The first reported case of retrograde intubation (RI) was by Butler and Cirillo in 1960.¹ The technique involved passing a red rubber catheter cephalad through the patient’s previously existing tracheostomy. When the catheter exited the oral cavity, it was tied to the endotracheal tube (ETT), allowing it to be pulled into the trachea.

The first person to perform RI as presently practiced was Waters, a British anesthesiologist in Nigeria.² In 1963, he reported treating patients who had cancrum oris, an invasive gangrene that deforms the oral cavity, severely limiting mouth opening. His technique involved passing a standard Tuohy needle through the cricothyroid membrane (CTM) and feeding an epidural catheter cephalad into the nasopharynx. He “fished” the catheter out of the nasopharynx through the nares, using a hook he devised. The epidural catheter was then used as a stylet to guide the ETT through the nares and into the trachea.

Over the ensuing years, RI did not gain clinical acceptance because of its invasiveness and the potential for complications from the CTM puncture. After 1964, when fiberoptic technology became available, RI was irregularly but occasionally discussed in the literature.¹^–¹⁶⁵ In 1993, RI was designated as part of the anesthesiologist’s armamentarium by the American Society of Anesthesiologists (ASA) Difficult Airway (DA) Task Force.³

The term retrograde intubation, used by Butler and Cirillo, is a misnomer.⁴ The technique is actually a translaryngeal guided intubation, but for historical reasons we continue using the name retrograde intubation.

II Indications

The RI technique has been used both in the hospital setting and in prehospital mobile units (in the field).^5,⁶ It has been employed with both anticipated and unanticipated DAs^2,⁵^–²⁰; after failure to intubate by conventional means (direct laryngoscopy,^6,^9,¹¹ blind nasal intubation,^10,¹¹ bougie,^13,²¹ laryngeal mask airway [LMA], and fiberoptic laryngoscopy^18,²²^–²⁴); and in both humans and animals.^25,²⁶ In the literature, including my own experience (31 patients), there have been approximately 807 cases (670 patients and 137 cadavers) in which RI was used as a means of securing the airway. Although in most cases RI has been used to place a single-lumen ETT, one case report described placement of a double-lumen ETT through RI.⁷

A wide variety of airway diseases have necessitated RI (Box 20-1). RI has been most frequently associated with limited range of motion of the neck (153 trauma victims with potential cervical spine injury), and its use has been reported in facial trauma. It has been employed in both adults and pediatric patients with success (Box 20-2).

Box 20-1

Number of Retrograde Intubations in the Literature (539 Patients and 137 Cadavers)

Oral Cavity

Cancrum oris, 27

Mandibular or maxillary fracture, 31

Mandibular prognathia, 1

Perimandibular abscess, 1

Ankylosis of temporomandibular joint, 8

Hypopharyngeal tumors, 2

Cervical

Spinal cord injury, 98

Ankylosing spondylitis, 19

Rheumatoid arthritis, 32

Pharynx and Larynx

Laryngeal cancer, 163

Pharyngeal abscess, 1

Epiglottitis, 1

Pharyngeal edema, 1

Laryngeal edema after burn, 3

Others

Pediatric anomalies, 27

Obesity, 13

Coronary artery bypass grafting (failure to intubate), 28

Tracheostomy stoma, 16

Trauma (failure to intubate), 34

Disease not specified (failure to intubate), 21

Cadaver studies, 137

From Sanchez AF: The retrograde cookbook, Irvine, 1993, University of California, Department of Anesthesia.

Box 20-2

Characteristics of Retrograde Intubations (RIs) in the Literature

Number of adult patients (RI): 509

Number of pediatric patients (RI): 30

Age range: 1 day (weight, 2.9 kg) to 84 yr

Average time for technique: 0.2 min (range, 0.5–15 min)

Elective cases: 249 patients

Failed attempts using RI: 12 patients

From Sanchez AF: The retrograde cookbook, Irvine, 1993, University of California, Department of Anesthesia.

Not all reports have described the amount of time required to perform the technique, but in one study involving emergency medical service personnel (paramedics and registered nurses) using training mannequins, the average time was 71 seconds (range, 42 to 129 seconds).²⁷ Barriot and Riou described 13 patients with maxillofacial trauma who could not be intubated in the field using direct laryngoscopy (six attempts; average time, 18 minutes).⁵ Intubation was subsequently performed in these patients on the first RI attempt, with an average time of less than 5 minutes. An additional 6 patients were intubated in less than 5 minutes when RI was used as the initial method of choice. Slots and colleagues reported a modified technique using a Mini-Trach II set on 20 cadavers with an average time for intubation of 6.7 seconds (range, 3 to 10 seconds); it was subsequently used on an emergency basis on 3 patients with an average time of 10 seconds. The investigators concluded that the RI technique was a rapid, efficacious method for intratracheal intubation of trauma patients, especially patients with maxillofacial trauma.²⁸

Historical indications for RI are the following:

1. Failed attempts at laryngoscopy, LMA, or fiberoptic intubation (FOI)

2. Urgent establishment of an airway where visualization of the vocal cords is prevented by blood, secretions, or anatomic derangement in scenarios in which ventilation is still possible

3. Elective use when deemed necessary in clinical situations such as unstable cervical spine, maxillofacial trauma, or anatomic anomaly

III Contraindications

Contraindications to RI have been cited, often anecdotally (Box 20-3). Most are relative contraindications and can be divided into four categories: unfavorable anatomy, laryngotracheal disease, coagulopathy, and infection.

Box 20-3

Contraindications to Retrograde Intubation and Examples

Unfavorable Anatomy

Lack of access to cricothyroid muscle (severe flexion deformity of the neck)

Poor anatomic landmarks (obesity)

Pretracheal mass (thyroid goiter)

Laryngotracheal Disease

Malignancy

Stenosis

Coagulopathy

Infection (Pretracheal Abscess)

From Sanchez AF: The retrograde cookbook, Irvine, 1993, University of California, Department of Anesthesia.

A Unfavorable Anatomy

Because in most cases RI is performed above or below the cricoid cartilage, absolute lack of access to this region, as in patients with severe flexion deformity of the neck, poses a contraindication if not an impossibility.^29,³⁰ For the same reason, the patient with nonpalpable landmarks,³¹^–³³ obesity,³⁴ overlying malignancy,³² or large thyroid goiter should be approached cautiously.³² Shantha reported a case of RI in a patient with a large thyroid goiter.³⁵ After failure of conventional intubating methods (including FOI), the surgeons dissected down to the CTM and subsequently passed the catheter cephalad. Thirteen cases of RI have been reported in obese patients without major complications.^12,^18,^34,³⁶

B Laryngotracheal Disease

Theoretically, laryngotracheal stenosis may contraindicate RI because narrowing of the trachea or larynx could be made worse by either the needle puncture or the catheter.^2,³¹ However, RI has been used in patients with laryngeal cancer,³⁷ epiglottitis,³⁸ and laryngeal edema resulting from burn injuries.^16,²⁴ It should not be used if laryngeal tracheal stenosis is present directly under the intended puncture site.

C Coagulopathy

Preexisting bleeding diathesis should be considered a relative contraindication.^31,^32,³⁹^–⁴⁴ Although there is a potential for bleeding, the CTM is considered to be a relatively avascular plane (see later discussion). A small, self-limited hematoma was reported in a patient who underwent a coronary artery bypass grafting with intraoperative heparin and postoperative disseminated intravascular coagulation.⁹

D Infection

RI in the presence of preexisting infection over the puncture site or in the path of the puncture, as in pretracheal abscess or Ludwig’s angina, could result in transmittal of bacterial flora into the trachea and should be avoided. This, again, should be considered a relative contraindication, because transtracheal aspiration is performed to obtain a sputum sample in patients with pneumonia despite the possibility of pretracheal abscess (see “Complications”).^31,^32,^43,⁴⁵^–⁴⁷

IV Anatomy

The performance of RI requires basic anatomic knowledge of the cricoid cartilage (Fig. 20-1) and the structures above and below it to minimize complications and failure. Indeed, regardless of the intubation technique planned, the cricoid cartilage and CTM should be identified preoperatively in every patient.⁴⁸ Cartilage and membrane, vascular structures, and the thyroid gland are relevant anatomic structures.

Figure 20-1 Anatomy of the cricoid cartilage. Midsagittal view of the larynx and trachea.

(From Sanchez AF: The retrograde cookbook, Irvine, 1993, University of California, Department of Anesthesia.)

A Cartilage and Membrane

The cricoid cartilage has the shape of signet ring (see Fig. 20-1). It consists of a broad, flat, posterior plate called the lamina and a narrow, convex, anterior structure called the arch.^49,⁵⁰ In most cases the cartilage can be easily palpated by identifying the thyroid notch and running a finger down the midline in a caudad direction until a rigid, rounded structure is encountered. The vertical height of the arch is 0.5 to 0.7 cm (Fig. 20-2).⁵⁰ The CTM connects the superior border of the arch to the inferior border of the thyroid cartilage and measures approximately 1 cm in height and 2 cm in width.^49,⁵¹ The lateral borders are the paired cricothyroid muscles.⁴⁵ The cricotracheal ligament connects the inferior border of the arch to the upper border of the first tracheal ring and measures 0.3 to 0.6 cm in height.⁵² The distance between the inferior border of the thyroid cartilage and the vocal cords varies with gender but is approximately 0.9 cm.⁵³

Figure 20-2 Midsagittal view of the larynx and trachea showing the distance between the vocal cords and the upper border of the first tracheal ring. Only a small portion of the Murphy eye of the endotracheal tube is below the vocal cords.

(From Sanchez AF: The retrograde cookbook, Irvine, 1993, University of California, Department of Anesthesia.)

B Vascular Structures

There are paired major blood vessels above and below the cricoid cartilage: the cricothyroid artery and the superior thyroid artery (Fig. 20-3).

Figure 20-3 Vascular anatomy above and below the cricoid cartilage.

(Modified from Naumann H, editor: Head and neck surgery, Philadelphia, 1984, WB Saunders.)

The cricothyroid artery,^50,^51,⁵³^–⁵⁵ a branch of the superior thyroid artery, runs along the anterior surface of the CTM, usually close to the inferior border of the thyroid cartilage. In some cases, the cricothyroid arteries anastomose in the midline and give rise to a descending branch that feeds the middle lobe of the thyroid gland when present. On the basis of dissections that I have performed on cadavers, the cricothyroid artery becomes insignificant in size as it approaches the midline. No major venous plexus could be found mentioned in the literature, and none was found in my own dissections.

The anterior branch of the superior thyroid artery runs along the upper border of the thyroid isthmus to anastomose with its counterpart from the opposite side.^50,⁵³^–⁵⁵ The inferior thyroid artery also anastomoses with the superior thyroid artery at the level of the isthmus. The arteries are remarkable for their large size and frequent anastomoses. In fewer than 10% of the population, an unpaired thyroid artery ascends ventral to the trachea (from either the aortic arch or the brachiocephalic artery) to anastomose at the level of the isthmus. It is usually small but may be very large. A rich venous plexus is formed in and around the isthmus.

C Thyroid Gland (Isthmus, Pyramidal Lobe)

The isthmus of the thyroid gland (see Fig. 20-3) is rarely absent and generally lies anterior to the trachea between the first and fourth tracheal rings (usually between the second and third), although there are many variations. Its size and vertical height vary; the average vertical height and depth are 1.25 cm. Extending from the isthmus, the highly vascular pyramidal lobe (Fig. 20-4) is well developed in one third of the population. It is found more frequently on the left of the midline and may extend up to the hyoid bone (the upper continuation is usually thyromuscular).^25,^49,^50,^54,⁵⁵

Figure 20-4 Pyramidal lobe of the isthmus.

(Modified from Naumann H, editor: Head and neck surgery, vol 4, Philadelphia, 1984, WB Saunders.)

V Physiology

Sympathetic stress responses—increased heart rate, blood pressure, intraocular pressure, and intracranial pressure and elevated catecholamine levels—have been reported with laryngoscopy, endotracheal intubation, coughing, translaryngeal local anesthesia, laryngotracheal anesthesia, and FOI. Therefore, concern is appropriate when performing RI in patients with coronary artery disease, elevated intraocular pressure, or elevated intracranial pressure.^29,^30,⁵⁶^–⁶² It is reasonable to argue, however, that RI, performed skillfully, is not more stimulating than any other technique for managing the airway.

No apparent significant changes in hemodynamics were reported in multiple case reports of patients with cardiac disease (congenital anomalies, ischemic coronary artery disease, valvular disease, pericarditis, and congestive heart failure) who underwent RI, both awake with topical anesthesia and under general anesthesia.^9,^11,^15,^17,^38,⁶³^–⁶⁸ Casthely and colleagues reported on 25 patients with DA due to rheumatoid arthritis who underwent open heart surgery (coronary artery bypass graft and valve replacement).⁹ The patients had invasive monitors (Swan-Ganz catheters and peripheral arterial lines) placed preoperatively. The initial 24 patients underwent a cardiac induction of anesthesia (diazepam 10 mg, fentanyl 25 to 30 µg/kg, and pancuronium 0.1 mg/kg) before rigid laryngoscopy followed by RI. Comparison of hemodynamic responses to rigid laryngoscopy (Macintosh and Miller blades) versus RI demonstrated that the former approach was more stressful (Table 20-1). Patient 25 underwent RI before induction of anesthesia after application of topical anesthesia with no significant hemodynamic response.

TABLE 20-1 Hemodynamic Effects of Retrograde Intubation

	Laryngoscopy	Retrograde Intubation
HR	Increase	No change from baseline
MAP	Increase	No change from baseline
CI	Decrease	No change from baseline
PCWP	Increase	No change from baseline
ECG	3-mm ST depression	No change from baseline

CI, Cardiac index; ECG, electrocardiogram (ST segment changes in lead V5); HR, heart rate; MAP, mean arterial pressure; PCWP, pulmonary capillary wedge pressure.

Modified from Casthely PA, Landesman S, Fynaman PN: Retrograde intubation in patients undergoing open heart surgery. Can Anaesth Soc J 32:661, 1985.

Two case reports documented patients with a previous history of DA and intracranial pathology (pseudotumor cerebri and intracranial tumor with elevated intracranial pressure) who underwent elective, awake RI after topical anesthesia with no evidence of further increase in intracranial pressure.^8,¹⁴ I myself, unmedicated except for topical lidocaine, underwent awake RI with no significant hemodynamic changes.⁴⁸

VI Techniques

A Preparation

1 Positioning

The ideal position for RI is the supine sniffing position with the neck hyperextended.^69,⁷⁰ In this position, the cervical vertebrae push the trachea and cricoid cartilage anteriorly and displace the strap muscles of the neck laterally. As a result, the cricoid cartilage and the structures above and below it are easier to palpate. RI can also be performed with the patient in a sitting position,⁴⁸ which may be the only position in which some patients can breathe comfortably. Potential cervical spine injury or limited range of motion of the cervical spine may necessitate RI with the neck in a neutral position, which is a well-documented practice (see Box 20-1).

2 Skin Preparation

Although most documented RIs have not been elective, every effort should be made to perform RI using aseptic technique. Recommendations have been made for prophylactic antibiotics in diabetic or immunocompromised patients, who may be more susceptible than others to infection.⁷¹

3 Anesthesia

If time permits, the airway should be anesthetized to prevent sympathetic stimulation, laryngospasm, and discomfort. In the literature, many different combinations of techniques have been described:

1. Translaryngeal anesthesia during intravenous sedation or general anesthesia ^2,⁸^–¹⁰ ^,16 ^,18 ^,72

2. Translaryngeal anesthesia with superior laryngeal nerve block ^5,^6,⁷³

3. Translaryngeal anesthesia with topicalization of the pharynx (aerosolized or sprayed)^7,^14,^74,⁷⁵

4. Glossopharyngeal nerve block and superior laryngeal nerve block with nebulized local anesthetic ¹⁷

(Refer to Chapter 11 for a detailed description of neural blockade of the airway.)

In my own experience,⁴⁸ an awake RI can be performed using translaryngeal anesthesia (4 mL 2% lidocaine) supplemented with topicalization (nebulized or sprayed local anesthetics) of the pharynx and hypopharynx. Special caution should be exercised when performing the translaryngeal anesthesia, because coughing, grunting, sneezing, or swallowing causes the cricoid cartilage to travel cephalad, with the potential for breaking the needle in the trachea.^76,⁷⁷ To avoid this, one can insert a 20-G angiocatheter and remove the needle before injecting the local anesthetic.

4 Entry Site

The transtracheal puncture for RI can be made either above or below the cricoid cartilage. The CTM is relatively avascular and has less potential for bleeding (see “Anatomy”). The disadvantages of the CTM are that initially only 1 cm of ETT is actually placed below the vocal cords, and the angle of entry of the ETT into the trachea is more acute. An initial puncture performed at the cricotracheal ligament or lower affords the added advantage of allowing the ETT to travel in a straighter path as well as allowing a longer initial length of ETT below the vocal cords. The disadvantage is that this site (below the cricoid cartilage) has more potential for bleeding (although none has been reported). Both entry sites have been used successfully. In cadaver studies, the success rate for RI was higher with less vocal cord trauma when the cricotracheal ligament rather than the CTM was used. Vocal cord trauma has not been reported in living patients.^78,⁷⁹

B Classic Technique

The classic technique of RI is performed percutaneously using a standard 17-G Tuohy needle and epidural catheter.

After positioning (Fig. 20-5), skin preparation, and anesthesia, a right hand–dominant person should stand on the right side of the supine patient. The left hand is used to stabilize the trachea by placing the thumb and third digit on either side of the thyroid cartilage. The index finger of the left hand is used to identify the midline of the CTM and the upper border of the cricoid cartilage.

Figure 20-5 Classic technique. Midsagittal view of the head and neck.

(From Sanchez AF: The retrograde cookbook, Irvine, 1993, University of California, Department of Anesthesia.)

Because the Tuohy needle is blunt, a small incision through the skin and subcutaneous tissue with a no. 11 scalpel blade is recommended. Because of the significant force required to perforate the skin and the CTM, there is a risk of perforating the posterior tracheal wall as well. This has been verified in cadaver studies with the use of a fiberoptic bronchoscope (FOB).³⁴

The right hand then grasps the Tuohy needle and saline syringe like a pencil (using the fifth digit to brace the right hand on the patient’s lower neck) and performs the puncture, aspirating to confirm placement in the lumen of the airway (Figs. 20-6 and 20-7).

Figure 20-6 Classic technique. Standard no. 17 Tuohy needle (with saline-filled syringe) is advanced (with bevel pointing cephalad) through the cricothyroid membrane at a 90-degree angle, trying to stay as close as possible to the upper border of the cricoid cartilage. Entrance into the trachea is verified by aspiration of air.

(From Sanchez AF: The retrograde cookbook, Irvine, 1993, University of California, Department of Anesthesia.)

Figure 20-7 Classic technique. The angle of the Tuohy needle is changed to 45 degrees with the bevel pointing cephalad, again verifying position by aspirating air.

(From Sanchez AF: The retrograde cookbook, Irvine, 1993, University of California, Department of Anesthesia.)

Once the Tuohy needle is in place, the epidural catheter is advanced into the trachea (Fig. 20-8). When advancing the epidural catheter, it is important to have the tongue pulled anteriorly to prevent the catheter from coiling up in the oropharynx. The catheter usually exits on its own from either the oral (Fig. 20-9) or the nasal cavity. A hemostat should then be clamped to the catheter at the neck skin line to prevent further movement of the epidural catheter. If the catheter has to be retrieved from the oropharynx, my preferred instrument is a nerve hook (V. Mueller NL2490, Baxter, Deerfield, IL). Magill forceps have been used, but these were designed to grasp large structures such as an ETT and may not grip the relatively small catheter (the distal tips of the forceps do not completely occlude); in addition, they may traumatize the pharynx. Arya and associates described an innovative atraumatic method of retrieving catheters from the oral pharynx in patients with limited mouth opening. They used a “pharyngeal loop” that they devised from a ureteral guidewire that was threaded through a 3-mm uncuffed polyvinyl chloride ETT and doubled up to form a loop.⁸⁰

Figure 20-8 Classic technique. An epidural catheter is advanced through the vocal cords and into the pharynx. During this time, the patient is asked to stick the tongue out or tongue can be pulled out manually. Most of the time, the epidural catheter comes out of the mouth on its own. The Tuohy needle is then withdrawn to the caudal end of epidural catheter.

(From Sanchez AF: The retrograde cookbook, Irvine, 1993, University of California, Department of Anesthesia.)

Figure 20-9 Classic technique. Pull the epidural catheter out of the mouth to an appropriate length; then clamp a hemostat flush with the skin.

(From Sanchez AF: The retrograde cookbook, Irvine, 1993, University of California, Department of Anesthesia.)

Originally, the catheter was threaded through the main distal lumen (beveled portion) of the ETT. Bourke and Levesque modified the technique by threading the catheter through the Murphy eye (Fig. 20-10), reasoning that this would allow an additional 1 cm of ETT to pass through the cords.⁸¹ Lleu and coworkers,^78,⁷⁹ in cadaver studies, showed that using the cricotracheal ligament as the puncture site in combination with threading the epidural catheter through the Murphy eye enhanced success compared with the original technique.

Figure 20-10 Classic technique. Cross section of larynx and trachea with endotracheal tube (ETT) and catheter guide passing through the cricothyroid membrane. A, Catheter passes through end of ETT, and 1 cm of ETT passes the cords. B, The catheter exits the side hole, allowing 2 cm of ETT to pass beyond the vocal cords.

(From Sanchez AF: The retrograde cookbook, Irvine, 1993, University of California, Department of Anesthesia.)

When the ETT is being advanced over the epidural catheter (Figs. 20-11 through 20-13), a moderate amount of tension should be employed.^13,⁴⁰ Excessive tension pulls the ETT anteriorly, making it more likely to be caught up against the epiglottis, vallecula, or anterior commissure of the vocal cords. If there is difficulty in passing the opening of the glottis, the ETT can be rotated 90 degrees counterclockwise or exchanged for a smaller tube.^13,^23,³⁴