Abstract
Objectives
Approximately 0.2% to 2.7% of children with congenital heart disease require a tracheostomy after cardiac surgery with the majority having single ventricle (SV) type heart lesions. Tracheostomy in SV patients is reported to be associated with high mortality. We hypothesized that short- and long-term survival of patients with SV heart disease would vary according to tracheostomy indication.
Methods
This is a single center, 20-year, retrospective review of all patients with SV heart disease who underwent tracheostomy. Demographic, cardiac anatomy, surgical, intensive care unit, and hospital course data were collected. The primary outcome was survival following tracheostomy. Secondary outcome was the completion of staged palliation to Fontan.
Results
In total, 25 patients with SV heart disease who underwent tracheostomy were included. Indications for tracheostomy included one or more of the following: tracheobronchomalacia (n = 8), vocal cord paralysis (n = 7), tracheal/subglottic stenosis (n = 6), primary respiratory insufficiency (n = 4), diaphragm paralysis (n = 3), suboptimal hemodynamics (n = 2), and other upper airway issues (n = 1). Survival at six months, one year, five years, and ten years was 76%, 68%, 63%, and 49%, respectively. Most patients completed Fontan palliation (64%). Patients who underwent tracheostomy for suboptimal hemodynamics and/or respiratory insufficiency had a higher mortality risk compared to those with indications of upper airway obstruction or diaphragm paralysis (hazard ratio 4.1, 95% confidence interval 1.2-13.7; P = .02).
Conclusions
Mortality risk varies according to tracheostomy indication in patients with SV heart disease. Tracheostomy may allow staged surgical palliation to proceed with acceptable risk if it was indicated for anatomic or functional airway dysfunction.
Introduction
Approximately 0.2% to 2.7% of children with congenital heart disease require a tracheostomy after congenital heart surgery.1–6 Many of these patients have single ventricle (SV) heart lesions. 2 Although rare, previously reported experiences with tracheostomy in SV heart disease have demonstrated poor outcomes with up to a 44% in-hospital mortality in patients with hypoplastic left heart syndrome (HLHS) 2 and overall poor long-term survival. Edwards et al reported a cohort of six subjects with SV heart lesions who underwent tracheostomy and among this group, three (50%) died and two were unable to complete Fontan palliation. 7 Similarly, Cotts et al reported 25 subjects with SV heart lesions who underwent tracheostomy and just nine (36%) of these patients survived during their seven-year study period. 4 However, despite the risks, in some cases tracheostomy may be the only option in providing a patient with a stable airway and plan for eventual hospital discharge. Although it has become an increasingly common procedure over the past two decades, 2 indications for tracheostomy in this patient population have not been well defined. We hypothesized that short- and long-term survival of patients with SV heart disease would vary according to tracheostomy indication. Specifically, we hypothesized that patients undergoing tracheostomy for known anatomic airway disease would have better survival than those who underwent tracheostomy for primary respiratory insufficiency or suboptimal hemodynamics.
Methods
Subjects
This is a retrospective cohort study of all patients with SV-type congenital heart disease who underwent tracheostomy at C.S. Mott Children's Hospital from January 2001 to December 2020. The study was approved by the Institutional Review board at the University of Michigan. Patients were identified from existing institutional databases and exhaustive medical record review. All patients undergoing tracheostomy at any point in their surgical palliative course were included in the cohort.
Data Collected
Demographic, cardiac anatomy, cardiac surgery, intensive care unit, and hospital course data were collected. These included age at tracheostomy, sex, race and ethnicity, type of SV congenital heart disease, presence of a chromosomal abnormality, presence of a congenital diaphragmatic or airway anomaly, presence of a congenital neurologic abnormality, surgical history and surgical palliative stage at tracheostomy, need for mechanical ventilation prior to the cardiac surgery associated with tracheostomy, surgical palliation associated hospitalization for tracheostomy, cardiopulmonary bypass time, cross-clamp time, and circulatory arrest time, use of delayed sternal closure, and postoperative complications including cardiac arrest, need for extracorporeal membrane oxygenation, diaphragm paralysis, vocal cord paralysis, chylothorax, acute kidney injury, venous thrombosis, neurologic injury, necrotizing enterocolitis, and bacterial infection requiring antibiotics. Duration of postoperative mechanical ventilation, the requirement for and duration of noninvasive positive pressure ventilation (NIPPV), and results of any bronchoscopy performed were also ascertained. The indication for tracheostomy was determined based on chart review and was categorized as one following: anatomic airway abnormalities including tracheal, subglottic stenosis, or “other” upper airway issues; functional airway abnormalities including tracheobronchomalacia and vocal cord paralysis; respiratory insufficiency including diaphragm paralysis or “other” primary respiratory insufficiency requiring mechanical ventilation; and suboptimal hemodynamics, which was defined as the inability to wean from positive pressure ventilation due to afterload reducing the effect of positive pressure ventilation and/or the inability to tolerate the metabolic demand of free breathing. Airway evaluation was only completed in those patients who were symptomatic (ie, respiratory distress and stridor) or were unable to wean from a ventilator or NIPPV. Evaluation generally consisted of an initial awake laryngoscopy, followed by a more detailed laryngoscopy and flexible bronchoscopy under general anesthesia before tracheostomy. For patients with diaphragm paralysis, the diagnosis was typically made with ultrasound or rarely with fluoroscopy.
Outcomes
The primary outcome was survival following tracheostomy. Secondary outcomes were the eventual completion of staged SV palliation to Fontan.
Analysis
The standard descriptive statistics were reported including frequency and percentage or median with interquartile range (IQR) as appropriate. Survival post-tracheostomy was estimated using Kaplan-Meier curve. Univariate Cox proportion hazard model was used to examine associations of patient demographics, indication for tracheostomy, and surgical characteristics with mortality. Comparisons were also made according to a stage of palliation and separated into post-stage I, stage II, and stage III. Univariate comparisons were also made in patient and clinical characteristics between patients who completed Fontan palliation and those who did not. Fisher exact test or χ2 test for categorical variables and Wilcoxon rank-sum test for continuous variables were used as appropriate. All analyses were performed using SAS version 9.4 (SAS Institute), with a statistical significance level of .05 using two-sided tests.
Results
Baseline Characteristics
In total, 27 patients with SV heart disease underwent tracheostomy during the study period. In all patients, tracheostomy occurred during a hospitalization for a surgical SV palliation. Two patients were excluded from the study due to tracheostomy being performed at an outside institution prior to surgery at our center and a lack of adequate documentation. Patient demographics and pre-, intra-, and postoperative characteristics are listed in Table 1. About half (52%) of the patients were male. Five patients (20%) were born prematurely (<37 weeks gestation) and four patients (16%) had a genetic abnormality including: Trisomy 21, Fryns syndrome, Turner syndrome, and a pericentric inversion of chromosome 9. Additionally, four patients (16%) were born with congenital neurologic abnormalities including: encephalocele, vermian hypoplasia, congenital hydrocephalus, and a neonatal stroke. The most common subtype of SV lesion was HLHS (64%) (Figure 1).
Single ventricle types. Abbreviations: AVSD, atrioventricular septal defect; HLHS, hypoplastic left heart syndrome; DILV, double inlet left ventricle; DORV, double outlet right ventricle
Demographics and Surgical Characteristics.
Significant p value indicated in bold.
Data are presented as N (%) for categorical variables and median (interquartile range) for continuous variables.
P value from χ2 test or Fisher exact test for categorical variables and Wilcoxon rank sum test for continuous variables.
Comparisons were made as HLHS versus all others.
Comparisons were made as Norwood versus all others.
Abbreviations: AKI, acute kidney injury; AVSD, atrioventricular septal defect; CPB, cardiopulmonary bypass; DILV, double inlet left ventricle; DORV, double outlet right ventricle; ECMO, extracorporeal membrane oxygenation; HLHS, hypoplastic left heart syndrome; N, nerve; NEC, necrotizing enterocolitis; Preop, preoperative; Rec, recurrent
The median duration of follow-up post tracheostomy was 5.2 years (IQR 0.8-9.6 years). The median age at tracheostomy was 2.9 months with a median weight of 3.8 kg. Tracheostomy occurred between stages I and II of SV palliations in seventeen (68%) of patients, between stages II and III in five patients (20%), and after stage III in three patients (12%). Thirteen of the patients who underwent tracheostomy between stages I and II were initially palliated with a Norwood procedure; however, it differed by heart disease in some and included a Starnes procedure in the single patient with Ebstein's anomaly, aortopulmonary shunts in two patients with DILV and tricuspid atresia, and a pulmonary artery band with arch reconstruction in the single patient with DORV. Stage II palliation consisted of a hemi-Fontan procedure in all patients and stage III was a lateral tunnel-fenestrated Fontan procedure in all patients. About two-thirds (60%) of the patients required preoperative mechanical ventilation prior to their cardiac surgery. The majority of patients failed extubation postoperatively with 15 patients failing extubation attempts at least once. Three patients were able to be extubated but failed to wean from NIPPV and progressed to a tracheostomy. The median number of days intubated was 16.5 (IQR 11.5-40 days). A majority of patients (68%) had an associated postoperative complication following SV palliation, with the most common complications being recurrent laryngeal nerve injury or chylothorax. Indications for tracheostomy included one or more of the following for each patient: tracheobronchomalacia (n = 4), vocal cord paralysis (n = 8), tracheal/subglottic stenosis (n = 7), primary respiratory insufficiency (n = 4), and suboptimal hemodynamics (n = 2). Of the two patients with suboptimal hemodynamics, one had severe atrioventricular valve insufficiency with associated moderately depressed ventricular systolic dysfunction, and the other required takedown of the superior cavopulmonary anastomosis after stage II due to hypoxemia secondary to inadequate pulmonary blood flow secondary to elevated pulmonary vascular resistance. All patients with diaphragm paralysis (five total) underwent diaphragm plication, but respiratory failure persisted in three patients, who had additional upper airway obstruction, and required tracheostomy. The median time from surgery to tracheostomy was 43 days (IQR 20-68 days).
Outcomes
A total of 14 patients survived post-tracheostomy throughout the entire follow-up period. Overall median survival post-tracheostomy was 9.0 years. Survival was 76%, 68%, 63%, and 49% at six months, one year, five years, and ten years since tracheostomy, respectively (Figure 2). Only two patients (8%) died prior to hospital discharge after tracheostomy. An additional nine patients (36%) died after hospital discharge. Among those who died, the median time from tracheostomy to death was 0.5 years (IQR 0.2-3.8). Among those who died, three were ventilator dependent and eight were tracheostomy-dependent without the need for a ventilator. Causes of death included heart failure (n = 4), sepsis (n = 3), sudden unexplained death (n = 2), postoperative bleeding (n = 1), and a multifactorial cause (n = 1). Four patients were evaluated for a heart transplant and two patients were ultimately listed for transplant with one ultimately receiving a transplant and one dying prior to transplant. Fourteen patients (56%) were sent home on mechanical ventilation. Among them, five (45.5%) are deceased and nine (64%) survived through our follow-up period. There was not a statistically significant association between home mechanical ventilation and survival (P = .43). Twelve (48%) of patients in our cohort were ultimately decannulated. Four (16%) of patients were ventilator-dependent through our follow-up period.
Overall survival (n = 25).
Most tracheostomy-specific complications were rare and included bleeding (n = 2) and pneumothorax (n = 1). No patients had pneumomediastinum, stomal ulceration, or tracheal stenosis associated with their tracheostomy. All patients required treatment for bacterial tracheitis at least once during our follow-up period. Importantly, despite the reported risk of infection associated with sternotomy with a tracheostomy, 8 no patients in our cohort developed mediastinitis after tracheostomy. One patient developed a superficial sternal wound infection. One patient developed endocarditis. And one patient developed sepsis while a tracheostomy was in place.
Sixteen patients (64%) completed Fontan palliation. There were no significant differences in demographics, tracheostomy indications, or surgical characteristics in patients who completed Fontan palliation and those who did not.
Demographics, surgical characteristics, and tracheostomy indications were compared among survivors and deceased patients (Tables 1 and 2). There were no significant mortality differences in demographics and surgical characteristics. Prematurity (P = .09) and age at tracheostomy (P = .06) trended toward increased risk of mortality but did not reach statistical significance. Timing of tracheostomy in reference to the stage of palliation did not confer a significant mortality risk. Postoperative complications following SV palliation were not significantly different between groups. There was a significantly increased risk of mortality among patients with delayed sternal closure (P = .03). Patients who underwent tracheostomy for primary respiratory insufficiency or suboptimal hemodynamics had an increased risk of mortality compared to those requiring tracheostomy for anatomic or functional airway abnormalities (hazard ratio 4.1, 95% confidence interval 1.2-13.7; P = .02). Mortality among groups was not significantly different for all other indications.
Survival by Tracheostomy Indication.
Data are presented as N (%).
P value from Fisher exact test.
Discussion
We found tracheostomy for specific indications in SV heart disease to be relatively safe with few tracheostomy-specific complications and survival similar to the general population of patients with SV heart disease. To our knowledge, this is the first study reporting outcomes specific to tracheostomy indication in SV heart disease.
Among our tracheostomy patient cohort, five-year survival was 63%. This is similar to survival reported in the “Single Ventricle Reconstruction (SVR) trial,” the largest prospective data set of SV patients to date, in which the authors death in 218 of 549 (39.7%) patients at the six-year follow-up mark.9,10 Our findings are starkly different from previously reported survival statistics in this population with up to 44% in-hospital mortality 2 and low post-hospital discharge survival.4,6 The reason for improved survival among our cohort is unclear, however, may be partly related to our center's increasing experience with a large volume of patients with SV heart disease, including those with tracheostomy, and more contemporary development of interstage monitoring programs and a robust home ventilator program. Interestingly, the type of SV heart lesion was not associated with outcomes after tracheostomy, which differs from published experiences reporting HLHS to be a particularly high-risk lesion with this intervention.2,4 Timing of tracheostomy in reference to the stage of palliation also did not confer a significant mortality risk.
In this study, we also examined detailed outcomes by tracheostomy indication. The majority of patients who underwent tracheostomy for diaphragm paralysis, vocal cord paralysis, subglottic stenosis, and/or airway malacia completed Fontan palliation and survived through our review period. These airway complications and abnormalities are commonly associated with the SV patient population.8,11–13 Diaphragm and vocal cord paralysis are particularly common complications after the stage I Norwood procedure.11,12 Our findings suggest that tracheostomy in this population for anatomic or functional airway dysfunction does not portend the poor prognosis that has been previously reported. 4 Additionally, mortality risk was significantly higher in patients with delayed sternal closure. Our center routinely attempts to close the sternum in the operating room but will use delayed sternal closure in episodes of significant bleeding or hemodynamic instability. Thus, this risk may be more closely related to suboptimal hemodynamics rather than infection, which has been previously described as a risk factor among patients with a tracheostomy after cardiac surgery. 14 Importantly, there were no patients in our study who developed mediastinitis after tracheostomy.
Our experience differed from other studies examining tracheostomy in congenital heart disease in that the time to tracheostomy was longer than other reported experiences. The median time from surgery to tracheostomy was 43 days in our cohort. Among the more general population of patients with congenital heart disease, timing ranges from 27 to 39 days.1,3,4 The prolonged time to tracheostomy in our population is likely related to attempts to address anatomic or physiologic causes of respiratory failure, such as diaphragm plication, prior to pursuing tracheostomy.
Tracheostomy tended to be safe among our patient cohort with few life-threatening complications. Ortman et al found that tracheostomy dislodgement or obstruction was common causes of death among their patient population. 3 This was not the case in our cohort and may be attributable to intensive monitoring through our home ventilator and interstage follow-up programs. Just four (16%) of our cohort had tracheostomy-specific complications, none of which were life-threatening. Additionally, the majority of our patients were discharged on conventional, bi-level ventilation. The need for a ventilator was not associated with poorer outcomes, which differs from Cotts et al, 4 who reported just 1 in 12 patients who were able to survive to Fontan completion that required bi-level ventilator settings.
Our study is limited inasmuch that it is a retrospective review of a single-center experience and practitioner-specific clinical variability may affect outcomes. Caution should be taken when applying these data prospectively. Additionally, tracheostomy in congenital heart disease is rare, and our patient cohort was small, which may limit more broad application of our results.
Conclusion
Tracheostomy for SV patients after cardiac surgery may be a viable option for patients with anatomic or functional airway dysfunction. In this group, tracheostomy was generally safe and did not prohibit Fontan completion. Overall survival of the entire tracheostomy cohort was similar to the reported general population of patients with SV heart disease. Future, multicenter studies are necessary to continue to clarify the safety of tracheostomy in this population and standardize the approach to airway evaluation and treatment.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.