Comparison of two percutaneous tracheostomy techniques, guide wire dilating forceps and Ciaglia Blue Rhino: a sequential cohort study

Tracheostomy is usually performed in patients who need prolonged mechanical ventilation, frequent suctioning of bronchopulmonary toilet or have obstruction of the upper airway. The percutaneous tracheostomy is a minimally invasive, effective and reliable procedure and has become the alternative to surgical tracheostomy [1]. Almost all percutaneous procedures in The Netherlands are performed with one of the three following techniques: guide wire dilating forceps (GWDF) tracheostomy, Ciaglia Blue Rhino (CBR) tracheostomy, and sequential dilation tracheostomy (classic Ciaglia) [2]. We have extensive experience with the first two techniques [3, 4]. This study is a sequel to our previous reports. Several studies have compared different percutaneous techniques [5‐12], but because CBR is relatively new, a comparison with GWDF has been made only twice in two small prospective cohorts [5, 12]. The strength of the present study is the large group of patients, so the incidence of relevant complications is more meaningful.

The aim of this study was to compare GWDF and CBR. The study not only focuses on the immediate peri-operative complications but also describes the long-term sequelae of both techniques.

This is a retrospective analysis of all patients who underwent percutaneous tracheostomy in the University Medical Centre Nijmegen between March 1997 and April 2003. We compared the two historic data sets that we have published previously [3, 4], but we specifically focused on the precise definition of early complications and long-term sequelae. Between March 1997 and February 2000 we performed percutaneous tracheostomy on 171 patients, using the GWDF technique. Between March 2000 and April 2003 we performed percutaneous tracheostomy on a further 171 patients, using the CBR technique. Indications, contra-indications and technique for percutaneous tracheostomy are standardised [3, 4]. Patients or family gave informed consent before the procedure. Ethical approval from the institution's medical ethical committee was not obtained because the standard of care was provided and no other experimental treatments were introduced. Published data cannot be reduced to a single recognisable patient. All data were recorded prospectively on pre-designed forms. 'Procedure time' was defined as the time from incision to successful placement of the cannula. A 'peri-operative complication' was defined as a complication related to the procedure and occurring during or within 24 hours of the procedure. Postoperative complications were divided into 'complications while cannulated' and 'late complications'. A 'complication while cannulated' was defined as a complication occurring in the period between 24 hours after the procedure until removal of the cannula. A 'late complication' was defined as a complication occurring after removal of the cannula up to a follow-up of 3 years. Complications were divided into minor and major (see Tables 1, 2, 3). Moreover, complications were classified as procedure-specific and procedure-non-specific. Hypotension was defined as a systolic blood pressure of less than 90 mmHg. Hypoxaemia was defined as an arterial oxygen saturation of less than 90%. It was considered minor when lasting less than 5 min, and major when lasting 5 min or longer. Information regarding late complications was obtained by structured interviews with patients who were decannulated successfully. Patients or close relatives were asked about voice changes, dyspnoea, stridor, pain, and cosmetic problems. Patients were also asked to grade specific problems as absent, minor or major.

All data were analysed with Statistical Product and Service Solutions (SPSS) version 11.0. All variables were checked for normal distribution. Data are given as means ± SD or medians. Continuous variables were compared with Student's t-test or the Mann–Whitney test as appropriate. Bonferroni's correction for multiple comparisons was used. Categorisable variables were compared with the χ² test. A cut-off level of P < 0.05 was accepted as statistically significant.

Demographic data are shown in Table 4. The procedure was successful in 165 of 171 patients (96.5%) in the GWDF group and in 169 of 171 patients (98.8%) in the CBR group. Most tracheostomies were performed by an intensivist or a fellow (under supervision). More procedures were performed by a fellow in the CBR group than in the GWDF group (51 versus 27, respectively; P < 0.01).

In this study we have compared two different techniques of percutaneous tracheostomy, GWDF and CBR. Both techniques are frequently used in The Netherlands and are replacing the surgical technique [2]. This study showed no significant differences in clinically relevant complications between the two techniques. This is in agreement with two other studies comparing these techniques [5, 12]. Although the total number of complications in the two groups in the study of Ambesh and colleagues was not significantly different, the authors noticed an increased rate of minor peri-operative bleeding in the GDWF group [5]. This was balanced by an increase in the number of patients with one or more tracheal ring fractures in the CBR group (30%). The increase in major peri-operative bleeding with the GDWF technique might be explained by the poorly controllable dilation with the forceps [9]. Although the study of Añón and colleagues did not find any significant differences, in three of 26 patients in the GWDF group there was an inability to insert the cannula [12].

Several other studies comparing sequential dilation (classic Ciaglia) and CBR [6, 8], and comparing sequential dilation and GWDF [7, 9‐11], have been described in the literature. Van Heurn and colleagues concluded that sequential dilation and GWDF are both reliable but that sequential dilation has fewer early complications than GWDF [7]. Nates and colleagues also preferred sequential dilation to the GWDF technique, because of fewer surgical complications, less peri-operative and postoperative bleeding, and easier use [9]. Añón and colleagues found a comparable complication rate, but the procedural time of the GWDF method was significantly shorter [10]. Unfortunately, comparing these studies is difficult because complications were not defined uniformly.

In our study, a major complication while cannulated was obstruction of the cannula, which occurred in four patients. These figures correspond to the prevalence of cannula obstruction in the literature (0.3–3.5%) [13‐15]. Strict adherence to nursing protocols and a low threshold for cleaning the inner cannula should be the standard of care in the intensive care unit. An outreach team from the intensive care unit should visit patients, discharged to the general ward with a cannula in place, on a daily basis.

There are only few data available concerning late complications of percutaneous tracheostomy. Unfortunately, many confounders might be present, such as the disease process itself, the duration of endotracheal intubation, and other treatments in the intensive care unit (such as sedation or physical therapy). Moreover, both patients and caregivers often interpret late complications subjectively. The total number of late complications in our study was not significantly different between the two groups. Subjective voice changes and hoarseness were more frequent in the CBR group (P < 0.01). An explanation might be the longer mean endotracheal intubation time, because this is possibly the most important cause of voice problems. With sequential dilation tracheostomy, the incidence of voice problems ranges between 0% and 21% [16‐22]. More patients in the GWDF group complained of cosmetic problems. Only a few studies have mentioned cosmetic complaints, but differences of opinion between patient and caregiver are frequent [23]. In each group in our study, one patient developed a critical symptomatic tracheal stenosis. More patients might have had an asymptomatic tracheal stenosis, but because no additional diagnostic tests such as computed tomography or magnetic resonance imaging scans were performed, the actual incidence is unknown. Several studies have incriminated the GWDF technique as a cause of tracheal stenosis, but no studies with the CBR have been described. The incidence varied from 0% to 63% [18, 23‐27]. Most of these tracheal stenoses were asymptomatic.

Several factors might decrease the strength of our conclusions. First, the study used historical data sets with a sequential design; a time bias is therefore possible. As experience with percutaneous tracheostomy increases, the number of complications will decrease, even if another technique is used, although in our study this might well have been balanced by the fact that over time more fellows performed the procedure. Second, scoring of the peri-operative complications by different physicians might be variable because of different interpretations. Despite these shortcomings, we conclude from our study that, although the CBR technique has more minor peri-operative complications, the two techniques are comparable. More prospective, randomised studies are required to compare these different tracheostomy techniques adequately. We are currently conducting a prospective, randomised study in which we compare GWDF and CBR tracheostomies; we are specifically looking for the occurrence of precisely defined early and late complications. The occurrence of tracheal stenosis will be analysed using the forced oscillation technique and magnetic resonance imaging.