Pediatric tracheostomy: A Changing procedure?

Author(s): Ralph F Wetmore, Mary E Thompson, Roger R Marsh, Lawrence W C Tom 
Publication title: The Annals of Otology, Rhinology & Laryngology. St. Louis: Jul 1999. Vol. 108, Iss. 7; Part 1. pg. 695, 5 pgs 

Copyright Annals Publishing Company Jul 1999

In 1982, the experience with tracheostomy at The Children's Hospital of Philadelphia was reported for 1971 through 1980. We have now reviewed 450 cases for the period from 1981 through 1992, and compared the characteristics of these cases with those in the previous review. Long-term follow-up was available on 83% of cases, and the median follow-up was 2.96 years. Patients received a tracheostomy for airway obstruction (38%), chronic ventilation (53%), or multiple indications (9%). The mean duration of tracheotomy (adjusted for death and loss to follow-up) was 2.13 years. The tracheostomy-related mortality was 0.5%, and the nontracheostomy-related mortality was 22%. Nineteen percent of patients had complications in the first postoperative week, and 58% had 1 or more late complications. In comparison with the previous study from our institution, there was a great increase in long-term tracheostomy and a continuing trend away from tracheostomy for short-term airway management. Better monitoring and improvements in parental teaching may have contributed to a decrease in tracheostomy-related mortality.


In 1982, we reported the pediatric tracheostomy experience at The Children's Hospital of Philadelphia from January 1971 through December 1980.1 Several important trends were evident from that review. The yearly incidence of tracheostomy remained relatively constant during the decade of study. Upper airway obstruction (24%) and neurologic disorders (24%) were the primary diagnoses. The most common indication for tracheostomy was prolonged intubation (53%), with the other major indications being upper airway obstruction (39%) and pulmonary toilet (4%). The overall complication rate was 49%, with an early complication rate of 28% and a late complication rate of 53%. The tracheostomyrelated mortality was 2%. To determine if there had been a change in the demographics and complication rates of pediatric tracheostomy over the past 2 decades, we conducted a review of the tracheostomy experience from 1981 through 1992 for comparison with our previous study. 


Tracheostomies performed at The Children's Hospital of Philadelphia from January 1981 through December 1992 were reviewed retrospectively. A list was compiled from the operative logs and included tracheostomies performed by the otolaryngology and general surgery services. A total of 450 cases were identified. These were analyzed with respect to the following variables: age, diagnosis, indication for procedure, duration of tracheostomy, early and late complications, and mortality. Early complications were defined as those occurring in the operating room or during the first postoperative week. At the end of the first week, the tracheostomy stoma was felt to be mature, and the first tube change was performed. Late complications were defined as those occurring after the first week. Documentation was variable, because some patients were transferred to other institutions within days of their surgery and because parts of the chart could not be located in some cases. Demographic data, diagnoses, and indications for tracheostomy were available for at least 97% of cases. Records were available for 373 patients during the first week. Long-term follow-up data (90 days or until death or decannulation) were available for 373 patients. Some children in the early complication analysis were lost to long-term follow-up or died within the first week. An equal number of cases excluded from that analysis received long-term tracheostomy care by our department and are included in the late complication series. 


The great majority (67%) of the pediatric tracheostomies were performed during the first year of life (Fig 1). The number of tracheostomies performed remained relatively constant during the years of the study, without obvious changes in age distribution. As with the previous study, there was a slight male predominance (58%). The primary indications for tracheostomy remained virtually unchanged from the 1971-80 study, with prolonged ventilation remaining the most frequent indication for surgery (53% of cases in both series). Airway maintenance was the primary indication in 39% of cases in the first series and in 38% in this series. The remaining cases had other or multiple indications. 

Bronchopulmonary dysplasia (BPD) was the primary diagnosis of greatest incidence (29%), followed by neurologic disorders (24%; Fig 2). Use of tracheostomy for the management of croup was much less common in the 1981-92 study compared to the 197180 study: 1.1 % versus 7.6%, a significant reduction (p < .001). The duration of tracheostomy ranged from 5 days to 12.8 years. The median was 778 days, as estimated by using the Kaplan-Meier statistic to adjust for intercurrent death and incomplete followup. An increase in tracheostomies of greater than 24 months' duration in this study compared to the previous one (53% versus 17% as estimated from KaplanMeier curves) was due to the increase in chronically ventilated patients (Fig 3). Of the 373 patients with long-term follow-up, 190 were decannulated, 94 died, and 89 were still cannulated. For these cannulated patients, the median follow-up was 2.96 years. 

Among the 373 patients with 1 week of followup, 71 patients suffered 92 early complications (Table 1). Obstruction of the tracheostomy tube was the most common early complication, seen in 10% of cases. Accidental decannulation was seen less frequently than in the 1971-80 study (5% versus 24%). In examining the prevalence of early complications, there was no statistically significant difference between children under 1 year of age and those older than 1 year. 

Among the 373 patients with long-term followup, there were 409 late complications affecting 216 patients (Table 2). Of the late complications, tracheal granulomas were most common (41%), while the prevalence of tracheocutaneous fistula remained almost exactly the same as in the 1971-80 study (20%). Children under 1 year of age were significantly more likely than older children to have complications; 66% of the younger and 42% of the older children had 1 or more complications (p < .001). 

In reviewing overall mortality, the Kaplan-Meier survival curves, adjusted for cases decannulated or lost to follow-up, show a higher short-term mortality in the 1971-80 series, which included many children receiving a tracheostomy for cardiac surgery (Fig 4). The adjustment for decannulated patients permits comparisons between the 2 series in spite of their very different durations of tracheostomy and illustrates the survival experience of those patients who were tracheostomized for long periods, but it does exaggerate the actual mortality per tracheostomy. Tracheostomy-related mortality declined from 2% in the 1971-80 study to 0.5% in the current series. Mortality from other causes, not adjusted for loss to follow-up, was 22% in this series and 26% in the previous series. 


The incidence of tracheostomy in the 1981-92 study remained relatively constant with only a slight decrease in number compared to the 1971-80 study from our institution: 37.5 cases per year versus 42 cases per year.l Figure 5 demonstrates a declining incidence of tracheostomy when compared to the number of hospital admissions. Figure 1 illustrates the peak prevalence at 3 to 4 months of age - a peak noted by other authors.2 Nearly three quarters of patients in this series were under the age of 2, compared to 58% in the 1971-80 series. This trend toward younger age is illustrated in Fig 6 and has been reported by others.3 In several series of pediatric tracheostomy patients, the male-female ratio ranges from equal to a male predominance of 2 to 1.2-9 Our series had a male-female ratio of 1.41 to 1. 

fig1.1.jpg (278052 bytes)
Fig 1.1 
Fig 2. 
Fig 3. 

table1.jpg (254871 bytes)
Fig 4. 
Fig 5. 

In comparing the 1971-80 and 1981-92 series, there has been essentially no change in the proportion of indications for tracheostomy. These values are similar to those reported by Waki et al,8 but differ from those of Crysdale et al,4 who reported a much greater incidence in children suffering from airway obstruction. Ward et al9 have reported an increasing incidence of tracheostomy for prolonged ventilation and a coinciding decrease for airway obstruction, but their study included only children under 5 years of age. 

In the 1971-80 study, upper airway obstruction and neurologic disorders were the most frequent diagnoses. In the current study, the incidence of neurologic disorders has remained the same, while BPD has become the most frequent diagnosis of pediatric tracheostomy patients. The statistically significant increase in patients with BPD in the 1981-92 study as compared to the 1971-80 study (29% versus 12%; p < .001) is a reflection of the longer survival of patients with this disorder who would not have survived in the past. While the incidence of tracheostomy for BPD has increased, there has been a coinciding decrease in the need for tracheostomy in the treatment of laryngotracheobronchitis (croup) that is a reflection of improved airway management by pediatric intensivists at this hospital. 

In the 1971-80 study, 34% of the tracheostomies were of less than 1 month's duration, adjusted for losses to follow-up and death. Use of tracheostomy for cardiac and craniofacial surgery and to manage the airway in acute epiglottitis accounted for many of these cases. In the 1981-92 review, only 6% of tracheostomies were of less than 1 month's duration, and approximately 72% were of up to a year's duration. This increase in the duration of tracheostomy is demonstrated in Fig 3 and reflects an increasing need for chronic ventilation and increased survival of such patients - a change noted by other authors.3,6 

The prevalence of early complications in the 1981-92 study mirrors that of the previous study, although obstruction of the tracheostomy tube occurred more commonly in the 1981-92 study. Other major complications include accidental decannulation and infection involving either the stoma or the trachea. These 3 major complications were reported commonly in other series.49 Pneumothorax and pneumomediastinum occurred in a few patients in this series and were the most frequent complications in series by Arcand and Granger,3 Line et al,5 and Waki et al.8 

Tracheal granulation and persistent tracheocutaneous fistula were the most commonly reported late complications, as observed by others.3,4,8,9 Accidental decannulation occurs more frequently than the 6% represented in the current series. Many episodes of accidental decannulation remain unreported unless significant morbidity occurs. 

The tracheostomy-related mortality in the current series, at 0.5%, was one quarter of that seen in the 1971-80 series. This mortality compares favorably to tracheostomy-related mortalities seen in other series, which range from 0.9% to 5%.2-6,8,9 The decrease in tracheostomy-related mortality may be due in large part to the introduction of practical pulse oximetry monitors; the first Nellcor model was cleared for marketing by the US Food and Drug Administration in late 1982 and rapidly gained acceptance at this institution. Parent education has also gained more attention over the years, and may have contributed further to the decline in tracheostomyrelated mortality. The non-tracheostomy-related mortality of 22% in this series compares to 26% in the previous series and a range of 14% to 33% in the literature.1-6,8,9 


fig6.jpg (58578 bytes)

fig 6

Pediatric tracheostomy has shown a slight decrease as a percentage of hospital admissions over the past 20 years. The survival of a greater number of critically ill infants who require chronic ventilation has placed a greater demand upon the procedure of tracheostomy to provide access to the airway. As a result, pediatric tracheostomy was performed more often in younger children, especially infants, during the past decade than in the previous one. Early and late complications remain unchanged, although there have been shifts in their incidence. Better monitoring and improvements in parental teaching may have contributed to the decrease in tracheostomy-related mortality.


1. Wetmore RF, Handler SD, Potsic WP. Pediatric tracheostomy. Experience during the past decade. Ann Otol Rhinol Laryngol 1982;91:628-32.
2. Zeitouni A, Manoukian J. Tracheotomy in the first year of life. J Otolaryngol 1993;22:431-4. 3. Arcand P, Granger J. Pediatric tracheostomies: changing trends. J Otolaryngol 1988;17:121-4.
4. Crysdale WS, Feldman RI, Naito K. Tracheotomies: a 10-year experience in 319 children. Ann Otol Rhinol Laryngol 1988;97:439-43.
5. Line WS, Hawkins DB, Kahlstrom EJ, MacLaughlin EF, Ensley JL. Tracheotomy in infants and young children: the changing perspective 1970-1985. Laryngoscope 1986;96:5105.
6. Palmer PM, Dutton JM, McCulloch TM, Smith RJH. Trends in the use of tracheotomy in the pediatric patient: the Iowa experience. Head Neck 1995;17:328-33.
7. Swift AC, Rogers JH. The changing indications for tracheostomy in children. J Laryngol Otol 1987;101:1258-62. 8. Waki EY, Madgy DN, Zablocki H, Belenky WM, Hotaling AJ. An analysis of the inferior based tracheal flap for pediatric tracheotomy. Int J Pediatr Otorhinolaryngol 1993;27:4754.
9. Ward RF, Jones J, Carew JF. Current trends in pediatric tracheotomy. Int J Pediatr Otorhinolaryngol 1995;32:233-9.


From the Department of Pediatric Otolaryngology, The Children's Hospital of Philadelphia and the University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania.
CORRESPONDENCE - Ralph F. Wetmore, MD, Dept of Pediatric Otolaryngology, The Children's Hospital of Philadelphia, 34th St and Civic Center Blvd, Philadelphia, PA 19104.