group had a favorable PSG. Using both favorable MLB and PSG to predict increased the sensitivity from 93% to 98% and the specificity from 46% to 69%. However, 31% of the patients in the group had a favorable MLB and PSG. It is likely that this group had subtle abnormalities during evaluation that led to the clinician s decision not to decannulate, despite favorable MLB and PSG. This finding highlights the complexity of our study population, which included older children with complex airway abnormalities. Nonetheless, this group of patients would benefit most from a comprehensive evaluation of the anatomic and physiologic function of the airway prior to decannulation, which is the approach used at our institution. Our study has certain limitations. First, this is a retrospective study with a heterogeneous population with various diagnoses. However, this is the nature of the population requiring decannulation. Since the number of patients undergoing decannulation is relatively small, it is not possible to examine a homogeneous population. Second, PSG with capping trial is done in approximately 25% of the patients getting ready for decannulation at our center. Therefore, there is a potential issue for selection bias. In fact, our patients are older and are likely to have complex medical problems prompting physicians to obtain additional assessments.

7 Gurbani et al 655 Further studies are necessary to evaluate whether our findings are applicable to all patients who undergo tracheostomy decannulation. Third, our decision to do a tracheostomy capping PSG is depending on the result of MLB. Patients judged to be ready for a capping PSG would have airway evaluation by MLB first. Therefore, MLB and PSG results are not totally independent of each other. Finally, a few patients were included in both the and groups due to some patients in the group undergoing airway surgeries that may have led to. Conclusion Clinicians use various methods to assess for decannulation readiness, and the approach is often variable. MLB is the gold standard for evaluation of airway anatomy prior to decannulation. Recently, the role of PSG has expanded in the pediatric patient population with obstructive and central sleep apnea, neuromuscular disorders, and ventilatordependent patients. Based on our study, certain PSG parameters including OI, AHI, and hypoventilation along with a favorable MLB appear to be good predictors of in patients with a complex airway. Using MLB and PSG together increases the predictability of in these children over either factor alone. Therefore, PSG should be considered as an important part of the evaluation process for decannulation, especially in children with complex airway problems. Larger studies are needed to validate these specific PSG parameter thresholds in all pediatric patients undergoing decannulation. Acknowledgments The authors thank J. Denise Wetzel, CCHMC medical writer, for editing the manuscript. Authors Note Dr Gurbani was selected as a finalist for the Young Investigator Award at CIPP 211. Author Contributions Neepa Gurbani, conception and design, data acquisition, analysis, drafting and critical review of the manuscript, final approval, accountability for all aspects of the work; Ussana Promyothin, data acquisition, critical review of the manuscript, final approval, accountability for all aspects of the work; Michael Rutter, data interpretation, critical review of the manuscript, final approval, accountability for all aspects of the work; Matthew C. Fenchel, data analysis, critical review of the manuscript, final approval, accountability for all aspects of the work; Rhonda D. Szczesniak, data analysis, critical review of the manuscript, final approval, accountability for all aspects of the work; Narong Simakajornboon, conception and design, data interpretation and analysis, critical review of the manuscript, final approval, accountability for all aspects of the work. Disclosures Competing interests: None. Sponsorships: None. Funding source: Cincinnati Children s Hospital Research Fund, study design and conduct; collection, analysis, and interpretation of the data; and writing or approval of the manuscript. References 1. Leung R, Berkowitz RG. Decannulation and outcome following pediatric tracheostomy. Ann Otol Rhinol Laryngol. 25; 114: Willis R, Myer C, Miller R, Cotton RT. Tracheotomy decannulation in the pediatric patient. Laryngoscope. 1987;97: Carron JD, Derkay CS, Strope GL, Nosonchuk JE, Darrow DH. Pediatric tracheotomies: changing indications and outcomes. Laryngoscope. 2;11: Wetmore RF, Marsh RR, Thompson ME, Tom LW. Pediatric tracheostomy: a changing procedure? Ann Otol Rhinol Laryngol. 1999;18(7 pt 1): Gray RF, Todd NW, Jacobs IN. Tracheostomy decannulation in children: approaches and techniques. Laryngoscope. 1998; 18(1 pt 1): Iber C, Ancoli-Israel S, Chesson A, Quan SF. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. 1st ed. Weschester, Illinois: American Academy of Sleep Medicine; Ceriana P, Carlucci A, Navalesi P, et al. Weaning from tracheotomy in long-term mechanically ventilated patients: feasibility of a decisional flowchart and clinical outcome. Intensive Care Med. 23;29: Gao C, Zhou L, Wei C, Hoffman MR, Li C, Jiang JJ. The evaluation of physiologic decannulation readiness according to upper airway resistance measurement. Otolaryngol Head Neck Surg. 28;139: Mallory GB Jr, Reilly JS, Motoyama EK, Mutich R, Kenna MA, Stool SE. Tidal flow measurement in the decision to decannulate the pediatric patient. Ann Otol Rhinol Laryngol. 1985;94(5 pt 1): Tunkel, McColley SA, Baroody FM, Marcus CL, Carroll JL, Loughlin GM. Polysomnography in the evaluation of readiness for decannulation in children. Arch Otolaryngol Head Neck Surg. 1996;122: Mukherjee B, Bais AS, Bajaj Y. Role of polysomnography in tracheostomy decannulation in the paediatric patient. J Laryngol Otol. 1999;113: