Tracheostomy is considered the airway management of choice in long-term mechanically ventilated intensive care unit (ICU) patients. However, conventional surgical tracheostomy is associated with a number of disadvantages and serious complications (1,2). Therefore, simpler and safer procedures for tracheostomy were studied. Besides various surgical modifications of the standard tracheostomy technique by Jackson (3), new percutaneous procedures have been introduced in the last few years. Because severe complications regarding percutaneous dilatational tracheostomy (PDT) using Ciaglia’s technique (4) are reportedly rare (5–8), PDT is considered a safe and easy technique that can readily be performed at the patient’s bedside. Therefore, it has become an appropriate alternative to conventional surgical tracheostomy and has gained worldwide acceptance (7,8). Yet, PDT is not always feasible or practicable, and problems can occur, especially in children and young adults. The increased elasticity of the anterior tracheal wall in these patients may result in total compression of the tracheal lumen during the procedure. Besides temporary problems with oxygenation, the risk of accidental traumatization of the posterior tracheal wall during puncture or dilation, possibly resulting in massive bleeding or tracheo-esophageal fistula, is also increased (9). Today, these problems can be avoided with a new percutaneous technique using a translaryngeal approach (translaryngeal tracheostomy; TLT), as described by Fantoni et al. (10).
The primary aim of our study was to evaluate the effects of PDT and TLT techniques on intra- and postoperative oxygenation and on the incidence of bleeding, aspiration, and tracheostoma infection in long-term ventilated, critically ill patients.
During a 21-mo period, we performed a prospective, nonrandomized study of 2,003 patients who had undergone thoracic and cardiovascular interventions and who were admitted to our ICU within that period. In 90 (4.5%) patients (43 male, 47 female), elective tracheostomy was performed when the patient was expected to be ventilator-dependent for 10 days or more from the date of tracheostomy. Tracheostomy was performed either according to Ciaglia’s (Patients 1–45) or Fantoni’s (Patients 46–90) techniques. The leading diagnoses for admission to the ICU and complications that result in long-term ventilation are shown in Table 1.
Infection of the tracheostomy site and severe coagulopathy were considered contraindications for tracheostomy, whereas infection of the sternotomy wound itself, even in combination with mediastinitis, was not considered a contraindication. In all cases, tracheostomy was performed at the bedside in the ICU. Anesthesia of the sedated patients was accomplished by IV administration of fentanyl (0.2 mg), propofol (3.0 mg/kg), and pancuronium (0.1 mg/kg). Before the procedure, patients were ventilated at an FIO2 of 1.0 for 5 min. After tracheostomy, the FIO2 was reduced to the preoperative level. If, under this regimen, SaO2 measured by pulse oxymetry decreased to below 92%, FIO2 was increased stepwise until the preoperative SaO2 was reached. Arterial blood gases were obtained both 1 h before and after the procedure. Throughout tracheostomy, continuous monitoring of the patients with electrocardiogram, invasive blood pressure, and pulse oximetry was performed.
Regardless of the technique used, the patient’s neck was positioned slightly extended, and the surgical area was cleaned and prepared with drapes. In the case of PDT, the Ciaglia Percutaneous Tracheostomy Introducer Set (Cook Critical Care, Bjaeverskov, Denmark) was used. First, the endotracheal tube in place was pulled back under bronchoscopic control so that the tracheal lumen could be punctured without problems. When the lumen was clearly identified, the trachea was punctured between the second and third tracheal ring, using a 14-gauge Teflon® (DuPont, Wilmington, DE) cannula. Correct position of the cannula in the middle of the trachea was confirmed bronchoscopically. A guidewire was then introduced through the cannula and armed with a thin, synthetic catheter. Thereafter, a 1.5-cm transverse skin incision was made, and, once the trachea could be identified, it was dilated to 36F, using Seldinger’s technique. The tracheostomy tube was then inserted into the trachea over a smaller dilator. The tubes used were the 9.0-mm (n= 14) and 10.0-mm (n= 31) Rueschelit Ultra (Ruesch, Kernen, Germany).
When TLT was performed, the Translaryngeal Tracheostomy Kit (Mallinckrodt, Mirandola, Italy) and tracheal cannulas of 9.5-mm internal diameter were used. After careful suctioning of the oropharynx, the cuff of the in-place endotracheal tube was deflated, and the tube was pulled back under laryngoscopic view until the cuff could be visualized just below the glottis. The cuff was then reinflated with 4–5 mL of air, and the tube was secured in this position. Puncture of the trachea between the second and third tracheal cartilage was done under bronchoscopic control. The guidewire was then passed through the introducer needle and moved retrograde toward the oropharynx, parallel to the tube. When the guidewire was caught in the patient’s mouth, its cranial tip was connected to the pointed end of the conical tracheal cannula. The patient was then tracheally extubated and reintubated orally with the set’s 5.0-mm internal diameter endotracheal tube. The tube’s cuff was positioned right above the carina, distal to the puncture site. By pulling on the caudal end of the guidewire and by digital compression, the tracheal cannula was advanced through pharynx and larynx into the trachea. Once inside the trachea, a small skin incision was made at the initial puncture site, and the pointed tip of the cannula was advanced through the anterior tracheal wall. The cannula was passed forward through the skin until the markings on the cannula showed that its intratracheal end was just 3 cm below the skin level. Thereafter, the pointed tip of the cannula was cut off, the thin endotracheal tube was removed, and the cannula was rotated 180° by means of an obturator. It was then connected to the respirator.
Correct position of the tracheostomy tube was confirmed in both PDT and TLT by bronchoscopy and auscultation of the lungs. Regardless of the technique, every procedure was terminated by one final bronchoscopy. Any aspired blood and secretions were suctioned off.
Bleeding of more than 10 mL and aspiration of blood and/or secretions requiring bronchoscopic suctioning were considered complications. Obstruction of a segmental bronchus was defined as mild, whereas obstruction of a main bronchus with concomitant decrease in SaO2 below 85% was considered severe aspiration. Overall oxygenation before and after the procedure was evaluated in terms PaO2/FIO2 ratio, where a decrease of more than 20% was considered mild, and that of more than 40% severe. Pre- and postoperative PaCO2 levels were determined.
Microbiological samples were taken from the wound edges every time the tracheostomy tube was replaced. Macroscopically visible signs of infection, such as rubor or induration, were considered an infection of the tracheostoma, as were growth of pathogenic bacteria and/or fungus from the samples. Growth of apathogenic bacteria from the skin was not considered significant when there were no macroscopic signs of infection and when laboratory markers (white blood cell count, C-reactive protein, procalcitonin) were unresponsive.
Once the homogeneity of the data was confirmed, the Wilcoxon-Mann-Whitney test was used to compare the data in terms of mean and SD, whereas the χ2 test was used to compare contingencies. All statistical calculations were performed using GraphPad InStat® Version 3.00 (GraphPad Software, Inc., San Diego, CA). Statistical significance was confirmed with a P value < 0.05.
A total of 90 long-term ventilated cardiosurgical ICU patients received elective tracheostomy for 21 mo. Tracheostomy was performed in 45 patients each by the minimally-invasive PDT and TLT techniques at the patient’s bedside.
Before tracheostomy, the patients were ventilated via an oral (n= 72) or nasal (n= 18) tube for 12.1 ± 9.1 days in PDT and 9.2 ± 8.0 days in TLT (P is not significant). Regardless of the technique performed, the PaO2/FIO2 ratio slightly decreased, and the postoperative ratio was significantly lower in PDT as compared with TLT, whereas the preoperative ratios were not significantly different (see Table 2). After PDT, in 12 (26.7%) of the patients, the postoperative PaO2/FIO2 ratio was more than 20% below the preoperative value, which was identical to the TLT group. A severe decrease of more than 40% was noted in nine (20.0%) of the patients in the PDT group, whereas such a decrease occurred in only three (6.7%) of the TLT patients (P= ns). Despite temporary deterioration of the PaO2/FIO2 ratio that was noted regardless of the technique used, no life-threatening hypoxia was noted during the procedures (e.g., cyanose, bradycardia, or arterial hypotension). With regard to arterial carbon dioxide tension, we noted a significantly increased PaCO2 after TLT when compared with the preoperative levels, whereas during PDT PaCO2 remained almost stable.
With regard to complications of tracheostomy itself, the cumulative rate was 12.5% in PDT. In three patients, the final bronchoscopy showed mild aspiration of bloody secretions, which were suctioned off via the bronchoscope. Severe oxygen desaturation caused by massive aspiration of bloody secretions occurred in one patient who received PDT, but without adverse sequelae because of its short duration. In another patient, the skin incision made for PDT was too large, and bleeding of about 80 mL required surgical intervention to adapt the wound edges. In those patients who received TLT, initial problems occurred in 14 (31.1%) patients regarding retrograde guidewire advancement toward the pharynx. In most of these cases, it took several trials until the guidewire’s oropharyngeal end was placed in the correct position, and, in one patient with a short neck and a goiter who had already initially been intubated bronchoscopically, the guidewire did not reach the oropharynx and could not be caught elsewhere in the patient’s airways with Magill’s forceps, because the glottis, vocal cords, and trachea could not be revealed under direct laryngoscopy. We had to convert to PDT, which was performed without complications. When the guidewire was pulled and rotated simultaneously, it tore in another patient. No clinical signs of infection were noted at the tracheostoma, and microbiological investigations never revealed contamination with pathological bacteria or fungus, in both PDT and TLT.
Presuming that the patients had not died, been decannulated, or been discharged elsewhere, the first elective exchange of the tracheostomy tube was performed 11.1 ± 2.1 (11.4 ± 1.4) days after PDT (TLT) in 21 (46.7%) of the patients who received PDT and 25 (55.6%) of those in which TLT was used. There were no problems during this elective procedure, however, in one case of an accidental tube dislocation two days after PDT, an emergency tube exchange was performed, and a mediastinal emphysema occurred, which we attribute to the prolonged and difficult procedure of tube reinsertion.
The overall survival rates were 62.2% in the PDT group and 46.7% in the TLT group (P is not significant). Regardless of the technique used for tracheostomy, no deaths could be attributed to perioperative complications of the tracheostomy procedures. On the basis of clinical findings, none of the patients discharged from our hospital after decannulation (n= 42 in total) developed stridor or tracheal stenosis. In a patient who required PDT twice during his stay in the hospital, no tracheal stenosis was noted either clinically or under bronchoscopic control 6 mo after final decannulation. Unacceptable scars were not noted in any patient, regardless of the technique used for tracheostomy.
Since 1994, percutaneous tracheostomy has been recognized as a safe alternative to the conventional surgical technique and, thus widely established in intensive care medicine (5–7,9,11). Recently, Ciaglia’s PDT technique (4) is most preferred, and the incidence of complications after conventional surgical tracheostomy could be decreased. In the meantime, the technique has been modified, principally by bronchoscopic guidance when the trachea is punctured and dilated (12–14). As a result, severe complications caused by accidental traumatization of the posterior tracheal wall during “blind” puncture and dilation have become rare. In 1997, Fantoni et al. (10) introduced their technique of TLT. TLT obviates the need of external dilation of the trachea, and thus limiting the dangers of compression of the tracheal lumen and injury of the posterior tracheal wall. Thus, TLT is also feasible in children and young adults with highly elastic tracheas (10). Another advantage is the reduction of cervical soft tissue injury during cannula placement as a result of digital stabilization of the anterior tracheal wall.
In our patients who received PDT, the rate of complications was 12.5% and within the range of complications reported in a number of other clinical trials that investigated bronchoscopically controlled PDT (5,12–14). During TLT, besides minor technical difficulties involving retrograde guidewire advancement, one major technical difficulty was noted, requiring TLT abortion and conversion to PDT, because the guidewire could not be caught or seen in the patient’s airway. However, the patient’s trachea was already extremely difficult to intubate. This complication shows that previously known or expected difficulties regarding endotracheal intubation should be strictly considered a contraindication for TLT, because the endotracheal tube has to be changed during the procedure. Furthermore, during endotracheal tube exchange, regurgitation and even massive aspiration may occur. Therefore, we feel that it is important to emphasize not useing TLT in patients who area at increased risk of aspiration (e.g., extreme obesity, ileus), and that enteral feeding in ICU patients should be stopped at least 4 hours before the procedure.
Because TLT is a relatively new method for tracheostomy, data with regard to this procedure are rarely available in the recent literature. Walz et al. (15,16) studied a total of 50 patients who underwent TLT and noted complications in 3 patients (6%), which consisted of paratracheal misplacement of the cannula in one patient, infection of the tracheostoma in another, and complete pull of the cannula through the skin in the third patient, making it necessary to perform PDT. Fantoni et al., who investigated 109 patients, including 14 neonates and children, reported an overall complication rate of 5.5% (10). Concerns have been raised that, in TLT, dislocation of bacteria from the oropharynx into the trachea during cannula advancement may increase the risk of stoma infections (17). Yet, in our patients, infection was not noted clinically or revealed by microbiological testing in either PDT or TLT.
Median sternotomy has often been considered a contraindication for tracheostomy as a result of the risks of sternal infection or mediastinitis, and cricothyroidotomy has been suggested as an alternative to minimize infection hazards (18,19). However, Jackson (20) showed that the risk for posttracheostomy tracheal stricture is directly related to the position of tracheostomy tube placement and is lower, as insertion is done more distally to the cricoid cartilage tube. Regardless, Brantigan and Grow (19) demonstrated that, after cricothyroidotomy, the incidence of late laryngotracheal injury was low, and it is widely recommended for gaining tracheal access more distally. Furthermore, recent studies have shown that, even if distal percutaneous tracheostomy is performed within the first few days after median sternotomy, the risks for sternal infection or mediastinitis are not increased when compared with late tracheostomy when the sternotomy has virtually healed (22). Likewise, in our patients, we saw no increased incidence of sternal infections or mediastinitis compared with those patients who received later tracheostomy.
Despite high concentrations of inspired oxygen, arterial oxygen tension of critically ill patients with severe respiratory insufficiency is often not satisfactory. Because gas exchange is usually impaired during airway manipulations, any manipulations should be held as short as possible. To perform PDT or TLT, much less time is needed than for conventional tracheostomy, and airway compromise can be minimized (11). Although TLT allows maintenance of oxygenation via the thin endotracheal tube almost throughout the procedure, postoperative PaCO2 levels were significantly higher than preoperatively, whereas during PDT PaCO2 did not increase significantly. Because during both PDT and TLT a bronchoscope was introduced into the endotracheal tube in place, the rise of arterial carbon dioxide tension during TLT can probably be attributed to ventilation via the thin replacement tube. Therefore, TLT should be considered carefully for patients with increased intracerebral pressure to avoid PaCO2 levels rising throughout tracheostomy (16). In contrast, for patients with respiratory failure, TLT should be preferred, because a severe decrease of the PaO2/FIO2 ratio of 40% and greater was not as likely during TLT as during PDT. These findings can be attributed to the fact that during PDT the tracheal lumen is repeatedly obstructed during dilation from external pressure and the dilators itself. In contrast, TLT is distinguished by the fact that no pressure is applied to the trachea from the outside. Also, with the exception of the puncture needle and the guidewire, no other instruments need to be introduced, and, hence, TLT does not significantly interfere with mechanical ventilation.
The initial exchange of the tracheostomy tube is generally no problem when conventional surgical tracheostomy has been performed, even within the first few days after the procedure. However, when percutaneous techniques have been used, the exchange of tubes is at least initially more difficult. With PDT and TLT, the original tube should be left in place for at least seven days (23), because the tracheostoma is dilated only, and within the first few days after PDT or TLT, the tissue will contract to close the tracheostoma once the tube is removed. Immediate reinsertion of the tracheal cannula is virtually impossible under these conditions, and severe, sometimes fatal, complications during the procedure have been reported (24,25). Thus, cannula replacement was not scheduled before the ninth day after tracheostomy. In one patient, tube dislocation two days after PDT required an emergency exchange that was difficult and prolonged. Mediastinal emphysema developed, but fortunately resolved without sequelae. We therefore believe that it is important to first translaryngeally reintubate the patient when an unscheduled tracheostomy tube exchange is necessary within the first few days after any percutaneous tracheostomy. Airway control via a translaryngeal tube must always be established first, and only then should semi-elective cannula exchange with a dilator be attempted (25).
According to our data, PDT and TLT are safe procedures and unlikely to result in major complications when precautions are carefully observed. PDT and TLT, therefore, are attractive techniques when elective tracheostomy in critically ill patients is required.
1. Stauffer JL, Olson DE, Petty TL. Complications and consequences of endotracheal intubation and tracheostomy:: a prospective study of 150 critically ill adult patients. Am J Med 1981; 70: 65–76.
2. Heffner JE, Miller KS, Sahn SA. Tracheostomy in the intensive care unit. Part 2. Complications. Chest 1986; 90: 430–6.
3. Jackson C. Tracheostomy. Laryngoscope 1909; 19: 285–7.
4. Ciaglia P, Firsching R, Syniec C. Elective percutaneous dilatational tracheostomy: : a new simple bedside procedure. preliminary report. Chest 1985; 87: 715–9.
5. Winkler WB, Karnik R, Seelmann O, et al. Bedside percutaneous dilatational tracheostomy with endoscopic guidance: : experience with 71 ICU patients. Intensive Care Med 1994; 20: 476–9.
6. Toursarkissian B, Zweng TN, Kearney PA, et al. Percutaneous dilatational tracheostomy:: report of 141 cases. Ann Thorac Surg 1994; 57: 862–7.
7. Friedman Y, Fildes J, Mizock B, et al. Comparison of percutaneous and surgical tracheostomies. Chest 1996; 110: 480–5.
8. Graham JS, Mulloy RH, Sutherland FR, Rose S. Percutaneous versus open tracheostomy: : a retrospective cohort outcome study. J Trauma 1996; 41: 245–50.
9. Walz MK, Peitgen K, Thürauf N, et al. Percutaneous dilatational tracheostomy:: early results and long-term outcome of 326 critically ill patients. Intensive Care Med 1998; 24: 685–90.
10. Fantoni A, Ripamonti D. A non-derivative, non-surgical tracheostomy: : the translaryngeal method. Intensive Care Med 1997; 23: 386–92.
11. Holdgaard HO, Pedersen J, Jensen RH, et al. Percutaneous dilatational tracheostomy versus conventional surgical tracheostomy. Acta Anaesthesiol Scand 1998; 42: 545–50.
12. Barba CA, Angood PB, Kauder DR, et al. Bronchoscopic guidance makes percutaneous tracheostomy a safe, cost-effective and easy-to-teach procedure. Surgery 1997; 118: 879–83.
13. Marelli D, Paul A, Manolidis S, et al. Endoscopic guided percutaneous tracheostomy: : early results of a consecutive trial. J Trauma 1990; 30: 433–5.
14. Paul A, Marelli D, Chiu RCJ, et al. Percutaneous endoscopic tracheostomy. Ann Thorac Surg 1989; 47: 314–5.
15. Walz MK, Hellinger A, Walz MV, et al. Translaryngeal tracheostomy:: technique and initial results. Chirurg 1997; 68: 531–5.
16. Walz MK, Peitgen K. Puncture tracheostomy versus translaryngeal tracheostomy::. a prospective randomized study of 50 intensive care patients. Chirurg 1998; 69: 418–22.
17. Mazzon D, Paolin A, Vigneri M. Peristomal infection after translaryngeal tracheostomy: : a risk linked to the colonization of the oropharynx [letter]? Intensive Care Med 1998; 24: 278–9.
18. Pierce WS, Tyers GF, Waldhausen JA. Effective isolation of a tracheostomy from a median sternotomy wound. J Thorac Cardiovasc Surg 1973; 66: 841–2.
19. Brantigan CO, Grow JB. Cricothyroidotomy:: elective use in respiratory problems requiring tracheotomy. J Thorac Cardiovasc Surg 1976; 71: 72–81.
20. Jackson C. High tracheostomy and other errors: : the chief cause of chronic laryngeal stenosis. Surg Gynecol Obstet 1921; 32: 392.
21. Ciaglia P, Graniero KD. Percutaneous dilatational tracheostomy: : results and long-term follow-up. Chest 1992; 101: 464–7.
22. Hübner N, Rees W, Böckelmann M, et al. Percutaneous dilatational tracheostomy done early after cardiac surgery: : outcome and incidence of mediastinitis. Thorac Cardiovasc Surg 1998; 46: 89–92.
23. Heuer B, Deller A. Early and long-term results of percutaneous dilatational tracheostomy (PDTCiaglia
) in 195 intensive care patients. Anaesthesiol Intensivmed Notfallmed Schmerzther 1998; 33: 306–12.
24. Walz MK, Eigler FW. Percutaneous dilatational tracheostomy: : operative technique. Chirurg 1996; 67: 436–43.
25. Bause H, Prause A, Schulte am Esch J. Indications and technique of percutaneous dilatation tracheostomy for intensive care patients. Anaesthesiol Intensivmed Notfallmed Schmerzther 1995; 30: 492–6.