Review of pediatric tracheotomies and long-term outcomes performed in a low-middle-income country

INTRODUCTION

Tracheotomy is a life-saving procedure wherein an incision is made into the neck to access the trachea. The resultant tracheostomy refers to the opening to the trachea, which is created as a result of the tracheotomy. The procedure can be performed emergently or electively. General indications for tracheotomy include (1) bypassing upper airway obstruction (UAO), (2) preventing laryngotracheal stenosis and reducing ventilatory dead space in patients who require prolonged ventilation, and (3) facilitating pulmonary toileting.^[1]

Since its inception, tracheotomies in children have been used primarily to relieve UAO secondary to infection, such as croup. Over the past few decades, thanks to antibiotics and advances in medicine, the indications have shifted to help manage complex patients with chronic illness.^[2,3]

In low-middle-income countries (LMICs) such as South Africa, living with pediatric tracheostomies is a unique challenge for both patient and family, particularly due to socioeconomic factors. In a general household survey of South Africa conducted in 2018, 30% of children lived in unemployed households, 30% lived without access to potable water, and 59% lived below the poverty line.^[4] In addition, prior studies analyzing the backgrounds of tracheotomized patients in South Africa found that 20–24% of parents caring for home-ventilated children lived in an informal house (shack), 53–66% had incomplete schooling,^[5,6] and that the primary caregivers were single 23% of the time.^[6] Due to the resource-constrained backgrounds of most families, home-based nursing assistance is rarely available. Thus, the primary caretakers must learn to take full responsibility for tracheostomy care. Another study by Din et al. showed that after tracheotomy, carers in full-time employment dropped from 72% to 28%.^[7]

The Red Cross War Memorial Children’s Hospital (RCWMCH) in Cape Town is a quarternary hospital that mainly treats children from poor socio-economic backgrounds. In the absence of formalized respiratory therapists, who typically assume the primary role in caring for tracheotomized patients,^[8] the Breatheasy © program has been in operation for 30 years as a multidisciplinary team. It functions to empower families of tracheotomized patients, teaching them how to care for their children at home without additional nursing assistance, especially under these challenging circumstances.

The objectives of this study were to describe the patient profile, clinical characteristics, and long-term outcomes of pediatric tracheotomies performed at RCWMCH in Cape Town, South Africa, from 1994 to 2023. In addition, we aimed to compare our findings with similar reports in the literature from various regions around the world. A secondary aim of the study was to inform multidisciplinary teams caring for children with tracheotomy (or who may need tracheotomy) and to lead to refining/modifying management protocols.

MATERIALS AND METHODS

RESULTS

Over the 29 years, a total of 671 patients had tracheostomies performed, one of whom did not have an indication recorded in the database. The number of tracheotomies per year was <10/year up until 2008. Over the next 4 years, that number jumped to 53 and has stayed between 30 and 70 from that time on [Figure 1].

Close

Figure 1: Number of tracheotomies performed each year by indication category. UAO: Upper airway obstruction.

Export to PPT

The median age at which tracheostomy was performed was 14 months (IQR 2–57). Almost half of the tracheotomies performed were done on patients under 1 year of age (311/671; 46.3%). Of those tracheotomies performed on patients under 1 year of age, 82% were performed on children under 6 months of age (267/325) [Figure 2].

Close

Figure 2: (a) Tracheostomies by years of age showing a significant proportion of tracheostomies done under 1 year of age. (b) Tracheostomies done by months of age further showing a significant proportion of procedures done on patients <1 month of age. (c) Tracheostomies by weeks of age showing the distribution of procedures in patients <8 weeks of age.

Export to PPT

When stratified by indication category, patients with UAO from laryngeal and midface sources had the lowest median ages at which tracheostomies were performed at 4 (IQR 1–15) and 2 (IQR 0–7.25) months of age, respectively. The indication categories that had the oldest patients were those requiring tracheostomies for traumatic injuries and acquired neurologic insults, with a median age of 79 (IQR 44–113) and 48 (IQR 16.5–84.5) months of age, respectively. The other indication categories, including uncategorized UAO, congenital muscular disorders, congenital neurological disorders, and pulmonary/cardiac, ranged between those extremes at a median of 13 (IQR 2–44), 43 (7.75–125.25), 27 (IQR 3–59.5), and 11(IQR 4–49) months, respectively [Table 1].

While most of the tracheotomies were performed on children who were <50 km from the hospital, 30.1% (207/669) of tracheotomies were performed on children who lived >50 km from the hospital.

In our patient population, the most common indication category was laryngeal UAO at 23.9% (160/670), followed by patients with skeletal/midface hypoplasia at 16.7% (112/670). The overall rate of decannulation in our population was 59.9% (323/670). The highest decannulation rate was found in the patients with acquired neurologic deficits, including infectious causes and tumors/masses such as Guillain-Barré Syndrome and ependymomas of the brainstem (72/77; 93.5%). Similarly, patients who required tracheostomies due to trauma were likely to be decannulated (57/65; 87.7%). The least likely indication category to be decannulated was the congenital muscular category at a rate of 14.8% (8/54). Patients with a laryngeal source of UAO were eventually decannulated more often than not (85/160; 53.1%) and were more likely to be decannulated than those with midface/ skeletal sources of UAO (42/112; 37.5%).

The overall mortality rate in our population was 19.6% (131/670). The indication category with the highest mortality was congenital neurological, with a rate of 42.9% (18/42), followed by those with cardiac/pulmonary and congenital muscular indications, 31.7% (13/41) and 29% (22/76), respectively. The three categories with the lowest mortality were traumas, acquired neurologic, and laryngeal UAO, with rates of 8.45% (6/71), 11.5% (10/87), and 13.1% (21/160), respectively [Figure 3].

Close

Figure 3: (a) Bar graph highlights the number of patients with current tracheostomy tubes, decannulated, or deceased by diagnosis category. (b) Pie chart highlighting the total percentage of tracheotomies performed by indication. (c) Table showing the numbers and percentages of the number of patients tracheotomized, decannulated, and deceased by indication category. UAO: Upper airway obstruction.

Export to PPT

Out of 670 patients, 30% of patients (201/670) were documented to require ventilation while tracheotomized. 97.3% (74/76) of patients with congenital muscular indications for tracheotomy required ventilation. Laryngeal UAO (likely due to bilateral vocal cord palsy, airway papillomatosis, or glottic-subglottic stenosis), led to the least percentage of patients needing ventilation (7/160; 4.4%). This was followed by midface/syndromic obstructions and traumas (13/110; 11.8% and 10/71; 14.1% respectively). Of those with congenital neurologic indications for tracheostomy, all who required ventilation had a diagnosis of Ondine’s curse or central hypoventilation syndrome. Furthermore, 59.5% (44/74) of those with congenital muscular problems continued to require their tracheotomies, while 81.4% of those with an acquired neurological issue were able to be decannulated (35/43) [Figure 4].

Close

Figure 4: (a) Percentage of tracheotomized patients requiring ventilation at some point in their care by indication category. (b) Number of ventilated patients who have a current trache in place, were decannulated, or are deceased. UAO: Upper airway obstruction.

Export to PPT

The median time to death after tracheotomy placement was 6 months. Patients with congenital neurologic indications for tracheotomy had the longest time to death (36; IQR 16–64) while laryngeal and skeletal/midface UAO had the shortest time to death (4; IQR 3–11 and 3; IQR 1–5, respectively) [Figure 5].

Close

Figure 5: Box and whisker plots of the time to death after tracheotomy based on indication category. UAO: Upper airway obstruction.

Export to PPT

Patients who received tracheostomies for congenital neurologic and muscular conditions kept their tracheostomies for the longest with medians of 61 (IQR 12.5–92.25) and 51.5 (IQR 20.75–102.75) months, respectively. Those with acquired neurologic injuries or traumas necessitating tracheostomies kept them for the least amount of time, with medians of 1 (IQR 0–3) and 0 (IQR 0–11.5) months, respectively. Patients with laryngeal and midface UAOs required tracheostomies for a median of 29 (IQR 12–57) and 30 (IQR of 11–64) months, respectively [Figure 6].

Close

Figure 6: For patients with tracheotomies who were eventually decannulated, these box and whisker plots describe the time to decannulation in months by indication category. UAO: Upper airway obstruction.

Export to PPT

DISCUSSION

A child’s need for a tracheotomy often carries a profound psychological impact for caregivers. In navigating this new reality, several critical questions frequently arise: Will the tracheostomy be permanent? Will the child require home ventilation? How will we manage their daily care? Most poignantly, parents often grapple with the fear of whether their child might die while living with a tracheostomy.

While several studies have addressed these questions, they lack the sample size and granular detail of the present work. Most existing literature relies on broad categories, primarily airway obstruction, prolonged ventilation, or neurologic impairment, which fail to capture the full spectrum of underlying indications and their implications. This study is the first to move beyond these general classifications, subcategorizing and comparing outcomes across eight distinct clinical phenotypes to provide a more precise clinical picture. The biggest advantages of this study are the subcategorization of indications into clinical phenotypes and the large sample size that makes these subcategories meaningful. Whereas prior studies have largely categorized indications for tracheotomy as either UAO, assisted ventilation, or neurologic impairment,^[9-15] this study goes a step further and subcategorizes the indications into eight categories. By grouping indications by anatomic level (for UAO) and pathophysiology (for prolonged ventilation), they offer a more meaningful framework.

In a broad sense, these questions have been answered by prior studies. Our demographics and outcomes align well with this published data. Published studies from around the world have shown that anywhere between a quarter and two-thirds of patients requiring tracheotomies are completed on patients under 1 year of age, while our data showed that about half of patients had a tracheotomy done under 1 year of age.^[9,12-14] The proportion of patients with UAO as a primary indication for tracheostomy appears to be similar as well, with a published rate of anywhere between a third to three quarters^[9-16] compared to our rate of about half. Decannulation and mortality rate were also similar to published rates of decannulation of anywhere between a third and three quarters and mortality of approximately a quarter^[9,11,13-15] compared to our decannulation and mortality rate of 60% and 20%.

By reviewing the data in our study, a picture of the natural history of the tracheotomized patient for the different subcategories begins to emerge. For example, a patient with a laryngeal obstruction is likely to have a tracheotomy from an early age and is also reasonably likely to have their tracheostomy removed. In addition, mortality in this group occurs at an early age. Patients with congenital neuromuscular conditions are likely to have a tracheostomy in place permanently. This is valuable information that can be used to help guide the counselling clinicians offer to families pre-procedurally.

In conclusion, despite the numerous challenges faced in an LMIC, the broad outcomes found in our cohort at RCWMCH mirror those found in high-income countries. Despite the lack of respiratory therapists (South Africa does not have this occupation) and despite the lack of resources and, in some cases, even the lack of running water or living in a house, patients and families are still able to have similar outcomes, thanks in large part to the work of the multidisciplinary team set up through the Breatheasy © program. This success suggests that with the use of a high-functioning multidisciplinary team and patient follow-up with the clinician, other LMIC may be able to achieve similar results to high-income countries for patients receiving tracheotomy.

This study has several limitations. Its retrospective nature and reliance on a maintained database led to instances of missing data and potential inaccuracies. The extended study period also introduces temporal bias, as evolving standards of care and management techniques may have influenced longitudinal outcomes. Furthermore, we lacked data on intermediary procedures performed between tracheotomy and decannulation. Finally, the absence of detailed comorbidity and mortality data restricts our ability to fully analyze the drivers of specific patient outcomes.

CONCLUSION

This 29-year retrospective review demonstrates that pediatric tracheotomy in an LMIC can yield long-term outcomes comparable to those in high-income settings, despite significant socio-economic challenges. By stratifying patients into eight distinct clinical phenotypes, we identified that patients with acquired neurologic deficits and traumas achieve the highest decannulation rates (93.5% and 87.7%, respectively), while those with congenital muscular conditions face the lowest likelihood of decannulation (14.8%) and higher ventilation requirements.

Our findings highlight that the “natural history” of a tracheostomized child is heavily dependent on their specific indication category, with laryngeal and midface UAO patients requiring shorter durations and experiencing mortality at an earlier age compared to the prolonged needs of the congenital neurologic group. These data provide a framework for more precise pre-procedural counseling and clinical management. Ultimately, the success of these patients, 60% of whom were successfully decannulated, underscores the efficacy of a multidisciplinary team model, like the Breatheasy © program, in empowering caregivers and overcoming barriers to care in resource-constrained environments.