Customize your JAMA Network experience by selecting one or more topics from the list below.
Pereira KD, MacGregor AR, McDuffie CM, Mitchell RB. Tracheostomy in Preterm Infants: Current Trends. Arch Otolaryngol Head Neck Surg. 2003;129(12):1268–1271. doi:10.1001/archotol.129.12.1268
To study the indications for and outcomes of tracheostomy in a population of preterm infants.
Retrospective analysis of case records.
Two university-affiliated tertiary care children's hospitals.
We identified premature infants who required tracheostomies from January 1, 1997, through January 31, 2001. Information on weight, gestational age, comorbid conditions, indication for tracheostomy, and outcomes was collected. Infants were divided by birth weight into group 1 (<1000 g; n = 19 [very low birth weight]) and group 2 (≥1000 g; n = 14). Comorbid conditions were scored and a total score was calculated for each patient.
Group 1 had a higher incidence of patent ductus arteriosus, bronchopulmonary dysplasia, intraventricular hemorrhage, and retinopathy of prematurity. The incidence of congenital or genetic defects was equal in groups 1 and 2 (11 infants [58%] and 8 infants [57%], respectively). Group 1 had a higher average number of failed extubations (5.17 vs 3.18) and a higher oxygen requirement (48.7% vs 30.3%) compared with group 2. Weight at tracheostomy was essentially equal in groups 1 and 2 (3.6 vs 3.7 kg). Subglottic stenosis and laryngotracheomalacia were equally common findings in groups 1 and 2. The average comorbidity score for group 1 was higher than that for group 2 (6.7 vs 2.8). The most common indication for tracheostomy was ventilatory dependence (n = 24 [73%]), compared with airway obstruction (n = 6 [18%]) and pulmonary toilet (n = 3 [9%]). Overall, 6 patients (18%) had a complication related to the tracheostomy.
Severity of pulmonary disease was the most significant factor associated with the need for tracheostomy in preterm infants. A tracheostomy can safely be performed in these infants with minimal morbidity.
ADVANCES IN the treatment of preterm infants in neonatal intensive care units (NICUs) have dramatically increased their survival rates.1 An important component of these advances is the management of the neonatal airway. Neonatologists have enhanced their understanding and techniques of caring for the infant requiring ventilatory support. The trend is toward fewer days of intubation and ventilation. However, a tracheostomy may at times be required in the preterm infant for prolonged ventilatory support or upper airway obstruction, which precludes safe extubations. Ward et al2 reported an increase in the number of tracheostomies performed for prolonged ventilation and a concomitant decrease in tracheostomies performed for airway obstruction from 1980 through 1990. The rates of tracheostomy in premature neonates with congenital airway abnormalities requiring hospitalization or evaluation by an otolaryngologist have been reported as 10% to 14%,3,4 with comorbid factors present in 77% of these patients.5 Comorbid factors refer to those disease processes that occurred after birth as a result of prematurity and low birth weight. More recent studies have shown a general trend toward lower rates of tracheostomies. The children who undergo tracheostomy tend to be very small and more critically ill.6
Tracheostomy in children has been associated with significant morbidity and mortality.2,7-11 A higher complication rate has been noted in infants, with prematurity increasing the risks associated with the procedure.12-16 The literature is replete with reviews of the complications of pediatric tracheostomies, but none have studied the factors influencing the need for tracheostomy in the preterm infant. To assess these factors, we reviewed the records of infants admitted to the NICU who required a tracheostomy, analyzed their comorbidities using a scoring system, and identified indications for the procedure and its complications. To the best of our knowledge, this is the first study of its kind in the literature.
A census was obtained of all patients admitted to the NICUs at Memorial Hermann Children's Hospital in Houston, Tex, and the University of New Mexico Medical Center in Albuquerque, between January 1, 1997, and January 31, 2001. We performed a retrospective review of all premature infants (gestational age, <37 weeks) who received a tracheostomy and direct laryngoscopy during this time and created a database. The neonatologists at our institution request a tracheostomy when ventilatory support of longer than 8 weeks is anticipated in a neonate who weighs more than 2000 g or one of the same weight who has failed at least 3 extubations.
During the period reviewed, 3793 patients were admitted to the study NICUs. Of these patients, 1660 were born prematurely and 33 of them required tracheostomy. These infants were then divided into 2 groups on the basis of birth weight. Group 1 included 19 infants born weighing less than 1000 g (very low birth weight); group 2, 14 infants born weighing at least 1000 g. Data on weight, gestational age, number of intubations, comorbid conditions, indications for tracheostomy, and outcomes were collected and recorded for each infant. A scoring system was devised to grade patients according to their significant comorbidities. One point was given for each congenital or genetic anomaly, and one each for the presence of patent ductus arteriosus, necrotizing enterocolitis, or retinopathy of prematurity. Two points were given for bronchopulmonary dysplasia and intraventricular hemorrhage, as these conditions had a more significant impact on overall health. The following points were also given for oxygen requirement at the time of tracheostomy: 2 for oxygen requirement of at least 40%; 1 for oxygen requirement of less than 40%; and 0 for no requirement of supplemental oxygen. A total score was then calculated for each patient.
We analyzed ventilatory requirements at discharge or transfer, oxygen requirement after tracheotomy, and time until discharge to confirm the correct indication for tracheostomy. The status of the patient with respect to decannulation at the time of last follow-up was noted.
Table 1 illustrates the study group characteristics. Thirty-two infants (97%) underwent intubation owing to infant respiratory distress syndrome, with only 1 patient undergoing intubation owing to known laryngotracheomalacia.
Table 2 illustrates the primary indication for tracheostomy. At the time of tracheostomy, a direct laryngoscopy and bronchoscopy was performed on 32 infants (97%). Laryngeal findings were similar between the 2 groups, excluding vocal cord paralysis, which was diagnosed in only 1 patient in group 2 and no patients in group 1. Subglottic stenosis or edema was found in 6 infants in group 1 (32%) and 5 in group 2 (36%). Laryngotracheomalacia was found equally in groups 1 and 2, at 26% (5 infants) and 29% (4 infants), respectively. Most infants received a tracheostomy for prolonged ventilatory dependence due to poor pulmonary function and after several failed extubations. All 6 patients (100%) who received a tracheostomy for upper airway obstruction were discharged receiving no greater than 30% oxygen for no longer than 45 days, whereas only 8 (30%) of the 27 patients who received a tracheostomy for prolonged ventilatory dependence were discharged or transferred under similar conditions.
The comorbid conditions present in each group of infants are illustrated in Table 3. Group 1 infants had significantly more comorbidities related to prematurity than did group 2, including patent ductus arteriosus, bronchopulmonary dysplasia, intraventricular hemorrhage, and retinopathy of prematurity. Both groups had equal incidence of various congenital or genetic defects: 11 infants (58%) and 8 infants (57%) in groups 1 and 2, respectively. The average comorbidity score for group 1 was higher than that for group 2 (6.7 vs 2.8).
Overall, 6 patients (18%) had a complication related to the tracheostomy (Table 4). Tracheostomy-related complications developed in 3 infants (16%) in group 1 and 3 infants (21%) in group 2. Only minor complications, including skin breakdown or cellulitis, significant air leak around the tube, and postoperative bleeding, were noted during follow-up. There was no mortality associated with tracheostomy.
The survival rates were similar between the 2 groups, with 14 infants (74%) alive in group 1 and 11 infants (79%) in group 2, with a minimum follow-up of 10 months. No patient had undergone decannulation at the time of last follow-up. The average comorbidity score of those infants who died was 5.8, compared with an average score of 5.0 for the 25 infants who survived.
Infant respiratory distress syndrome secondary to bronchopulmonary dysplasia is the most common indication for endotracheal intubation in the neonate.17 In early experience with mechanical ventilation, tracheostomy was performed on patients requiring ventilatory assistance for even a short period.17 However, the development of the polyvinyl chloride endotracheal tube has made endotracheal intubation the standard of care for infants requiring prolonged ventilatory support. Infants are often intubated for many weeks or even months before tracheostomy is considered.18
The average gestational age and weight of infants admitted to the NICU and requiring mechanical ventilation has not significantly changed during the past 25 years. The mean gestational age has remained constant at 31 weeks. The average birth weight has risen slightly from 1670 to 1807 g in the same period.19,20 However, rapid advances in neonatology currently enable the survival of larger numbers of extremely premature infants with very low birth weights. Many of these neonates weigh less than 500 g at birth, and about 1 to 3 kg several months later.12 These infants have very immature cardiopulmonary systems that require high levels of ventilatory support for many months. As increased numbers of these infants survive, there is a definite need for further research and refinement in respiratory care, instrumentation, and equipment to avoid a concomitant increase in the number of tracheostomies.
The most common indications for tracheostomy in the premature infant population include prolonged ventilator dependence secondary to bronchopulmonary dysplasia and upper airway obstruction due to craniofacial or other structural abnormalities of the upper airway.10 Previous studies have shown upper airway obstruction to be the most common indication for tracheostomy in preterm infants, with the need for prolonged ventilatory support being second.12,21 More recent reports have shown a trend toward prolonged mechanical ventilation as the most common indication for tracheostomy in infants.2,14 In 1978, Heroy et al19 reported a 2.7% incidence of tracheostomy among patients admitted to an NICU. A total of 35.7% of the children who underwent mechanical ventilation required tracheostomy. More recent studies have shown a steady decline in the number of tracheostomies performed on neonates, from 1.77% in 1987 to 0.55% in 2001.1,20 Our study supports this trend, with an overall tracheostomy rate of 0.87% (33/3793) in patients admitted to the NICU. Although abnormal laryngotracheal findings were noted in the airway at the time of tracheostomy, most were related to endotracheal tube placement (subglottic edema and laryngeal edema, 21 patients [64%]) rather than congenital abnormalities (laryngotracheomalacia and vocal cord paralysis, 10 patients [30%]). The overwhelming indication for tracheostomy in our study was prolonged ventilator dependence after several failed extubations (24 infants [74%]).
Few reports in the literature have reviewed the multiple comorbidities that inevitably accompany the preterm infant. Gianoli et al14 studied tracheostomy in the first year of life and devised a comorbidity scoring system to quantify the severity of the child's overall state of health or illness at time of tracheostomy. They concluded that the higher the comorbidity score a patient had, the worse was the prognosis for decannulation and survival. Very low-birth-weight preterm infants had significantly higher comorbidity scores than preterm and full-term infants. Our experience agrees with these findings, with group 1 infants having comorbidity scores averaging 6.7, compared with 2.8 for group 2 infants. In addition, the average comorbidity score for the patients who died during the study was higher than those who ultimately survived (5.8 vs 5.0). The scoring system used in our study placed more weight on the infant's pulmonary status. We believe that this corresponded fairly well with the general health of the neonate. This scoring system, however, has not been validated. Further development and refinement of this scoring system might predict survival in future studies.
Long-term endotracheal intubation has been associated with risks for airway damage, including laryngeal edema, granulation tissue formation, ulceration, anterior web, vocal cord paralysis, subglottic stenosis, tracheomalacia, necrotizing tracheobronchitis, subglottic cysts, tracheal perforation, tracheal stenosis, and other less serious lesions of the airway.1,22-31 With improved neonatal intensive care, the incidence of neonatal subglottic stenosis has progressively decreased during the past 25 years, from 0.9% to 8.3% in the period of 1971 to 1979 to 0.0% to 0.63% in the period of 1990 to 1999.20 Only 11 (0.29%) of 3793 neonates in our institutions required a tracheostomy for subglottic stenosis. Tracheostomy as a procedure has been associated with complications that include tracheocutaneous fistula, tracheal stenosis, tube plugging, accidental decannulation, false passage creation, stomal cellulitis, and stomal granuloma or keloid formation. Less commonly, suture abscess, pneumomediastinum, pneumothorax, and innominate artery erosion have also been reported.10 Previous studies have reported complication rates in preterm infants ranging from 12% to 70%.2,7-16 We had an overall complication rate of 18% with no major complications or mortality in the study group. Three patients in each group experienced tracheostomy-related complications. It must be stated that the retrospective nature of this study limited the data available for collection and analysis.
More high-risk preterm infants with very low birth weights are surviving as a result of greatly improved neonatal intensive care. Concomitantly, the need for tracheostomy in this population has steadily declined. Less than 1% of neonates admitted to the NICU will require a tracheostomy.
The indications for tracheostomy in neonates have also changed. Fewer children require tracheostomy for upper airway obstruction. Those who do are usually older preterm infants with congenital or genetic abnormalities. Very low-birth-weight preterm infants generally require a tracheostomy for prolonged ventilatory support owing to respiratory distress syndrome and its consequences. Their higher comorbidity scores in this study reflect in part the severity of their pulmonary disease.
The incidence of complications of tracheostomy in the neonatal population has decreased secondary to improved technique, anesthesia, transport, and postoperative care. Our low complication rate of 18% indicates that a tracheostomy can safely be performed in the preterm infant with minimal morbidity.
Corresponding author: Kevin D. Pereira, MD, Department of Otolaryngology, The University of Texas at Houston Medical School, 6431 Fannin St, Suite 6.112, Houston, TX 77030 (e-mail: email@example.com).
Submitted for publication November 18, 2002; final revision received April 17, 2003; accepted April 17, 2003.
This study was presented as a poster at the American Academy of Otolaryngology–Head and Neck Surgery Annual Meeting; September 22, 2002; San Diego, Calif.