Tracheostomy Placement in Children Younger Than 2 Years: 30-Day Outcomes Using the National Surgical Quality Improvement Program Pediatric

Importance  Analysis of the American College of Surgeons National Surgical Quality Improvement Program (NSQIP) Pediatric demonstrated that the highest contribution to composite morbidity in otolaryngology is seen in children younger than 2 years undergoing tracheostomy.

Objective  To determine predictive factors for complications following tracheostomy placement in patients younger than 2 years that, if targeted for reduction in quality initiatives, might result in improved surgical outcomes.

Design, Setting, and Participants  The NSQIP Pediatric reports predefined 30-day postoperative outcomes for surgical cases from participating institutions for quality improvement. All 206 cases of elective tracheostomy performed in children younger than 2 years from 2012 to 2013 among 61 participating institutions and documented in the NSQIP Pediatric public use file were included. Data analysis was conducted from December 1, 2014, to June 30, 2015.

Interventions  Surgical placement of tracheostomy in children younger than 2 years.

Main Outcomes and Measures  Demographics and clinical and perioperative characteristics for children who did and did not experience a major complication from elective tracheostomy were compared. Continuous variables were compared using Wilcoxon rank sum tests and categorical variables were compared using Pearson χ² tests or Fisher exact tests where appropriate. A logistic regression model was fit to identify adjusted odds ratios (aORs) with 95% CIs of preoperative factors predictive of occurrence of a major complication.

Results  Of the 206 children younger than 2 years who underwent tracheostomy, 50 (24.3%) experienced a major complication within 30 days. The most common complications were pneumonia (16 [7.8%]), postoperative sepsis (12 [5.8%]), death (12 [5.8%]), and deep or organ space surgical site infections (8 [3.9%]). Neonatal age (aOR, 2.38; 95% CI, 1.06-5.37; P = .04), intraventricular hemorrhage (aOR, 2.72; 95% CI, 1.01-7.32; P = .048), and comorbid cardiac risk factors (relative to none: minor risk factors, aOR, 2.94; 95% CI, 1.19-7.24; major or severe risk factors, aOR, 1.31; 95% CI, 0.44-3.84; P = .04 for all cardiac risk factors) were independently predictive of major complications.

Conclusions and Relevance  Young children undergoing tracheostomy tube placement have high rates of morbidity. This analysis identifies the need for additional procedure-specific outcome variables and improved variable definitions to incorporate into a detailed module for NSQIP Pediatric that will more effectively promote national, specialty-specific targeted quality improvement efforts.

Pediatric tracheostomy is known to have a high rate of complications, reported at 36% to 60%.^1,2 The difference in rates of complications between children and adults is owing to variances in anatomy between pediatric and adult patients.³ Common postoperative complications found in children include hemorrhage, decannulation, localized infection, and localized pressure ulcers.^1,4,5 Multiple single-institution quality improvement initiatives and retrospective studies that focused on changes to surgical technique and postoperative tracheostomy care are described in the literature with mixed results. In addition, no standardization of protocols for tracheostomy placement and care exists between institutions.^4,6-9

The American College of Surgeons established the National Surgical Quality Improvement Program (NSQIP) in 2004 and the NSQIP Pediatric in 2008 to measure quality of care among surgical patients.^10,11 With the passage of the 2008 Medicare Improvements for Patients and Providers Act and the 2010 Patient Protection and Affordable Care Act, which call for setting national priorities for improvement in quality and recommend use of performance measures in federal programs to achieve these priorities,¹² the NSQIP Pediatric provides an opportunity to apply and measure the effect of multi-institutional quality improvement initiatives in improving surgical outcomes for children. Previous analysis of the NSQIP Pediatric data has demonstrated an overall low rate of major postoperative morbidity among collected pediatric otolaryngology surgical cases.¹³ However a complication rate of 23% was noted for children younger than 2 years undergoing tracheostomy placement, which had the highest contribution to composite morbidity of all procedures analyzed.

With more than 4800 pediatric tracheostomies performed in the United States annually, 33% of which are performed on infants,³ the American Society for Pediatric Otolaryngology NSQIP Pediatric task force convened an initiative to identify metrics for measuring procedure-specific quality of care and to set goals for reducing poor outcomes among children younger than 2 years undergoing tracheostomy placement. This task force will recommend a national quality improvement, tracheostomy-specific module for inclusion in the 2016 NSQIP Pediatric platform, improving outcomes for these children. To develop this module, it is important to understand what factors might be associated with poor outcomes. The purpose of this study was to determine predictive factors for complications following tracheostomy placement in children younger than 2 years that may be used to accurately measure the effect of multi-institutional quality initiatives.

The NSQIP Pediatric is a multispecialty program managed by the American College of Surgeons that reports peer-reviewed, risk-adjusted 30-day postoperative outcomes for surgical procedures performed on patients younger than 18 years from participating institutions for benchmarking and quality improvement. A dedicated surgical clinical reviewer at each participating institution collects data on patients selected by the NSQIP 8-day cycle-based systematic sampling of 35 procedures per cycle. At least 124 variables are collected from the medical records and the patients and their families, including information on demographics, preoperative and intraoperative risk factors, and postoperative outcome and discharge variables.^10,14-16 This study was approved by the Nationwide Children’s Hospital Institutional Review Board. All data were deidentified prior to analysis.

All instances of tracheostomy performed as the primary procedure on children younger than 2 years in the 2012-2013 NSQIP Pediatric public use file were identified using Current Procedural Terminology codes 31600 and 31601 and age at surgery. Children who underwent an emergency or urgent operation were excluded. We compared demographics, preoperative risk factors, and 30-day outcome characteristics between children who did and did not experience a major complication within 30 days of surgery. Neonates were defined as all children who underwent surgery at or before 44 weeks’ gestational age. Premature infants were defined as all children born at or before 37 weeks’ gestational age. Cardiac risk factors were classified as minor, major, and severe based on International Classification of Disease, Ninth Revision, Clinical Modifications, diagnostic codes. Pulmonary risk factors were defined as a diagnosis of asthma, obstructive sleep apnea, cystic fibrosis, preoperative ventilator dependence, preoperative oxygen requirement, or a preoperative tracheostomy. Anesthesia time was defined as total time during which the child was under the direct care of an anesthesiologist for the surgical procedure. Corticosteroid use was defined as use of oral or parenteral corticosteroid medication within 30 days before surgery. Major complications were defined as deep or organ space surgical site infection (SSI), wound dehiscence, reintubation, pulmonary embolism, pneumonia, progressive renal insufficiency or acute renal failure, seizure, nerve injury, cerebrovascular accident, coma lasting more than 24 hours, cardiac arrest, venous thromboembolism, postoperative sepsis, unplanned related reoperation, or death. Minor complications were defined as superficial SSI or urinary tract infection within 30 days after surgery.

Continuous variables were described using medians and interquartile ranges and frequencies and percentages were used to describe categorical variables. To compare children with and without a major complication, Wilcoxon rank sum tests were used for continuous variables, and Pearson χ² tests or Fisher exact tests, where appropriate, were used for categorical tests. P < .05 was considered statistically significant. We fit a multivariable logistic regression model to identify independent risk factors for occurrence of a major complication. Adjusted odds ratios and 95% CIs are reported for the included factors in the logistic regression model. All risk factors, including patient demographics, surgery details, and preoperative comorbidities, for which patients differed with P < .20 after univariable analysis were considered for inclusion in the multivariable model, which was fit using backward stepwise regression. Model fit was assessed using the Hosmer-Lemeshow goodness-of-fit test and model discrimination was checked using the C statistic. Acceptable discrimination is commonly reported to be 0.7 or greater.¹⁷ Owing to the sparseness and quasicomplete separation of the data for some variables, the Firth penalized likelihood bias-reduction method was used to determine the risk factors for a major complication.¹⁸ However, these variables eventually fell out of the model and the Firth penalized likelihood bias-reduction method was not used for the final model. All analyses were performed using SAS, version 9.3 (SAS Institute Inc). Data analysis was conducted from December 1, 2014, to June 30, 2015.

Table 1 presents demographics and preoperative risk factors for children younger than 2 years undergoing tracheostomy placement. Of 206 patients, 70 (34.0%) were neonates, 112 (54.4%) were born prematurely, 201 (97.6%) were categorized as American Society of Anesthesiologists class 3 or higher, 156 (75.7%) required nutritional support, and 50 (24.3%) experienced a major complication within 30 days of surgery. Children who experienced a major complication were of similar age as those who did not sustain a major complication; both groups also had similar proportions of patients who presented with failure to thrive or required supplemental nutrition. However, children who experienced a major complication were more likely to have a history of immune system disease or use of immunosuppressants (3 [6.0%] vs 0; P = .01).

Postoperative outcomes are presented in Table 2. Median postoperative length of stay was 28 days (interquartile range, 16-61 days) for all patients. There were no differences in postoperative length of stay or need for intraoperative or postoperative transfusions within 72 hours of surgery between children with or without a major complication. The most common major complications were pneumonia (16 patients [7.8%]), postoperative sepsis (12 [5.8%]), death (12 [5.8%]), and deep or organ space SSI (8 [3.9%]).

Table 3 presents a predictive multivariable logistic regression model of major complications for children younger than 2 years undergoing tracheostomy placement. Neonatal age (adjusted odds ratio, 2.38; 95% CI, 1.06-5.37; P = .04), and preoperative history of intraventricular hemorrhage (adjusted odds ratio, 2.72; 95% CI, 1.01-7.32; P = .048) were independent risk factors for major complications. Presence of a cardiac risk factor was also predictive of major complications. The adjusted odds of a major complication among those with major or severe cardiac risk factors were 1.31 times (95% CI, 0.44-3.84; P = .63) greater relative to those with no cardiac risk; the adjusted odds among those with minor cardiac risk factors were 2.94 times (95% CI, 1.19-7.24; P = .02) greater. The model demonstrated good fit but did not achieve acceptable discrimination (area under the curve = 0.679), so there are other preoperative risk factors that need to be included in the prediction model that were not captured in the NSQIP Pediatric.

Of the children younger than 2 years who underwent tracheostomy placement, 24.3% experienced a major complication within 30 days of surgery. Neonates, children with preoperative intraventricular hemorrhage, and children with cardiac risk factors were more likely to experience a major complication following tracheostomy placement. The risk factors identified in this study are different from those identified in studies in children undergoing other surgical procedures and in adults undergoing otolaryngology procedures. In a study by Raval et al,¹⁶ American Society of Anesthesiologists class 3 or higher and the need for nutritional support were predictors of postoperative complications in all children undergoing surgery, but neither of these factors was identified as a predictor of postoperative complications in our study. In addition, a previous analysis of adults undergoing otolaryngology procedures indicated that American Society of Anesthesiologists class 3 or higher, prolonged operative time, presence of cerebrovascular disease, and preoperative corticosteroid use were associated with postoperative complications.¹⁹ Our study does not indicate that these preoperative risk factors are associated with postoperative complications for children younger than 2 years undergoing tracheostomy. We cannot assume that predictors of complications in adults undergoing otolaryngology procedures or in children undergoing other surgical procedures will accurately describe expected complication rates in the current study population.

The multivariable model prepared for this study demonstrated good fit but not acceptable discrimination. The reduced discrimination of the logistic regression model in the setting of good fit indicates that variables not included among the NSQIP’s measured variables may better discriminate among the patients who experienced a postoperative complication. Previous analysis of the NSQIP has demonstrated that, for adult patients, hospital quality measures can be adequately risk adjusted with a limited number of variables.²⁰ To limit the burden of data collection, the NSQIP has reduced the variables needed to those necessary for risk adjustment. Owing to all of the reasons mentioned, it is essential to include some procedure-specific risk factors in our upcoming NSQIP Pediatric module. Other procedure-specific quality improvement initiatives, such as those targeting spinal fusion, pancreatectomy, and vascular surgery, have demonstrated the importance of defining risk factors and outcomes.^21-24 Risk factors to be included in the procedure-specific module will likely include the primary indication for a tracheostomy, type of tracheostomy collar, stomal and/or retention suture placement, and type of protective dressing.

The NSQIP Pediatric was originally established to report surgical outcomes for all specialties; unfortunately, it currently does not provide a comprehensive list of procedure-specific outcomes of interest to otolaryngologists. For example, tracheostomy-related pressure ulcers occur in 8% to 29% of pediatric patients,^4,7 but this complication is not documented in the NSQIP Pediatric. Therefore, our study may be underestimating the true number of complications seen in young children undergoing a tracheostomy. Our initiative will examine additional procedure-specific outcomes relevant to children undergoing tracheostomy so that we may be able to target complications of interest to otolaryngologists for quality improvement and identify areas for other cross-institutional quality improvement initiatives in pediatric otolaryngology. In addition to tracheostomy-specific pressure ulcers, other relevant outcomes that should be included are accidental decannulation, tracheostomy tube occlusion, pneumothorax, breakdown of the skin of the neck, or tracheal wall laceration.^1,5,25,26 Wound dehiscence and superficial SSI are clearly defined by the NSQIP Pediatric to be separate definitions and do not include pressure ulceration, thus adding to the necessity for inclusion in future research.

We note several limitations of this study. First, the NSQIP administrators attempt to validate the data collected, use clear variable definitions, and train clinical case reviewers, but those entering the data may not be trained in the nuances of otolaryngology procedures. An otolaryngologist may also define a complication differently than does the NSQIP. For example, physiologic tracheal secretions following tracheostomy placement are not uncommon, and an otolaryngologist would recognize these secretions as physiologic, but a data collector might not recognize this finding and might document it as consistent with SSI or pneumonia. Second, compared with existing studies in adults, logistic regression modeling in our study is restricted by the small sample size and limited variables included in the NSQIP Pediatric. Further accrual of patients in the NSQIP Pediatric and participation by additional institutions will allow for future analysis with increased power.

Our study offers guidance on variables to include in the future NSQIP Pediatric specialty and procedure-specific module for tracheostomy in children younger than 2 years. Inclusion of cardiac risk factors, intraventricular hemorrhage, and preoperative nutritional support is important for characterizing preoperative factors that might be drivers for the high rate of complications in this patient population and determining whether differences in complication rates are owing to these preoperative factors or to the effectiveness of potential quality improvement initiatives. At the same time, consistent with other quality improvement initiatives developed following initial analysis of NSQIP data,^21-24 otolaryngology-specific preoperative risk factors and postoperative sequelae will be included to identify meaningful postoperative outcomes for actionable quality improvement.

In this study, 24.3% of children younger than 2 years undergoing tracheostomy tube placement experienced a major complication. Neonatal age, cardiac risk factors, and intraventricular hemorrhage were independently associated with a higher likelihood of experiencing a major complication; these preoperative risk factors will be incorporated in an upcoming otolaryngology NSQIP Pediatric quality improvement initiative that will also include tracheostomy-specific risk factors and outcomes.

Submitted for Publication: August 21, 2015; final revision received October 9, 2015; October 29, 2015.

Corresponding Author: Kris R. Jatana, MD, Department of Otolaryngology–Head and Neck Surgery, Nationwide Children’s Hospital, 555 S 18th St, Ste 2A, Columbus, OH 43205 (kris.jatana@nationwidechildrens.org).

Published Online: January 28, 2016. doi:10.1001/jamaoto.2015.3302.

Author Contributions: Ms Asti and Dr Jatana had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Mahida, Asti, Boss, Shah, Jatana.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Mahida, Asti, Jatana.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Mahida, Asti.

Administrative, technical, or material support: Asti, Boss, Shah, Jatana.

Study supervision: Boss, Shah, Deans, Minneci, Jatana.

Conflict of Interest Disclosures: None reported.

Funding/Support: This project is supported by intramural funding from the Department of Otolaryngology–Head and Neck Surgery and the Center for Surgical Outcomes Research at Nationwide Children’s Hospital and by grant 1K08 HS22932-01 from the Agency for Healthcare Research and Quality , the Johns Hopkins Clinician Scientist Award, and the American Society of Pediatric Otolaryngology Career Development Award (Dr Boss).

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The American College of Surgeons National Surgical Quality Improvement Program and the hospitals participating in the program are the sources of the data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.

Previous Presentations: This study was presented at the American Society of Pediatric Otolaryngology Annual Meeting; April 24, 2015; Boston, Massachusetts; and the American College of Surgeons National Surgical Quality Improvement Program Annual Meeting; July 26, 2015; Chicago, Illinois.