This study was approved by the Institutional Review Board of Dong-A University Hospital (IRB No. DAUHIRB-20-036). The Board waived the requirement for informed consent due to the retrospective study design.

This was a retrospective observational study of patients who underwent PDT without endotracheal guidance at a single tertiary hospital from January 1 to December 31, 2018. Patients who underwent PDT were initially screened, and the corresponding electronic medical records, including procedure notes, nurse notes, vital signs, and physician progression notes, were manually reviewed.

Patients ≥18 years of age were included if they underwent PDT without endotracheal guidance, which was an almost routine practice for tracheostomy during the period. Patients were excluded if PDT was performed by other physicians or by endotracheal guidance (bronchoscopy, video laryngoscopy, ultrasound, etc.) and if they initially underwent surgical tracheostomy.

The primary outcome was defined as successful PDT. PDT failure was defined as reintubation or airway management using a supraglottic airway device (SAD) for any reason or conversion to surgical tracheostomy after the initiation of PDT. Secondary outcomes were defined as immediate (within 24 hours) and delayed (within 7 days) procedure-related complications, such as hypoxia, bleeding and hypotension within 24 hours after PDT and delayed problems associated with tracheostomy within 7 days after PDT. Hypotension was defined as a systolic blood pressure of 70 mm Hg or less; major bleeding was defined as bleeding requiring an aggressive intervention, such as embolization, electrocoagulation, and wound revision; and minor bleeding was defined as hemorrhage requiring direct compression at the procedure site within an hour. Hypoxia was defined an oxygen saturation on pulse oximetry of 88%. Survival and discharge were also reviewed.

PDT was routinely performed by a single medical intensivist using the Griggs technique PDT kit (Portex Percutaneous Tracheostomy Kit; Smith Medical, Kent, England) without any endotracheal guidance. The physician in this study, as a medical intensivist, had experience with more than 100 cases of PDT for more than 3 years and performed PDT with bronchoscopic endotracheal guidance in 20 cases and without endotracheal guidance in 52 cases in the previous 1 year.

The PDT procedures were routinely performed as follows. The patient was prepared by the administration of analgesics and sedatives and was placed in the supine position with the neck hyperextended. Continuous remifentanil infusion from 0.06 µg/kg/hr to 0.14 µg/kg/hr was used for analgesia, and bolus 5 mg of midazolam was used for procedural sedation. A neuromuscular blockade agent was not used during PDT. Blood pressure, heart rate and pulse oximetry were continuously monitored. The neck of the patient was widely disinfected with chlorhexidine gluconate solution and covered with sterile drapes for the aseptic procedure. The operator palpated the neck carefully from top to bottom to identify two important landmarks: the cricothyroid membrane and the sternal notch. The incision site was placed under 1 fingerbreadth from the cricothyroid membrane, targeting between the 2nd and 3rd tracheal cartilaginous rings. The sternal notch was a landmark warning not to proceed below this line. Local anesthesia with 2% lidocaine and 1:100,000 epinephrine was given subcutaneously at the incision site. After making an approximately 2-cm transverse skin incision, dissection of the pretracheal soft tissue was performed. The assistant changed the fraction of inspired oxygen setting to 100% and withdrew the endotracheal tube to 15 cm on the incisor while the operator palpated the tracheal membrane at the insertion site. Puncture of the trachea was cautiously performed on the midline of the incision, and the endotracheal position was confirmed by air bubble aspiration of a syringe containing 0.9% normal saline. Then, a guide wire was inserted, followed by removal of the needle. The initial stoma was made using a dilator, followed by second stoma formation with dilating forceps by the Griggs technique. Then, the tracheostomy cannula was placed using the Seldinger technique along the guide wire, and the patient was totally extubated by the assistant. The proper endotracheal position of the tracheostomy tube confirmed the smooth passage of the rubber suction tip into the cannula. During the procedure, most events concerning the patient were recorded on the electronic medical record system. All patients were confirmed by chest X-rays, and vital signs were recorded after the procedure.

The data were analyzed using IBM SPSS ver. 21.0 (IBM Corp., Armonk, NY, USA) for Windows. Nominal variables were examined by the Fisher’s exact test, and continuous variables were examined by the Shapiro-Wilk and the Kolmogorov-Smirnov test for assessing the normality of the distribution. If the test result was significant (P<0.05), the sample data distribution was considered non-normal. Over half of the continuous variables followed a normal distribution and are presented as the mean and standard variation. The other continuous variables followed a non-normal distribution and are presented as the median (interquartile range). Nominal variables are presented as the number and incidence rate (%). Values were rounded off to one decimal place.

From January 1 to December 31, 2018, 78 patients underwent PDT in the ICU and coronary ICU. The mean age was 71.9±11.5 years, and 29 patients (37.2%) were female. More than half of the patients were admitted to pulmonology (46 patients, 59%), followed by cardiology (7 patients, 9%), nephrology (5 patients, 6.4%), and others (16 patients, 20.5%). The mean Acute Physiology and Chronic Health Evaluation (APACHE) II score was 25.9±5.8, and 71 patients (91%) had received mechanical ventilation during their whole hospitalization. Fifty patients (64.1%) had been on mechanical ventilation during PDT; ventilator modes were synchronized intermittent mandatory ventilation (35 patients, 44.9%), pressure support ventilation (9 patients, 11.5%), and controlled mandatory ventilation (6 patients, 7.7%). Their mean FiO₂ was 40.0% (30.0–40.0), and the duration from mechanical ventilation to PDT was 14.0±7.5 days. Seventeen patients (21.8%) needed vasopressors to maintain their blood pressure. Thirty patients (38.5%) had taken anticoagulants or antiplatelets within 7 days before PDT, and their mean duration was 2.0 days (2.0–3.3 days): aspirin (9 patients, 11.5%), clopidogrel (9 patients, 11.5%), enoxaparin (14 patients, 17.9%), heparin (3 patients, 3.8%), warfarin (3 patients, 3.8%), and dalteparin (1 patient, 1.3%). Detailed baseline characteristics of the patients are shown in Table 1.

The primary outcome of the study was successful PDT. Among the 78 study patients, successful PDT was achieved in 74 (94.9%) at the first cannulation attempt. Four patients experienced initial cannulation failure. In two of these patients, a SAD was used to maintain airway patency, followed by successful PDT. The other two patients were reintubated and converted to surgical tracheostomy. All of the patients recovered completely after PDT or surgical tracheostomy without any hypoxia-related sequelae.

In the assessment of secondary outcomes, hypotension was the most common complication (16 patients, 20.5%); eight patients (10.3%) required crystalloid replacement only, four patients (5.1%) required vasopressors, and four patients (5.1%) required both crystalloid replacement and vasopressors. Minor bleeding at the tracheostomy site occurred in five patients (6.4%), and they needed only supportive care, such as direct manual compression or simple dressing. Major bleeding requiring surgical intervention or transfusion was not reported (Table 2). There were no early complications concerning the airway requiring therapeutic procedures or interventions within a week after the procedure.

After PDT, 47 patients (60.3%) were transferred to another hospital, five patients (6.4%) were discharged home, and 26 patients (33.3%) died. The mean time from the day of PDT to death was 23.0 days (11.3–67.0 days) (Table 3).

Clinical Characteristics of Patients with PDT Failure on the First Attempt

Four patients failed at the first attempt of PDT due to hypoxia. Clinical characteristics of patients with initial PDT failure are described in Table 4. Four patients experienced failed weaning from mechanical ventilation for various reasons and underwent PDT after different durations of mechanical ventilation. In the first case of failure, a 65-year-old woman with non-small-cell lung carcinoma needed mechanical ventilation due to consistent atelectasis and pneumonia. The 1st PDT attempt failed, and the SpO₂ dropped to 67% during the 1st attempt. A SAD was inserted for oxygenation, and the 2nd PDT attempt was successful. In the second case, a 59-year-old man with olivopontocerebellar atrophy required long-term mechanical ventilation due to postoperative complications of gastric ulcer perforation. After the 1st PDT attempt, the mechanical ventilator was connected, but the tidal volume was not received from the mechanical ventilator due to tracheostomy site obstruction. The patient’s SpO₂ dropped to 76%, and a SAD was inserted immediately; the 2nd PDT attempt was successful. In the third case, a 48-year-old woman with alcoholic liver cirrhosis failed to wean from mechanical ventilation after liver transplantation. The 1st PDT attempt failed, and the SpO₂ dropped to 75%. The patient was reintubated and underwent conversion to surgical tracheostomy. Lastly, an 84-year-old man with hypertension and diabetes mellitus underwent PDT because of poor airway protection after spinal stenosis surgery. The 1st PDT attempt failed, and the SpO₂ dropped to 79%. A SAD was inserted, and the subsequent 2nd attempt failed. This patient also underwent conversion to elective surgical tracheostomy.