O objetivo do presente trabalho foi analisar por meio da videolaringoscopia de alta velocidade o padrão vibratório das pregas vocais de indivíduos com e sem nódulo de prega vocal. Além disso, comparar o padrão vibratório das pregas vocais, de todos os grupos, analisado pela videolaringoscopia de alta velocidade com os resultados das análises acústica e quimográfica. Participaram deste estudo 36 indivíduos, do gênero feminino, com idades entre 18 e 45 anos, divididos em 3 grupos: grupo controle (GC), com 18 mulheres sem lesão nas pregas vocais e com vozes normais, grupo com nódulo vocal incipiente (GNI) com 8 mulheres com nódulo vocal incipiente e grupo com nódulo vocal (GN) composto por 10 mulheres com nódulo vocal de pequeno a grande. Procedimentos: gravação vocal, para análise acústica e perceptivo-auditiva, e exame laríngeo com a videolaringoscopia de alta velocidade para análise das imagens de alta velocidade (HSV) e da quimografia digital (DKG). Resultados: Encontraram-se tempos de fase do ciclo glótico semelhantes entre o grupo controle e o grupo com nódulo vocal incipiente, no entanto, significativamente menores do que os tempos de fase dos indivíduos com nódulo vocal, com exceção do tempo de fechamento. Observou-se também diferenças para o quociente de abertura (OQ), no grupo com nódulo vocal a proporção em que a prega vocal ficou aberta em um ciclo foi significativamente maior do que no grupo sem nódulo (OQ = 0,81 ± 0,07 e OQ = 0,68 ± 0,12, respectivamente). Além disso, por meio da quimografia digital pode-se observar que a amplitude de vibração das pregas vocais foi significativamente reduzida no grupo com nódulo vocal em relação aos indivíduos sem nódulo. Discussão e Conclusão: Concluiu-se neste trabalho, por meio da análise da videolaringoscopia de alta velocidade, que o padrão de vibração das pregas vocais foi regular em todos os grupos analisados, apresentando variações mínimas em relação ao período e à amplitude. Na comparação dos grupos, observou-se que o grupo com nódulo vocal apresentou redução da amplitude de vibração das pregas vocais em relação ao grupo controle. Além disso, no grupo com nódulo vocal o quociente de abertura indicou predomínio do tempo de fase aberta. Na comparação das metodologias empregadas concluiu-se que a videolaringoscopia de alta velocidade e a quimografia digital avaliam de forma distinta os tempos de fase do ciclo vibratório, e, consequentemente o quociente de abertura das pregas vocais, cujo cálculo é baseado nestes tempos de fase. Estes resultados alertam a necessidade de estabelecerem-se valores diferentes para a normalidade entre a videolaringoscopia de alta velocidade e a quimografia digital. / This dissertation reports on the analysis of the vocal fold vibration pattern of subjects with and without vocal fold nodule by high-speed videoendoscopy. It also compares the pattern of vocal fold vibration in all groups analyzed with the results of acoustic and kymograph analyses. Thirty six women whose age ranged from 18 to 45 participated of the study. They were divided into 3 groups: control group (GC), with 18 women with no lesions on vocal folds; incipient vocal nodule group (GNI), composed of 8 women with incipient vocal nodule; and vocal nodule group (GN), composed of 10 women with a smallto-large vocal nodule. The procedures comprised voice recording for acoustic analysis and auditory-perceptual and highspeed videoendoscopy for high-speed image analysis and digital kymography (DKG). The results revealed phases of the vibration cycle (open phase, closed phase, opening phase and closing phase) similar to those of the control group and incipient vocal nodule group, but significantly lower than the phase times of the vibration cycle of vocal nodule group, except for the closing phase. Regarding the open quotient (OQ), in the group with vocal nodule the proportion at which the vocal fold was opened in a cycle was significantly higher than the control group (OQ = 0,81 ± 0,07 and OQ = 0,68 ± 0,12, respectively). The digital kymography allowed observing that the vibration amplitude of the vocal folds was significantly reduced in the group with vocal nodule in comparison to the control group. By means of the high-speed videoendoscopy analysis, it was possible to conclude that the pattern of vocal fold vibration was regular in all groups, with minor variations in the period and amplitude. By comparing the groups, the group with vocal nodule decreased the amplitude of vibration of the vocal folds in the control group. Moreover, in the group with vocal nodule the open quotient indicated predominance of the open phase. The comparison of the methodologies allowed concluding that high-speed videoendoscopy and digital kymography asses the time phase of the vibration cycle differently, and hence the open quotient the vocal folds, which is calculated based on these times of phase. The results emphasize the need to set up different values for the normality of high-speed videoendoscopy and digital kymography.

Made available in DSpace on 2014-06-11T19:30:29Z (GMT). No. of bitstreams: 0 Previous issue date: 2013-02-20Bitstream added on 2014-06-13T19:19:20Z : No. of bitstreams: 1 mansano_am_dr_botfm.pdf: 414250 bytes, checksum: d2c4771203d56a53dd54713db8d2edee (MD5) / O manejo correto das vias aéreas em pacientes pediátricos é fundamental da prática anestésica, em ambientes de terapia intensiva ou unidades de emergência. Avaliar preditores clínicos de laringoscopia difícil em pacientes pediátricos, bem como mensurar a eficácia da manipulação laríngea externa otimizada em melhor a visualização da glote durante a laringoscopia. 445 crianças abaixo de 12 anos foram avaliadas, de modo transversal, quanto à presença de características antropométricas relacionadas à laringoscopia difícil. Os pacientes foram divididos em três grupos de acordo com a idade. Peso, altura, circunferência do pescoço, distância interincisivos (DII), distância tireomentoniana, distância esternomentoniana, distância entre o plano frontal e o mento (DPFM) e a classificação de Mallampati modificada foram determinados e correlacionados com a classificação de Cormack e Lehane (CML). A incidência de laringoscopia difícil (CML III ou IV) foi de 11,23%, reduzindo para 3,59% após a manipulação laríngea externa otimizada. Os fatores que estiveram relacionados com dificuldade de laringoscopia difícil foram: distância interincisivos diminuída, DPFM aumentada, distância tireomentoniana ou esternomentoniana diminuídas ou a classificação de Mallampati modificada III ou IV. O índice entre a DPFM e o peso obteve a maior área sob a curva ROC entre as variáveis estudadas. A DPFM e o índice DPFM/peso são os melhores preditores de laringoscopia difícil em crianças. A distância interincisivos também está correlacionada com a laringoscopia difícil. Em pacientes capazes de obedecer ordens simples, o índice de Mallampati modificado também apresentou boa correlação com laringoscopia difícil, embora não possa ser utilizado de forma isolada... / Pediatric airway management is a priority during anesthesia, critical care and emergency medicine. The purpose of this study is to validate bedside tests that predict laryngoscopic difficulty in anesthetized children and to measure improvement in laryngoscopic view by optimal external laryngeal manipulation. 445 children under 12 years of age were recruited in a cross-sectional study to assess the value of some anthropometric measures as predictors of laryngoscopic difficulty. The patients were divided into three groups by age. Weight, height, neck circumference, interincisor distance (IID), short thyromental distance, sternomental distance, frontal plane to chin distance (FPCD) and the modified Mallampati index were determined and were correlated with the Cormack and Lehane classification (CML). The incidence of difficult laryngoscopy (CML III or IV) was 11.23%, been reduced to 3.59% after optimal external laryngeal manipulation. Factors that were significantly associated with laryngoscopic difficulty included short IID, thyromental distance or sternomental distance, high FPCD, and the modified Mallampati index III and IV. The FPCD/weight index exhibited a higher area under the ROC curve than any other variable considered. This study confirms that the FPCD and the FPCD/weight ratio are the most consistent predictors of laryngoscopic difficulty in pediatric patients. The IID also correlated with laryngoscopic difficulty. For children who... (Complete abstract click electronic access below)

Orientador: Norma Sueli Pinheiro Módolo / Banca: Yara Marcondes Machado Castiglia / Banca: Eliana Marisa Ganem / Banca: Leopoldo Muniz da Silva / Banca: José Fernando Amaral Meleti / Resumo: O manejo correto das vias aéreas em pacientes pediátricos é fundamental da prática anestésica, em ambientes de terapia intensiva ou unidades de emergência. Avaliar preditores clínicos de laringoscopia difícil em pacientes pediátricos, bem como mensurar a eficácia da manipulação laríngea externa otimizada em melhor a visualização da glote durante a laringoscopia. 445 crianças abaixo de 12 anos foram avaliadas, de modo transversal, quanto à presença de características antropométricas relacionadas à laringoscopia difícil. Os pacientes foram divididos em três grupos de acordo com a idade. Peso, altura, circunferência do pescoço, distância interincisivos (DII), distância tireomentoniana, distância esternomentoniana, distância entre o plano frontal e o mento (DPFM) e a classificação de Mallampati modificada foram determinados e correlacionados com a classificação de Cormack e Lehane (CML). A incidência de laringoscopia difícil (CML III ou IV) foi de 11,23%, reduzindo para 3,59% após a manipulação laríngea externa otimizada. Os fatores que estiveram relacionados com dificuldade de laringoscopia difícil foram: distância interincisivos diminuída, DPFM aumentada, distância tireomentoniana ou esternomentoniana diminuídas ou a classificação de Mallampati modificada III ou IV. O índice entre a DPFM e o peso obteve a maior área sob a curva ROC entre as variáveis estudadas. A DPFM e o índice DPFM/peso são os melhores preditores de laringoscopia difícil em crianças. A distância interincisivos também está correlacionada com a laringoscopia difícil. Em pacientes capazes de obedecer ordens simples, o índice de Mallampati modificado também apresentou boa correlação com laringoscopia difícil, embora não possa ser utilizado de forma isolada... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Pediatric airway management is a priority during anesthesia, critical care and emergency medicine. The purpose of this study is to validate bedside tests that predict laryngoscopic difficulty in anesthetized children and to measure improvement in laryngoscopic view by optimal external laryngeal manipulation. 445 children under 12 years of age were recruited in a cross-sectional study to assess the value of some anthropometric measures as predictors of laryngoscopic difficulty. The patients were divided into three groups by age. Weight, height, neck circumference, interincisor distance (IID), short thyromental distance, sternomental distance, frontal plane to chin distance (FPCD) and the modified Mallampati index were determined and were correlated with the Cormack and Lehane classification (CML). The incidence of difficult laryngoscopy (CML III or IV) was 11.23%, been reduced to 3.59% after optimal external laryngeal manipulation. Factors that were significantly associated with laryngoscopic difficulty included short IID, thyromental distance or sternomental distance, high FPCD, and the modified Mallampati index III and IV. The FPCD/weight index exhibited a higher area under the ROC curve than any other variable considered. This study confirms that the FPCD and the FPCD/weight ratio are the most consistent predictors of laryngoscopic difficulty in pediatric patients. The IID also correlated with laryngoscopic difficulty. For children who... (Complete abstract click electronic access below) / Doutor

O objetivo do presente trabalho foi analisar por meio da videolaringoscopia de alta velocidade o padrão vibratório das pregas vocais de indivíduos com e sem nódulo de prega vocal. Além disso, comparar o padrão vibratório das pregas vocais, de todos os grupos, analisado pela videolaringoscopia de alta velocidade com os resultados das análises acústica e quimográfica. Participaram deste estudo 36 indivíduos, do gênero feminino, com idades entre 18 e 45 anos, divididos em 3 grupos: grupo controle (GC), com 18 mulheres sem lesão nas pregas vocais e com vozes normais, grupo com nódulo vocal incipiente (GNI) com 8 mulheres com nódulo vocal incipiente e grupo com nódulo vocal (GN) composto por 10 mulheres com nódulo vocal de pequeno a grande. Procedimentos: gravação vocal, para análise acústica e perceptivo-auditiva, e exame laríngeo com a videolaringoscopia de alta velocidade para análise das imagens de alta velocidade (HSV) e da quimografia digital (DKG). Resultados: Encontraram-se tempos de fase do ciclo glótico semelhantes entre o grupo controle e o grupo com nódulo vocal incipiente, no entanto, significativamente menores do que os tempos de fase dos indivíduos com nódulo vocal, com exceção do tempo de fechamento. Observou-se também diferenças para o quociente de abertura (OQ), no grupo com nódulo vocal a proporção em que a prega vocal ficou aberta em um ciclo foi significativamente maior do que no grupo sem nódulo (OQ = 0,81 ± 0,07 e OQ = 0,68 ± 0,12, respectivamente). Além disso, por meio da quimografia digital pode-se observar que a amplitude de vibração das pregas vocais foi significativamente reduzida no grupo com nódulo vocal em relação aos indivíduos sem nódulo. Discussão e Conclusão: Concluiu-se neste trabalho, por meio da análise da videolaringoscopia de alta velocidade, que o padrão de vibração das pregas vocais foi regular em todos os grupos analisados, apresentando variações mínimas em relação ao período e à amplitude. Na comparação dos grupos, observou-se que o grupo com nódulo vocal apresentou redução da amplitude de vibração das pregas vocais em relação ao grupo controle. Além disso, no grupo com nódulo vocal o quociente de abertura indicou predomínio do tempo de fase aberta. Na comparação das metodologias empregadas concluiu-se que a videolaringoscopia de alta velocidade e a quimografia digital avaliam de forma distinta os tempos de fase do ciclo vibratório, e, consequentemente o quociente de abertura das pregas vocais, cujo cálculo é baseado nestes tempos de fase. Estes resultados alertam a necessidade de estabelecerem-se valores diferentes para a normalidade entre a videolaringoscopia de alta velocidade e a quimografia digital. / This dissertation reports on the analysis of the vocal fold vibration pattern of subjects with and without vocal fold nodule by high-speed videoendoscopy. It also compares the pattern of vocal fold vibration in all groups analyzed with the results of acoustic and kymograph analyses. Thirty six women whose age ranged from 18 to 45 participated of the study. They were divided into 3 groups: control group (GC), with 18 women with no lesions on vocal folds; incipient vocal nodule group (GNI), composed of 8 women with incipient vocal nodule; and vocal nodule group (GN), composed of 10 women with a smallto-large vocal nodule. The procedures comprised voice recording for acoustic analysis and auditory-perceptual and highspeed videoendoscopy for high-speed image analysis and digital kymography (DKG). The results revealed phases of the vibration cycle (open phase, closed phase, opening phase and closing phase) similar to those of the control group and incipient vocal nodule group, but significantly lower than the phase times of the vibration cycle of vocal nodule group, except for the closing phase. Regarding the open quotient (OQ), in the group with vocal nodule the proportion at which the vocal fold was opened in a cycle was significantly higher than the control group (OQ = 0,81 ± 0,07 and OQ = 0,68 ± 0,12, respectively). The digital kymography allowed observing that the vibration amplitude of the vocal folds was significantly reduced in the group with vocal nodule in comparison to the control group. By means of the high-speed videoendoscopy analysis, it was possible to conclude that the pattern of vocal fold vibration was regular in all groups, with minor variations in the period and amplitude. By comparing the groups, the group with vocal nodule decreased the amplitude of vibration of the vocal folds in the control group. Moreover, in the group with vocal nodule the open quotient indicated predominance of the open phase. The comparison of the methodologies allowed concluding that high-speed videoendoscopy and digital kymography asses the time phase of the vibration cycle differently, and hence the open quotient the vocal folds, which is calculated based on these times of phase. The results emphasize the need to set up different values for the normality of high-speed videoendoscopy and digital kymography.

Objective: To compare the effects of thiopental, propofol and alfaxalone on arytenoid cartilage motion and establish dose rates to achieve a consistent oral laryngoscopy examination. Study design: Prospective, randomised, blinded crossover study. Animals: Six healthy adult beagle dogs. Methods: Each dog was administered three induction agents in a random order with a one week washout period between treatments. No premedication was used prior to induction of anaesthesia. Thiopental, propofol or alfaxalone were administered at 7.5 mg/kg, 3 mg/kg and 1.5 mg/kg, over 1 minute to effect, for induction of anaesthesia, respectively. If the dog was deemed inadequately anaesthetised then top-up boluses of 1.8 mg/kg, 0.75 mg/kg and 0.4 mg/kg, respectively, were administered over 10 seconds, repeated every 20 seconds, until an adequate anaesthetic plane had been reached. Continual examination of the larynx, using a laryngoscope, commenced once an adequate anaesthetic depth had been reached until recovery from anaesthesia. The number of arytenoid motions and deep inspiratory efforts (vital breaths) were counted within three time periods and compared over time among treatments. Data were analysed using Friedman test, Mann-Whitney U test, Spearman's rho, linear mixed model with post-hoc pairwise comparison with Tukey correction. Results interpreted at a 5% level of significance. Results: The median (range) induction time was 2.8 (2.0, 3.0), 2.7 (2.0, 3.3) and 2.5 (1.7, 3.3) minutes for thiopental, propofol and alfaxalone, respectively (p = 0.727). The median (range) dose rate required to achieve an adequate depth of anaesthesia was 6.3 (6.0, 6.6), 2.4 (2.4, 2.4) and 1.2 (1.2, 1.2) mg/kg/minute for thiopental, propofol and alfaxalone, respectively. Therefore, the median (range) total dose administered over the induction time was 17.8 (13.2, 18.8), 6.8 (5.3, 8.3) and 3.2 (2.3, 4.1) mg/kg for thiopental, propofol and alfaxalone, respectively. There was no significant difference for the total number of arytenoid motions (p =0.662) or vital breaths (p = 0.789) among induction agents. The median (range) examination times were 14.1 (8.0, 41.8), 5.4 (3.3, 14.8) and 8.5 (3.8, 31.6) minutes for thiopental, propofol and alfaxalone, respectively (p=0.016). Conclusion and clinical relevance: There was no significant difference in the total number of arytenoid motions among the induction agents. However, at the dose rates used in this study, propofol provided adequate conditions for evaluation of the larynx within a shorter examination time which may be advantageous during laryngoscopy in dogs. / Mini Dissertation (MMedVet)--University of Pretoria, 2016. / Companion Animal Clinical Studies / MMedVet / Unrestricted

Measurements of intraoral pressure (IOP) and sound pressure level (SPL) were taken of four oboists as they performed two sets of musical exercises: (1) crescendo-decrescendo from pp to ff and back to pp on the pitches D4, G4, C5 and A5, and (2) straight and vibrato performances of the same four pitches at mf. Video images of the vocal tract were also made using flexible fiberoptic nasoendoscopy (FFN). IOP and SPL data were captured in real time by the WinDaq®/Lite software package, with the dB meter located 8-9 inches in directly front of the oboe bell. The study yielded minimum and maximum values from 21.04 to 57.81 mm Hg and from 65.53 to 100.89 dB across all pitches examined. Discussion is included for the following topics: (1) the oboe’s sound envelope, or functional range of IOP and SPL values at different pitch levels, including the nonlinearity in the relationship between IOP and SPL on the oboe, (2) the static activation and kinetic maintenance thresholds for reed vibration, (3) the effect of vibrato on IOP/SPL, (4) the utilization of the vocal tract during execution of dynamic changes and vibrato, and (5) the impact of player experience on control of physical variables.

- / The present study aims to establish the level of diagnostic reliability provided by the transmission of a videolaryngoscopy by videoconference (VC), evaluating the quality of the transmitted images and its impact on diagnoses. Two DVDs with 60 laryngoscopy (30 fiberscopies and 30 telescopies) images, corresponding to 30 original and 30 transmitted by VC, were evaluated by 6 laryngologists who answered a questionnaire on the quality of the selected images (scores from 0 to 3) and on the diagnoses of the related disease. A third DVD with 26 pairs of images, an original and a VC, or two original, was also analyzed by the examiners who had to select which image presented the best quality, which corresponded to the VC image, and the possible impact of such differences on the quality of diagnoses. A high level of agreement (K = 0.911) was found among the examiners in relation to the identification of the VC (96,15% of accuracy) and to the best quality of the original image (K = 0,850). Although a partial agreement has been found among the examiners (K = 0,371), diagnoses was not jeopardized by the quality of the VC image in 75,64% of the cases. Original videotelescopy images (83,33% very good) showed the best quality and videofiberscopy VC the worst (36,66% very good and good). A high degree of correctness was found for the diagnoses by VC images (86,67% by fiberscopy and by telescopy). Consequently, the use of a VC to transmit videolaryngoscopy images has proved to be efficient for the remote diagnoses of larynx diseases

INTRODUCTION: The anesthetic management for pediatric patients undergoing direct laryngoscopy and/or bronchoscopy (DLB) is administered based on the anesthesiologist’s preference. Objectives: The preliminary analysis of this study aims to identify variables that decrease surgical end-to-transport (SET) time and directly impacts patient outcomes. SET time is defined as the time of surgery end to the time of patient exit from the operating room. METHODS: After IRB approval, all DLBs performed at Boston Children’s Hospital (Boston, MA) by the Otolaryngology Department from June 2012 to December 2014 (n= 2419) were obtained from the Anesthesia Information Management System. With a 0.05 level of significance, a multivariate logistic regression was performed in SAS v9.3 with SET time as the dependent variable and surgery duration, age, gender, ASA status, airway device and extubation status as the independent variables. Airway device and extubation status were found to be moderately predictable of each other, so separate models were conducted. Spearman correlation testing was performed to evaluate the relationship between SET time and post-anesthesia care unit (PACU) duration. RESULTS: We excluded cases having ASA classification >2, age >21 years, regional nerve blocks, tracheostomy, nasal cannula, or unknown airway or extubation status. Remaining cases (n = 967) were arranged by SET times and dichotomized by the median value (14 minutes) into two groups (≤14 minutes and >14 minutes). After adjusting for other variables, we found that patients with an endotracheal tube (ETT) were 4.85 times more likely to have a SET time higher than the median, as compared with to those having with a laryngeal mask airway (LMA) (p = 0.0023, 95% CI: 1.76, 13.33). We also found that patients with an ETT were 2.89 times more likely to have a SET time higher than the median compared with those having a mask airway device. (p < 0.0001, 95% CI: 2.09, 3.98). In addition, there was a weak positive correlation between SET time and PACU duration (r = 0.09406, p = 0.0069). DISCUSSION: Preliminary analysis indicates that airway management has a significant impact on SET time after adjusting for surgery duration, age, gender, and ASA status. The use of either a mask or an LMA resulted in a lower SET time than the use of an ETT. The correlation of SET time and PACU duration suggests that reducing SET time does not negatively impact post-operative outcomes and may even be positively, though weakly, correlated. This study is limited by its retrospective nature. Future analysis will include the evaluation of commonly used perioperative anesthetics with dosage and timing variables and their correlation with SET time and patient outcomes.

- / The present study aims to establish the level of diagnostic reliability provided by the transmission of a videolaryngoscopy by videoconference (VC), evaluating the quality of the transmitted images and its impact on diagnoses. Two DVDs with 60 laryngoscopy (30 fiberscopies and 30 telescopies) images, corresponding to 30 original and 30 transmitted by VC, were evaluated by 6 laryngologists who answered a questionnaire on the quality of the selected images (scores from 0 to 3) and on the diagnoses of the related disease. A third DVD with 26 pairs of images, an original and a VC, or two original, was also analyzed by the examiners who had to select which image presented the best quality, which corresponded to the VC image, and the possible impact of such differences on the quality of diagnoses. A high level of agreement (K = 0.911) was found among the examiners in relation to the identification of the VC (96,15% of accuracy) and to the best quality of the original image (K = 0,850). Although a partial agreement has been found among the examiners (K = 0,371), diagnoses was not jeopardized by the quality of the VC image in 75,64% of the cases. Original videotelescopy images (83,33% very good) showed the best quality and videofiberscopy VC the worst (36,66% very good and good). A high degree of correctness was found for the diagnoses by VC images (86,67% by fiberscopy and by telescopy). Consequently, the use of a VC to transmit videolaryngoscopy images has proved to be efficient for the remote diagnoses of larynx diseases

Objetivos: Averiguar o papel da duração da intubação e de outros fatores de risco no desenvolvimento de lesão de laringe em crianças submetidas à intubação endotraqueal em unidade de terapia intensiva. Além disso, determinar a incidência de estenose subglótica (ESG) nessa mesma população. Delineamento: Coorte Prospectiva. Métodos: Foram elegíveis todas as crianças de zero a quatro anos internadas na Unidade de Terapia Intensiva Pediátrica do Hospital de Clínicas de Porto Alegre que necessitaram de intubação endotraqueal por mais de 24 horas. Foram excluídas as com história de intubação, doença laríngea prévia, presença de traqueostomia atual ou no passado, presença de malformações craniofaciais e aquelas consideradas portadoras de doença terminal pela equipe assistente. As crianças incluídas foram acompanhadas diariamente e, após a extubação, foram submetidas à fibronasolaringoscopia (FNL). Aquelas que tinham alterações moderadas a graves nesse primeiro exame foram submetidas à nova FNL em sete a dez dias. Se as alterações persistissem ou se desenvolvessem sintomas, independentemente do exame inicial, as crianças eram submetidas à laringoscopia sob anestesia geral. Resultados: Foram acompanhadas 142 crianças entre novembro de 2005 e outubro de 2010. Na FNL inicial, 58 crianças (40,8%) apresentaram alterações laríngeas moderadas a graves. Das 84 crianças sem lesões no exame inicial, uma desenvolveu sintomas laríngeos e foi submetida à laringoscopia direta, sendo diagnosticada ESG. Na FNL de revisão do grupo com alterações moderadas a graves, 15 crianças apresentaram ESG no seguimento, diagnosticada após exame sob anestesia geral. Ao final do estudo, a incidência de ESG foi de 11,3% (IC 95%: 7,1 – 17,5%). Após análise multivariada dos fatores estudados, encontrou-se que, para cada cinco dias adicionais de intubação, há um acréscimo de 50,3% no risco de desenvolver ESG e, para cada dose extra de sedação/dia, um incremento de 12%. Conclusões: Nesse estudo prospectivo em crianças foi encontrada uma incidência de ESG superior à da maioria dos trabalhos anteriores. O tempo de intubação e a necessidade de doses extras de sedação parecem ser fatores importantes para o desenvolvimento de ESG durante a intubação endotraqueal. / Objectives: To investigate the role of the length of intubation and other risk factors in the development of laryngeal lesions in children undergoing endotracheal intubation in an intensive care unit, as well as to determine the incidence of subglottic stenosis (SGS) in the same population. Design: Prospective cohort. Methods: Children aged zero to four years admitted to the Pediatric Intensive Care Unit of Hospital de Clinicas de Porto Alegre who required endotracheal intubation for more than 24 hours were eligible for study. Patients with previous intubation, history of laryngeal disease, current or past tracheostomy, presence of craniofacial malformations and those considered terminal by the staff were excluded from the study. Children were monitored daily and underwent flexible fiberoptic laryngoscopy (FFL) after extubation. Those who presented moderate to severe abnormalities in this first examination underwent another FFL between 7 and 10 days later. If these lesions persist or symptoms developed, regardless of the findings in initial examination, laryngoscopy under general anesthesia was warranted. Results: We followed 142 children between November 2005 and October 2010. In the first FFL 58 children (40.8%) had moderate to severe laryngeal lesions. Of the 84 children with a normal initial examination, only one developed laryngeal symptoms and underwent direct laryngoscopy, which revealed eventually SGS in the follow up. In the group with moderate to severe lesions in the first examination, 15 children presented SGS afterwards (after examination under anesthesia). At the end of the study, the incidence of SGS was 11.3% (95% CI: 7.1 to 17.5%). Multivariate analysis showed that for every five additional days of intubation there is a 50.3% increase in the risk of developing SGS, and that for each additional sedation doses/ day, there is a 12% increase in the same outcome. Conclusions: In this prospective study in children, we found a higher incidence of SGS than most previous studies. The length of intubation and the need for additional sedation doses appear to be key factors for the development of SGS during endotracheal intubation.