The incidence of bronchopulmonary dysplasia (BPD) continues to increase, likely secondary to the improved survival of extremely preterm infants [1]. The most severely affected subset of infants with BPD are classified as having grade 3 BPD (also called type 2 severe BPD), which is characterized by the ongoing need for invasive positive pressure ventilation (IPPV) at 36 weeks postmenstrual age (PMA) ^[2][3][2,3]. Infants with grade 3 BPD are six times more likely to need a tracheostomy for chronic ventilation compared to those with grade 2 BPD, which is defined as the need for non-invasive positive pressure (i.e., nasal constant positive airway pressure (CPAP), etc.) at 36 weeks PMA [4]. Tracheostomy is usually considered in infants that are of term-corrected age or older, unable to wean from IPPV within a reasonable time-frame, and/or have had multiple unsuccessful weaning attempts.

Severe BPD is the most common reason for tracheostomy placement in infants [5]. Furthermore, the proportion of infants and children with BPD who are dependent on positive pressure via tracheostomy has increased over time ^[6][7][8][6,7,8]. The BPD Collaborative recently reported using their registry data that 23% of a cohort of 524 patients with severe BPD had undergone tracheostomy [4]. It has been estimated that in the United States, at least 200 infants with severe BPD are discharged on home mechanical ventilation annually, with an estimated 2000 children with severe BPD on home ventilation via tracheostomy at any one time [9].

2. Outcomes following Tracheostomy

2.1. Mortality

The question most important to parents when discussing tracheostomy is “will my child survive?” by which they mean, “what are the chances that my child survives into adulthood and beyond, like other children?” Unfortunately, this question is surprisingly difficult to answer. Published data are limited regarding the mortality rates of infants with BPD and tracheostomy, and these limited data report highly variable mortality rates. Furthermore, it is difficult to quantify the percentage of infants with severe BPD that die prior to tracheotomy placement; one study reported that 7% of infants with severe BPD died prior to tracheostomy placement [10]. Some studies have reported the mortality rates from the time of tracheostomy to the time of initial hospital discharge with a range of 9–23% ^[11][12][11,12]. BPD-associated pulmonary hypertension (BPD-PH), prematurity, small for gestational age status, and tracheostomy placement before one year of age have all been identified as risk factors for mortality in patients with severe BPD who have a tracheostomy ^[13][14][15,23]. One study examined the impact of mean airway pressure and fraction of inspired oxygen (FiO₂) needed at the time of tracheostomy placement, and neither was associated with an increased risk for mortality ^[13][15]. The reported causes of death include tracheostomy complications (accidental decannulation, tracheal obstruction, mucous plugging of the tube), and those related to progression or severity of co-morbid conditions.

2.2. Respiratory Outcomes

Several studies ^{[6][7][10][12][13][15][16]}[6,7,10,12,14,15,16] report median ages for tracheostomy placement, age for discontinuation of positive pressure, and age for decannulation. In one study, 97% of surviving patients were liberated from positive pressure ventilation by five years of age, and those unable to do so were unlikely to ever achieve this milestone [7]. Other studies have examined longer-term respiratory outcomes for infants with BPD and tracheostomy. For example, one study examined childhood respiratory outcomes of patients with severe BPD both with and without tracheostomy and found lower childhood pulmonary function testing results in the tracheostomy group ^[17][24]. This study reported that patients with tracheostomy had a significantly lower mean childhood best forced expiratory volume in one second (FEV1) and mean FEV1/forced vital capacity (FVC) compared to infants with severe BPD without tracheostomy ^[17][24].

2.3. Readmission

As is true for most technology dependent patients, re-admission rates are high in tracheostomy-dependent BPD patients. For example, 73% of infants with BPD and tracheostomy in one multicenter cohort required hospital re-admission for respiratory reasons within the first 12 months of initial hospital discharge [6]. Ehrenkranz et al. ^[18][26] found that the hospital re-admission rate for all infants with severe BPD was 39%, while Jensen et al. [2] reported that 29% of infants with grade 3 BPD had 2 or more hospitalizations for respiratory reasons. The most common reasons for re-admission for infants with BPD and a tracheostomy include respiratory infections and tracheostomy-related complications ^[19][20][27,28].

2.4. Tracheostomy Complications

Tracheostomy complications can be categorized as early (post-operative days (POD) 0–7) and late (POD > 7). The early period is critical for appropriate maturation of the stoma. Early complications are primarily related to immaturity of the stoma and include posterior tracheal wall tear, accidental decannulation, false passage, pneumothorax, subcutaneous emphysema, bleeding, infection, and skin breakdown ^[5][21][5,29]. In practice, the first tracheostomy tube change often marks the transition from the early to the late post-operative period with confirmation of the maturation of the stoma and ability to resume routine care. Late complications include cellulitis, tracheitis, tracheo-innominate fistula, mucus plugging, airway obstruction, and granulation tissue formation [5].

2.5. Morbidities

Infants with severe BPD requiring tracheostomy often have other morbidities, with one multicenter study noting 58% of infants with a tracheostomy also had BPD-PH, of which 33% required outpatient pulmonary antihypertensive medications [6]. It has been reported that lesions of the large airways are also common in this population, including tracheobronchomalacia (TBM) (40–74%), subglottic stenosis (48%), and airway edema (48%) ^[6][22][6,30]. Concurrent large airway lesions, such as TBM, may result in sometimes dramatic increases in the work of breathing that can delay decannulation ^[23][31].

2.6. Growth and Feeding

.6. Growth and Feeding

All infants with severe BPD are at high risk for sub-optimal growth, which can impact short- and long-term pulmonary outcomes ^[21][24][29,32]. Intrauterine growth restriction, small for gestational age status, and postnatal undernutrition have all been associated with delayed alveolar development, abnormal lung healing, and reduced postnatal gains in lung function ^[25][33]. Alternatively, positive linear growth has been associated with an ability to wean from respiratory support in infants with BPD ^[26][34]. Patients who demonstrate catch-up weight gain and linear growth have also been shown to have improved pulmonary function testing in childhood ^[27][28][35,36]. Information about growth-related outcomes in patients with BPD requiring tracheostomy is limited. One single-center study examined growth velocity before and after tracheostomy placement in infants with severe BPD and found stable improvements in weight and length growth by four weeks after the first tracheostomy tube change ^[21][29]. In this study cohort, there was no change in pre- and post-operative respiratory severity scores and an overall decrease in caloric intake following tracheostomy ^[21][29], leading the authors to speculate that this improving growth may result from reduced stress and work of breathing following tracheostomy.

2.7. Neurodevelopment

All infants with severe BPD are at risk for neurodevelopmental impairment (NDI) ^[29][38], and many factors can impact neurodevelopment in severe BPD. For example, postnatal corticosteroid exposure, particularly dexamethasone, has been shown to increase NDI ^[30][39]. Infants with severe BPD are often exposed to prolonged analgesics and sedatives, which have also been shown to negatively affect neurodevelopment. Midazolam has been shown to impact hippocampal development and long-term learning memory ^[31][40]. Opioids have been shown to lead to long-term changes in memory and brain function secondary to apoptosis in microglial cells and neurons ^[31][40]. Additionally, infants with severe BPD often experience frequent skin-breaking laboratory draws and painful procedures which have been associated with worse neurodevelopmental outcomes in the first two years of life ^[32][41]. While all infants with BPD are at risk for NDI, infants with severe BPD and tracheostomy are at the highest risk ^[29][33][22,38]. It is unclear how tracheostomy placement impacts neurodevelopment in patients with severe BPD; it may be a selection bias for those with the severest disease and therefore the greatest exposure to the negative stimuli discussed above, or there may be something intrinsic in having a tracheostomy that negatively impacts neurodevelopment. Although studies are necessary to elucidate the mechanisms associated with the increased risk of NDI in severe BPD patients with tracheostomies, these patients require intensive child development interventions following tracheostomy placement and continuing throughout childhood. It is unclear if the timing of tracheostomy impacts neurodevelopmental outcomes. Given the importance of oral stimulation on neurodevelopment in infancy, it is reasonable to postulate that earlier tracheostomy, involving taking away the endotracheal tube and fixation devices, may have a positive impact on neurodevelopment in this extremely high-risk group. One retrospective cohort study examined the neurodevelopmental outcomes at 18–22 months of age in former preterm infants who underwent tracheostomy among 16 centers in the NICHD Neonatal Research Network to assess the association of tracheostomy with adverse neurodevelopmental outcomes ^[33][22].

3. Tracheostomy Decision Making

3.1. Tracheostomy Indications

There are no accepted national standards for tracheostomy indications, and therefore indications for tracheostomy vary from center to center as well as from provider to provider [10]. In most centers, tracheostomy is often considered for infants with severe BPD that require “long-term” IPPV or have structural airway problems that cannot be immediately surgically corrected. However, there are no standard definitions for long-term IPPV or even for structural airway problems. Other potential indications, either alone or in combination, may include infants that cannot be liberated from non-invasive positive pressure support (i.e., nCPAP), have significant growth failure, experience equipment interface difficulties, and/or need a relatively high and on-going supplemental oxygen requirement.

3.2. Timing of Tracheostomy

The decision to place a tracheostomy is usually very difficult for families and caregivers; this is in part due to fear and anxiety for families and a common feeling among healthcare providers of failure when an infant needs a tracheostomy. Therefore, it is common to delay tracheostomy decision making until well past the diagnosis of severe BPD which is made at 36 weeks PMA, and even well past corrected term age (40 weeks PMA). Another contributing factor to waiting so long to decide on tracheostomy is that it has been reported that mortality and postoperative complications related to tracheostomy are seen more commonly in preterm infants compared to term infants ^[34][43]. However, just as there is no consensus on indications for tracheostomy, there is currently no consensus on when the appropriate time for tracheostomy placement. The available literature consists entirely of observational studies and suggests a median age of tracheostomy placement between 43–51 weeks PMA, and that the timing of tracheostomy placement is highly center-dependent. The timing of tracheostomy placement often involves waiting for multiple failed attempts at weaning from mechanical ventilation ^[12][35][12,44]. The notion that “we should try one more time” often underlies this and is a deliberate attempt to avoid tracheostomy given the risks for mortality and morbidity as well as the implications for parents and caregivers of a technology dependent infant. Clinicians will often also try to optimize nutrition and lung growth, believing that with enough time this may allow some infants to be successfully liberated from mechanical ventilation without needing a tracheostomy ^[34][43].

3.3. Family-Centered Care

A family-centered approach that is built on shared decision making is absolutely necessary for all conversations and decisions regarding tracheostomy placement. When counseling families regarding tracheostomy, providing the available information regarding short- and long-term outcomes in an understandable fashion for each family is essential. Determining the most important factors influencing each family’s decision making is also key in arriving at a decision that will provide the very best outcome for a given patient and family. One study found that parents of infants with BPD often prioritize outcomes related to physical health and safety over outcomes related to neurodevelopment. In theis study, parents were more concerned about breathing, growth, feeding, and safety outcomes, and were more willing to accept difficulties with learning and behavior ^[36][45].

3.4. BPD, Tracheostomy, and Social Determinants of Health

There is a growing body of literature that has examined the effects of race/ethnicity and socioeconomic status on outcomes for infants and children with severe BPD. Infants born to Black mothers have been shown to have an increased likelihood of mortality and an increased length of hospital stay compared to infants born to White mothers ^[37][50]. Sociodemographic status measured by neighborhood deprivation index and neighborhood median household income have also been associated with an increased likelihood of mortality and higher rates of readmission in patients with BPD ^[38][39][51,52].

3.5. Care Coordination

It is important that patients be at a center that utilizes an interdisciplinary team for severe BPD management when deciding on tracheostomy ^[40][53]. Evidence suggests that a multidisciplinary care team can improve survival in infants with severe BPD and tracheostomy ^[41][42][19,42]. Additionally, multidisciplinary discussions with the family before tracheostomy regarding short- and long-term risks, outcomes, prognosis, discharge planning, and outpatient care are essential. Multidisciplinary team members should include the intensive care team, pediatric pulmonology, otolaryngology, pediatric surgery, palliative/supportive care, psychology, nursing, social work, and care management. Consideration of the long-term outpatient support needed for the infant if the family and team proceed with a tracheostomy is essential. Infants with tracheostomy require multidisciplinary care in the outpatient setting, with co-management by a general clinician and a respiratory subspecialist, such as a pediatric pulmonologist or neonatologist ^[43][54]. At least two trained caregivers are needed at home to care for the infant after discharge, one of whom should always be awake and present in the home ^[43][54].

3.6. Making the Decision to Place a Tracheostomy

First and foremost, there is no high-quality evidence on which to base a decision on, or the timing of, for tracheostomy in patients with severe BPD. There is clearly a group of infants with severe BPD who cannot be liberated from invasive positive pressure ventilation, and to facilitate airway, lung, and neurological development at some point a stable and safe airway (i.e., a tracheostomy) is necessary. However, currently reaching that decision can be quite difficult. Obviously, the decision to place a tracheostomy must be a shared decision including the parents of the patient and the various disciplines involved in the patients pre- and post-tracheostomy care. Occasionally, scholars have had the experience where families push for a tracheostomy; however, it is much more common that parents and/or caregivers want to delay tracheostomy placement and try “just one more time” to extubate the patient. Thus, this decision is often made only after some sort of agreement is reached that everything that has been tried to avoid tracheostomy placement has failed. This may be completely appropriate given that there are substantial risks with tracheostomy, but it may inadvertently cause undue stress for parents and caregivers, and it may make the decision seem arbitrary, subjective, and/or one-sided. To attempt to make this difficult decision at least a bit more objective and include longitudinal assessments, scholars have developed a tracheostomy scoring tool for the BPD unit that includes risk factors and assesses them longitudinally to monitor trends over time. The risk factors used in this tool are based on  the experience and include respiratory factors (prolonged requirement for high or increasing supplemental oxygen, inhaled nitric oxide, and/or anti-pulmonary hypertensive medications), growth factors (sub-optimal growth, especially linear growth, despite good nutrition), neurodevelopmental factors (ability to participate in developmentally appropriate activities), and medication needs (high-dose chronic systemic steroids, multiple neuro-sedative medications, etc.). Trending this tracheostomy score starting at 36 weeks in intubated patients facilitates conversations and family education around the potential for tracheostomy placement.

4. Post-Tracheostomy Management

4.1. Tracheostomy Care

The initial post-operative period is a time of high risk for complications related to accidental decannulation, with difficulty replacing the tracheostomy tube, development of a false passage, and wound and skin care complications. For this reason, tracheostomy patients are typically monitored closely by the surgical team until the first tracheostomy tube change, usually performed between post-operative days four and seven ^[44][55]. Following this initial period, care of the tracheostomy site includes tracheostomy tube changes every two to four weeks, daily tracheostomy tie changes, and frequent skin and stoma care and cleaning. Some patients may intermittently develop so called tracheostomy-associated tracheitis, which is a poorly defined clinical entity that most sources in the literature describe as an increase or change in secretions along with signs of clinical worsening, including fever or the need for increased respiratory support ^[45][56].

4.2. Discharge

The American Thoracic Society developed a clinical guideline for infants undergoing tracheostomy, focusing on discharge criteria, caregiver education, and chronic home ventilation needs ^[43][54]. At least two trained family caregivers need training and education on caring for the child at home. This training for parents is extensive and includes respiratory status assessment, tracheostomy care, tracheostomy tube changes, suctioning, and how to respond to emergencies such as tube displacement. Additionally, caregivers should receive training on the home ventilator, medication administration, and feeding tube management.

4.3. Outpatient Management

The clinical guidelines for the outpatient management of infants with BPD have become available recently ^[46][47][63,64]. These clinical guidelines suggest long-term monitoring with lung imaging and pulmonary function testing. Commonly used medications such as bronchodilators, steroids, and diuretics are also discussed. There are additional recommendations for the management of home ventilation and supplemental oxygen. These guidelines are based on systematic reviews of the available literature and expert option, and unfortunately the available literature to guide these recommendations is limited with low certainty of evidence ^[46][47][63,64]. This lack of evidence is a major contributing factor to the significant variation of care regarding the outpatient management of infants with severe BPD with tracheostomy ^[9][43][9,54].

4.4. Considerations for Decannulation

The American Academy of Otolaryngology issued a clinical consensus statement regarding many facets of tracheostomy care to help reduce care variations between clinicians, including the assessment of readiness for and accomplishment of decannulation ^[48][66]. Once a patient has been weaned from mechanical ventilation, including during episodes of illness, the status of swallowing and the patency of the airway should be assessed. There should be no documented ongoing aspiration that would necessitate the presence of the tracheostomy tube for pulmonary toilet and secretion clearance. The evaluation of airway patency includes awake flexible laryngoscopy, ideally revealing at least one mobile vocal fold, and micro-direct laryngoscopy to confirm airway patency distal to the glottis. The patient should tolerate capping of the tracheostomy tube all day. Once this is tolerated, patients should undergo either capped overnight polysomnography or a nighttime capping trial in the hospital setting. Finally, if a patient is admitted to the hospital for decannulation and observation on pulse oximetry monitoring for one to two nights before discharge home without the tracheostomy tube, a dressing should be kept over the stoma, and water precautions should continue until stoma closure is confirmed. Up to 65% of patients can have a persistent tracheocutaneous fistula (TCF) six weeks following decannulation ^[49][67] and require surgical fistula closure.

5. Conclusions

The decision to proceed with tracheostomy comes with significant risks for mortality and morbidity. However, there are currently no available alternatives for long-term invasive or non-invasive positive pressure support outside of the hospital setting. Thus, the decision to place a tracheostomy is often very stressful for both families and caregivers. The development of center-specific guidelines for assessment of need for tracheostomy placement can alleviate some of that stress and result in better shared decision making. However, high-quality evidence is urgently needed to determine the indications and timing for tracheostomy placement and specific risk factors to aid in identifying which patients are at the highest risk for mortality.