Risk Factors for Intubation-Related Laryngeal Granulomas: A Prospective Study in Surgical Patients Authors: Takashi Wada, Machiko Kimura, Yasuhiro Arai, Nobuhiko Oridate Published: 2026 ABSTRACT AbstractIntroductionThis study aimed to determine the frequency of intubation-related laryngeal granulomas (ILGs) after surgery under general anesthesia and identify risk factors, with a focus on tube type and duration.MethodsA single-center prospective cohort (November 2012–September 2013) of adults undergoing elective surgery expected to last at least 5 h with intubation was conducted. Fiberoptic laryngoscopy was performed on the day after extubation and at 1 week, 1 month, and 3 months. Candidate variables included tube type (single lumen vs. double lumen [DLT]), intubation duration, Cormack-Lehane grade, intraoperative transesophageal echocardiography (TEE), gastroesophageal reflux disease, body mass index, sex, and age. Associations were evaluated using univariate tests and multivariable logistic regression.ResultsForty patients were analyzed (median age, 73 years; 19 men). Cardiovascular surgery accounted for 25 of 40 cases (62.5%); DLTs were used in 6 of 40 (15.0%) and TEE in 22 of 40 (55.0%). Mucosal changes (erythema/white plaque) were observed on day 1 in 18 of 40 patients (45.0%). ILG was detected at 1 week in 10 of 40 patients (25.0%), including nine after cardiovascular surgery. Eight of 10 cases (80%) resolved spontaneously by 1 month, whereas 2 of 10 (20%) enlarged and required 2 months of inhaled steroid therapy, with complete resolution by 3 months. Univariate analysis revealed that both intubation duration ≥13 h (p = 0.014) and DLT use (p = 0.026) were significantly associated with ILG. In multivariable analysis, intubation duration ≥13 h (adjusted odds ratio [aOR]: 10.697, 95% confidence interval [CI]: 1.639–69.828; p = 0.013) and DLT use (aOR: 14.521, 95% CI: 1.531–137.680; p = 0.020) were independent risk factors.ConclusionIn this cohort, ILG was detected in 25% of patients at 1 week post-extubation and typically resolved within 1 month. Prolonged intubation and DLT use were independently associated with increased ILG risk. High-risk patients – particularly those with prolonged intubation or DLT exposure – may benefit from selection of the smallest appropriate DLT when required and targeted laryngoscopy within 1 month. These findings should be interpreted cautiously, given the small sample and single-center design. Introduction Laryngeal granulomas may result from mechanical and chemical irritation. In surgical practice, intubation-related granulomas (ILGs) are of particular concern because, although outcomes are generally favorable, these lesions can sometimes enlarge and compromise the airway [1–5]. Accordingly, ILGs are relevant not only to anesthesiologists involved in peri-extubation management but also to otolaryngologists responsible for postoperative voice and airway outcomes. Most available data on post-intubation laryngeal injury derive from intensive care unit (ICU) populations exposed to prolonged ventilation, in whom factors such as intubation duration and nasogastric tube placement have been implicated; however, the role of tube-related factors remains inconsistent across studies [6–9]. A recent prospective study by Plowman et al. [10] evaluated postoperative cardiovascular surgery patients using fiberoptic laryngoscopy within 72 h of extubation and focused primarily on vocal fold mobility impairment (VFMI); granuloma was reported as one component of mucosal injury rather than as the primary endpoint. However, the incidence, perioperative factors associated with ILGs, and short-term clinical course of ILGs in elective surgical patients undergoing planned endotracheal intubation remain insufficiently defined. Accordingly, this prospective study aimed to determine the incidence of ILGs in adult patients undergoing elective surgery under general anesthesia and to identify factors independently associated with ILG development, with particular focus on endotracheal tube type (double lumen tube [DLT] vs. single lumen tube [SLT]) and intubation duration. In addition, this study explored the potential association between intraoperative transesophageal echocardiography (TEE) and ILG occurrence. We hypothesized that DLT use and longer intubation duration would be independently associated with ILG development. Methods Study Design and Participants We conducted a single-center prospective observational study at Yokosuka Uwamachi Hospital between November 2012 and September 2013. Adults scheduled for elective surgery expected to last at least 5 h under general anesthesia with endotracheal intubation were screened. The 5-h threshold was prespecified to enrich the cohort for patients at higher risk of post-intubation laryngeal injury, given the longer ventilation time and the higher likelihood of additional intraoperative airway instrumentation (e.g., TEE probe insertion and manipulation in cardiothoracic cases), and was informed by prior literature linking intubation duration to laryngeal injury in ICU cohorts [6–9]. Exclusion criteria included lack of consent, preoperative laryngeal granuloma or tumor, and pre-existing vocal fold paralysis. Perioperative airway management and the postoperative laryngoscopy schedule followed institutional protocols in routine use during the study period. Data Collection and Variables Baseline demographics (age and sex), body mass index (BMI), gastroesophageal reflux disease (GERD) diagnosis and/or reflux symptoms from the preoperative interview, and the surgical department were obtained from medical records. Peri-intubation variables included endotracheal tube type (DLT vs. SLT), intubation duration, Cormack-Lehane laryngoscopy grade [11], intraoperative TEE use, and postoperative cough or hoarseness (assessed by patient questionnaire). Intubations were performed by attending anesthesiologists or supervised trainees following departmental airway standards (e.g., routine cuff pressure manometry and avoidance of unnecessary tube exchanges), thereby reducing operator-based variability. External tube diameter was documented from anesthesia charts but was not modeled as an independent exposure in the primary analyses because of collinearity with tube type (online suppl. Table S1; for all online suppl. material, see https://doi.org/10.1159/000553227). Postoperative Assessment Standardized flexible fiberoptic laryngoscopy was performed by an otolaryngologist on the day after extubation and again at 1 week, 1 month, and 3 months post-extubation. These intervals were selected to capture (1) early mucosal injury (day 1), (2) the typical timeframe for granuloma development within the first weeks, and (3) resolution or persistence over 1–3 months, consistent with prior reports of post-intubation laryngeal lesions [8, 9, 12]. Examinations were conducted by board-certified otolaryngologists using a standardized technique and a predefined case report form, with still image documentation to support consistent outcome classification across multiple visits. Definitions The primary outcome was the presence of ILG at 1 week post-extubation. ILG was defined as a newly developed nodular or polypoid lesion arising from the vocal process or posterior glottis on flexible laryngoscopy following recent endotracheal intubation. This definition aligns with established otolaryngology criteria for posterior glottic/contact lesions [9, 13, 14]. Secondary observations included erythema and white plaque at the posterior glottis. Obesity was defined as BMI ≥25 kg/m2. Criteria for Tube Selection (Clinical Practice) SLT sizes were selected by the attending anesthesiologist according to institutional practice, aiming for the smallest outer diameter that provided adequate ventilation and acceptable airway pressure, while accounting for sex, height, and airway anatomy. DLTs were used when required for thoracic or cardiac surgery (e.g., lung isolation), with size determined by patient habitus and radiographic airway estimates. High-volume, low-pressure cuffs were used in all cases. Departmental guidance encourages selection of the smallest feasible external diameter consistent with ventilation goals and routine cuff pressure monitoring with manometry, whereas the device of choice (DLT vs. SLT) remains indication-driven and was not protocolized by the study team. Statistical Analysis Patients were categorized according to ILG presence at 1 week. Categorical variables were compared using the chi-square or Fisher’s exact test, and continuous variables were analyzed using t tests or Wilcoxon rank-sum tests, as appropriate. For multivariable analysis, three prespecified models were constructed while accounting for the limited number of events:Primary (parsimonious) model: intubation duration ≥13 h and tube type (DLT vs. SLT), selected based on clinical plausibility, prior evidence linking longer intubation to laryngeal injury [6–9], and the need for parsimony given the limited number of events.Sensitivity A: primary model plus age, sex, BMI, and GERD (clinically relevant covariates).Sensitivity B (penalized): Firth logistic regression including age, sex, BMI, GERD, TEE use, Cormack-Lehane grade II–IV, and surgical category (cardiovascular vs. other) to evaluate robustness under small event conditions. Primary (parsimonious) model: intubation duration ≥13 h and tube type (DLT vs. SLT), selected based on clinical plausibility, prior evidence linking longer intubation to laryngeal injury [6–9], and the need for parsimony given the limited number of events. Primary (parsimonious) model: intubation duration ≥13 h and tube type (DLT vs. SLT), selected based on clinical plausibility, prior evidence linking longer intubation to laryngeal injury [6–9], and the need for parsimony given the limited number of events. Sensitivity A: primary model plus age, sex, BMI, and GERD (clinically relevant covariates). Sensitivity A: primary model plus age, sex, BMI, and GERD (clinically relevant covariates). Sensitivity B (penalized): Firth logistic regression including age, sex, BMI, GERD, TEE use, Cormack-Lehane grade II–IV, and surgical category (cardiovascular vs. other) to evaluate robustness under small event conditions. Sensitivity B (penalized): Firth logistic regression including age, sex, BMI, GERD, TEE use, Cormack-Lehane grade II–IV, and surgical category (cardiovascular vs. other) to evaluate robustness under small event conditions. Two-sided p values <0.05 were considered statistically significant. Given the limited number of events, results from sensitivity analyses were interpreted cautiously as supportive rather than definitive. Rationale for the 13-Hour Cutoff The 13-h threshold was prespecified to represent prolonged intubation in this surgical cohort and was informed by local case mix and prior ICU literature linking longer intubation to laryngeal injury [6–9]. Intubation duration was also analyzed as a continuous variable and by quartiles in sensitivity analyses (online supplement). Ethics The protocol was approved by the Institutional Review Board of Yokosuka Uwamachi Hospital (approval no: 20120006) and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants before enrollment. Results Patient Flow From November 2012 to September 2013, 48 patients were screened; 40 consented and were included in the analysis. Eight patients were excluded: six declined consent, and two had incomplete baseline data. Patient Characteristics and Subgroup Comparisons Table 1 presents the overall and subgroup characteristics (ILG vs. no ILG at 1 week). The median age was 73 years (range, 44–87), and 19 of 40 patients (47.5%) were men. Cardiovascular surgery accounted for 25 of 40 cases (62.5%). DLTs were used in 6 of 40 cases (15.0%), and intraoperative TEE was used in 22 of 40 cases (55.0%). Table 1.Baseline characteristics overall and by ILG status at 1 weekVariableOverallILGNo ILG p valueAge70.4±11.867.3±14.771.4±10.80.425Sex, n (%) Male19 (47.5)4150.721 Female21 (52.5)6150.721BMI23.9±4.222.3±4.424.4±4.10.134 BMI <25 kg/m2, n (%)27 (67.5)8190.451 BMI ≥25 kg/m2, n (%)13 (32.5)2110.451Intubation duration29.0±66.435.4±40.426.9±73.50.024* <13 h, n (%)29 (72.5)4250.014* ≥13 h, n (%)11 (27.5)650.014*GERD history, n (%) Yes5 (12.5)230.584 No35 (87.5)8270.584Cormack-Lehane grade, n (%) I31 (77.5)7240.665 II–IV9 (22.5)360.665Tube type, n (%) SLT34 (85.0)6280.026* DLT6 (15.0)420.026*Transesophageal echocardiography, n (%) Used22 (55.0)8140.082 Not used18 (45.0)2160.082Surgery type, n (%) Cardiovascular surgery25 (62.5)9160.060 Other15 (37.5)1140.060Postoperative throat symptoms, n (%) Present19 (47.5)7120.148 None21 (52.5)3180.148BMI, body mass index; ILG, intubation-related laryngeal granuloma; GERD, gastroesophageal reflux disease; SLT, single lumen tube; DLT, double lumen tube. *p < 0.05. Baseline characteristics overall and by ILG status at 1 week BMI, body mass index; ILG, intubation-related laryngeal granuloma; GERD, gastroesophageal reflux disease; SLT, single lumen tube; DLT, double lumen tube. *p < 0.05. Postoperative Laryngeal Findings and Outcomes Mucosal changes (erythema or white plaque) were observed on day 1 in 18 of 40 patients (45.0%). ILG was present at 1 week in 10 of 40 patients (25.0%) – nine after cardiovascular surgery and one after esophagectomy. As shown in Table 2, 8 of 10 ILG cases (80%) resolved spontaneously by 1 month, whereas 2 of 10 (20%) enlarged and required inhaled steroid therapy; complete resolution was achieved in all cases by 3 months. Despite conservative care before treatment, the two treated patients presented with broad-based posterior glottic granulomas and persistent dysphonia and throat clearing. Both patients were administered inhaled steroids and showed visible shrinkage by 4–6 weeks, achieving complete endoscopic and symptomatic resolution by 3 months without reintubation or surgical excision. These cases illustrate the clinical spectrum and highlight that a minority of ILGs (2/40, 5% of the cohort; 2/10, 20% of ILGs) can progress and require specific therapeutic intervention, despite a generally self-limited course. Table 2.Post-extubation laryngeal findings and clinical outcomes by ILG statusOutcomeILG, n/N (%)No ILG, n/N (%)Day 1 mucosal change (any)9/10 (90.0)9/30 (30.0)Spontaneous resolution by 1 month8/10 (80.0)​Required inhaled steroid therapy2/10 (20.0)​Complete resolution by 3 months10/10 (100.0)​ILG, intubation-related laryngeal granuloma. Post-extubation laryngeal findings and clinical outcomes by ILG status ILG, intubation-related laryngeal granuloma. Univariate Associations Univariate analysis (rightmost column, Table 1) revealed that both intubation duration ≥13 h (p = 0.014) and DLT use (p = 0.026) were significantly associated with ILG. TEE use, age, sex, BMI, GERD, Cormack-Lehane grade, and postoperative cough or hoarseness were not associated with ILG. Multivariable Models In the primary model, which included intubation duration ≥13 h and tube type (DLT vs. SLT; Table 3), both variables were independent risk factors:Intubation duration ≥13 h: adjusted odds ratio: 10.697, 95% confidence interval: 1.639–69.828, p = 0.013DLT vs. SLT: adjusted odds ratio: 14.521, 95% confidence interval: 1.531–137.680, p = 0.020 Intubation duration ≥13 h: adjusted odds ratio: 10.697, 95% confidence interval: 1.639–69.828, p = 0.013 Intubation duration ≥13 h: adjusted odds ratio: 10.697, 95% confidence interval: 1.639–69.828, p = 0.013 DLT vs. SLT: adjusted odds ratio: 14.521, 95% confidence interval: 1.531–137.680, p = 0.020 DLT vs. SLT: adjusted odds ratio: 14.521, 95% confidence interval: 1.531–137.680, p = 0.020 Table 3.Logistic regression for ILG at 1 week (primary and sensitivity models)PredictorModelaOR (95% CI) p valueModel NIntubation duration ≥13 hPrimary10.697 (1.639–69.828)0.013*40DLT vs. SLTPrimary14.521 (1.531–137.681)0.020*40Intubation duration ≥13 hSensitivity A41.128 (1.710–989.394)0.022*40DLT vs. SLTSensitivity A61.659 (1.139–3,339.259)0.043*40Age, yearsSensitivity A0.938 (0.807–1.089)0.40140Male sexSensitivity A0.127 (0.008–1.950)0.13940BMI ≥25 kg/m2Sensitivity A0.056 (0.001–3.086)0.15940GERDSensitivity A55.684 (1.004–3,088.255)0.05040Complete case analysis. Odds ratios derived from standard logistic regression. For auxiliary models with convergence issues owing to small event counts or separation, ridge regularization was applied (odds ratio reported without CI).aOR, adjusted odds ratio; CI, confidence interval; BMI, body mass index; ILG, intubation-related laryngeal granuloma; GERD, gastroesophageal reflux disease; SLT, single lumen tube; DLT, double lumen tube. *p < 0.05. Logistic regression for ILG at 1 week (primary and sensitivity models) Complete case analysis. Odds ratios derived from standard logistic regression. For auxiliary models with convergence issues owing to small event counts or separation, ridge regularization was applied (odds ratio reported without CI). aOR, adjusted odds ratio; CI, confidence interval; BMI, body mass index; ILG, intubation-related laryngeal granuloma; GERD, gastroesophageal reflux disease; SLT, single lumen tube; DLT, double lumen tube. *p < 0.05. In sensitivity A (further adjusted for age, sex, BMI, and GERD), associations for intubation duration and DLT use remained directionally consistent, although effect estimates were attenuated and confidence intervals widened (Table 3). In sensitivity B (Firth penalized regression; online suppl. Table S2), the results were directionally similar but imprecise because of the small number of events. Small Sample Caveat Because only 10 ILG events were observed, multivariable estimates – particularly in sensitivity models with additional covariates – should be interpreted with caution owing to potential instability and overfitting. This limitation is consistent with prior small sample reports in the post-intubation injury literature [6–9, 12]. Discussion This prospective study adds longitudinal data on ILG in elective surgical patients undergoing serial postoperative laryngoscopic follow-up. A recent prospective cardiovascular surgery study by Plowman et al. [10] evaluated 185 patients with fiberoptic laryngoscopy within 72 h of extubation and focused primarily on VFMI; in that study, granuloma was reported as one component of mucosal injury rather than as the primary endpoint. By contrast, our study prespecified ILG at 1 week as the primary outcome and followed patients through 3 months, allowing assessment of both early ILG development and short-term clinical course. In this cohort, intubation duration ≥13 h and DLT use were independently associated with ILG at 1 week. These findings should be interpreted as complementary to, rather than duplicative of, prior work. Whereas Plowman et al. [10] focused on VFMI in cardiovascular surgery during the immediate postoperative period, prior ICU-based studies have mainly examined laryngeal injury after prolonged ventilation in more heterogeneous critically ill populations [6–9]. The present data suggest that even in planned perioperative settings, tube type and intubation duration may influence the development of ILG detectable at 1 week. This distinction is clinically relevant because the timing, endoscopic appearance, and short-term course of ILG differ from those of VFMI-centered postoperative laryngeal injury. From a perioperative perspective, our results suggest several potentially modifiable practices. Avoiding unnecessary DLT use, selecting the smallest appropriate DLT when lung isolation is required, and minimizing intubation time may help reduce risk. In addition, targeted post-extubation laryngoscopy within 1 month may be reasonable for predefined high-risk patients, particularly those with prolonged intubation or DLT exposure. These implications should be interpreted cautiously given the exploratory nature of the study, but they provide a pragmatic framework for perioperative risk stratification. Most ILGs followed a favorable short-term course: 8 of 10 resolved by 1 month, whereas 2 enlarged and required inhaled steroid therapy, with complete resolution by 3 months. This pattern is broadly consistent with prior reports describing a largely self-limited course with occasional progression requiring intervention [8, 9, 12–17]. Given the rare but clinically important possibility of airway-compromising granuloma [4, 5], a pragmatic follow-up strategy may include symptom screening at discharge and targeted endoscopic assessment within 1 month for high-risk patients, with prompt otolaryngology referral for persistent dysphonia, dyspnea, or throat clearing. An additional analytic point concerns tube type versus diameter. The inferential analyses addressed tube type (DLT vs. SLT), not external diameter; the latter is presented descriptively in online supplementary Table S1. Because DLT use is closely linked to both larger outer diameter and procedure-specific indication, separate modeling of diameter in this small cohort would likely introduce collinearity and unstable estimates. Therefore, the observed association should be interpreted as relating to tube type as analyzed, whereas diameter-specific effects should be clarified in larger studies. Several limitations should be noted. The cohort was small (n = 40) with few events (n = 10), limiting statistical power and increasing the risk of model instability despite the use of parsimonious and penalized analyses. The single-center design and selective enrollment of procedures expected to last at least 5 h may limit generalizability, and residual confounding, including confounding by indication between cardiovascular surgery and DLT use, cannot be excluded. In addition, scheduled endoscopic assessments may have missed transient lesions that arose and resolved between study time points. Accordingly, these findings should be regarded as exploratory and hypothesis-generating rather than definitive. Despite these limitations, this study provides prospective longitudinal data on ILG in elective surgical patients, including detection at 1 week and short-term course through 3 months. Future multicenter studies should validate these associations, prospectively capture tube size as a standardized metric, and refine risk-based surveillance pathways for postoperative laryngeal injury. Such work may help define practical perioperative strategies that integrate anesthesiology and otolaryngology perspectives to improve both airway safety and postoperative voice outcomes. Conclusion In this prospective cohort of elective surgical patients undergoing general anesthesia, prolonged intubation (≥13 h) and DLT use were independently associated with ILG at 1 week. Most ILGs resolved within 1 month, although a minority progressed and required treatment. These findings extend prior work on postoperative laryngeal injury by characterizing the early course of ILG in surgical patients and identifying potentially modifiable perioperative factors. Acknowledgments We would like to thank Editage (www.editage.jp) for English language editing. Statement of Ethics This study was conducted at Yokosuka Uwamachi Hospital and approved by the Institutional Review Board of the Hospital, approval no. 20120006. All procedures complied with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants before enrollment. Conflict of Interest Statement The authors have no conflicts of interest to declare. Funding Sources This study was not supported by any sponsor or funder. Author Contributions T.W. contributed to conceptualization, study design, investigation, data curation, formal analysis, and writing – original draft; M.K. contributed to conceptualization; Y.A. contributed to data curation; N.O. contributed to methodology, formal analysis, and supervision; All authors reviewed and approved the final manuscript.