Cricoid Pressure Displaces the Esophagus: An Observational Study Using Magnetic Resonance Imaging

After obtaining institutional review board approval and informed consent, we studied 22 healthy volunteers aged 16 yr or older who did not have contraindications to CP. Sagittal views of the neck without CP and axial views of the neck with and without CP were obtained with the subject's head positioned neutrally in the cervical, thoracic, and lumbar spine coil on the MRI table. Twenty to 30 N of force (2 to 3 kg) was applied to the subject's cricoid. To minimize variability in the technique, a single investigator (K.J.S.) applied CP to all subjects, and the applied force was standardized by reproducing 2 kg on a weigh scale prior to each application of CP. This method has been used to reproduce accurate amounts of CP previously. 4,5Because of spatial limitations of the MRI machine, the investigator applied CP using a two-handed technique while positioned at the head of the MRI table. All subjects were asked whether the CP “felt midline” prior to imaging to maximize the reliability of the technique within the confines of the MRI machine. All imaging examinations were performed on a 1.5-T magnet (Twinspeed LX; GE Medical Systems, Milwaukee, WI) with sagittal and axial T1- and T2-weighted sequences. The images were taken independent of the respiratory phase as the dimensions of the structures of interest (cricoid, esophagus, and vertebral body) did not change significantly with the phase of respiration. The images were reviewed in digital format (GE PACS, Pathspeed 8.1; GE Medical Systems).

Images were presented to the participating radiologists (J.D. and G.Y.) for measurements in a blinded fashion to minimize bias. Using standard PACS measuring tools, measurements were taken from the most cephalad image in which the esophagus was seen clearly. The transverse diameters of the cricoid, the esophagus, and the vertebral body were defined as the transverse distance between the structure's right and left borders, which were demarcated with lines drawn along the anteroposterior axis using the PACS measuring tools. Similarly, the anteroposterior diameter of the cricoid was defined as the anteroposterior distance between the cricoid's anterior and posterior borders, which were demarcated with lines drawn along the transverse axis. The midline of each structure was defined as the line intersecting the midpoint of the transverse diameter along the anteroposterior axis (fig. 1). Displacement of the cricoid and displacement of the esophagus were defined as the lateral displacement of the midlines of these structures relative to the midline of the vertebral body. Unopposed esophageal tissue was defined as any part of the esophagus, along its transverse axis, that was not opposed by both the cricoid and the vertebral body (fig. 1). The dimensions of the esophagus and the airway, the length of soft tissue between the airway and the vertebral body, the amount of esophageal displacement relative to the midline of the vertebral body, the transverse distance between cricoid and esophageal midlines, the anteroposterior distance from the cricoid to the vertebral body, the amount of unopposed esophageal tissue, and the amount of laryngeal displacement relative to the midline of the vertebral body, with and without CP, were measured (fig. 1).

Each radiologist read the images independently. When disagreements occurred between the radiologists regarding the image to be used for measurements, we used the following rules. If both the cricoid and the esophagus were seen in both images, the more cephalad image was chosen. If both the cricoid and the esophagus were seen only in one of the two images, the image with both structures was chosen. If no image showed both the cricoid and the esophagus, the image showing the most cephalad portion of the esophagus was chosen.

Our primary objective was to qualitatively describe the anatomic changes in the neck with CP. A convenience sample of 22 subjects was chosen based on economic considerations. In the event of a zero incidence of esophageal displacement during CP in 22 subjects, the maximum risk of esophageal displacement in the entire population would be 13% (upper 95% CI). 10As 49% (95% CI, 35.3–62.7%) of patients had displacement of the esophagus in the absence of CP in our previous study using computed tomography, 7we thought that 22 patients would be adequate. Frequencies were expressed as proportions and percentages. For the extent of esophageal and laryngeal displacement, the measures of central tendency and dispersion were expressed as the mean and SD, respectively. We used the chi-square statistic to compare dichotomous measures with and without CP (e.g. , frequency of esophageal displacement), and we used the paired t  test to compare continuous measures with and without CP (e.g. , length of esophageal displacement). As these multiple comparisons were secondary outcomes, we applied the Bonferroni correction;P < 0.01 was considered statistically significant.

One subject did not complete the study. Table 1summarizes the demographics and relevant medical history of the remaining 21 subjects. For two subjects without CP, the MRI scan did not extend sufficiently caudad to view the esophagus; however, the cricoid and cricopharyngeal muscle were present, allowing for all measurements except esophageal dimensions and degree of displacement from the midline. The esophagus was observed in all subjects with CP.

There was 25% and 50% agreement on images selected with and without CP, respectively. The main sources of disagreement were difficulties in determining the origin of the esophagus and whether the structure was the esophagus or the cricopharyngeal muscle in an image, despite the high resolution of the MRI scans. To determine the interrater agreement for measurements, we defined an allowable measurement error of 2.0 mm, assuming an average esophageal diameter of approximately 15.0 mm, as a representative reproducible measurement based on a study demonstrating up to 15% measurement variability using MRI. 11Of the images selected identically by both radiologists, there was 80% and 70% agreement in the measurements for esophageal diameter with and without CP, respectively.

The average position of the cricoid was at the C6–7 intervertebral level (median, C6–7; range, C5 vertebral body to C7–T1 intervertebral level). Of the 19 subjects whose MRI images of the esophagus were available, the esophagus was clearly present at the level of the cricoid in only one subject. In the remaining 18 subjects, the cricopharyngeal muscle was at the level of the cricoid, which was 10.3 ± 4.5 mm superior to the esophagus.

Table 2summarizes the esophageal dimensions and related measurements. The esophagus was displaced, relative to the midline of the vertebral body, in 10 (52.6%) of 19 subjects without CP and 19 (90.5%) of 21 subjects with CP. In the 19 subjects who had images with and without CP available for comparison, CP shifted the esophagus relative to its initial position to the left in 13 subjects (68.4%) and to the right in four subjects (21.1%). In the presence of CP, the esophageal transverse diameter increased 1.0 ± 4.9 mm (P = 0.41;table 2), and the esophageal midline moved 3.5 ± 5.4 mm to the left, relative to the midline of the vertebral body (P = 0.013;table 2). The esophagus was not completely opposed between the cricoid and the vertebral body in 9 (47.4%) of 19 subjects without CP and in 15 (71.4%) of 21 subjects with CP (fig. 2). Greater than 40% of the esophageal diameter was unopposed in 1 (5.3%) of 19 subjects without CP and 7 (33.3%) of 21 subjects with CP. Figure 3illustrates leftward displacement of the esophagus with the application of CP.

The airway was displaced, relative to the midline of the vertebral body, in 7 (33.3%) of 21 subjects without CP and 14 (66.7%) of 21 subjects with CP. Airway compression, defined as a decrease in the anteroposterior diameter by at least 1.0 mm, was seen in 17 CP subjects (81.0%). CP significantly reduced the anteroposterior diameter by 4.0 ± 4.4 mm (P = 0.001), displaced the airway to the left by 3.4 ± 3.9 mm (P = 0.004), and reduced the distance between the cricoid and the vertebral body by 2.3 ± 3.5 mm (P = 0.008) (table 2).

The cricoid and esophagus moved apart by 1.0 ± 6.6 mm with CP. The transverse esophageal diameter and the transverse airway diameter increased 1.0 ± 4.9 mm and 0.3 ± 2.2 mm, respectively, with CP. All these changes were not statistically significant.

Despite all 21 subjects reporting the perception of midline compression, CP was asymmetrically applied in 9 (42.9%) of them with more pressure applied on the right side of the cricoid in 7 of the 9 subjects. Seven of the 17 subjects who were physicians described CP as more painful than expected. The remaining 10 physician subjects and all 4 nonphysician subjects found CP to be tolerable.

This study demonstrated that in the absence of CP, the esophagus was lateral to the cricoid in more than 50% of the population. The two-handed technique of CP was associated with a 1.5- and 3-fold increase in the frequency and the degree of lateral esophageal displacement, respectively, as well as a 1.5-fold increase in the frequency of esophagus unopposed between the airway and vertebral body, possibly providing a clear passage for regurgitation of gastric contents.

The mechanism for CP assumes that the cricoid, esophagus, and vertebral body are juxtaposed along the axial plane. Our results challenge this assumption. The anatomic structure lying posterior to the cricoid in nearly 95% of our subjects was the cricopharyngeal muscle. Together with the middle and inferior constrictor muscles, this muscle forms the inferior structure of the laryngopharynx, defined as the part of the pharynx that lies posterior to the larynx. 12It is a funnel-shaped structure beginning at the superior border of the epiglottis and extending to the inferior border of the cricoid, where it becomes continuous with the esophagus.

On average, the esophagus was located 10.3 ± 4.5 mm inferior to the cricoid. Furthermore, the esophagus was displaced laterally in more than one half of the subjects, which is similar to the results seen in previous studies of normal necks without CP using computed tomography and MRI. 7,13Below the level of the cricoid, the esophagus is expected to travel to the left of midline. 12 

Although the esophagus proper may not lie directly behind the cricoid, the laryngopharynx also serves as a conduit between the oropharynx and the gastric contents; therefore, compression of the laryngopharynx may prevent passive regurgitation. Because the lumen of the laryngopharynx was not visible on the MRI scans, we could not determine whether the cricopharyngeal muscle was compressed either partially or completely. We cannot determine whether regurgitation would be prevented.

Despite standardization in the technique of CP using a single trained investigator and a standardized pressure reproduced before each application, the frequency and degree of lateral displacement of both the esophagus and the larynx were significantly increased with CP. The CP technique used in this study was modified to accommodate the confines of the MRI machine and the subject who was awake. 10 

Because CP was not performed on unconscious and paralyzed subjects, it is possible that surrounding muscle tone and the swallowing reflex may have influenced the findings. The ideal imaging study would examine subjects without CP before general anesthesia and paralysis, with CP before anesthesia induction, with CP after anesthesia induction and paralysis, and without CP after anesthesia induction. The ethical issues and the technical challenges of applying CP and providing airway and ventilatory support to an anesthetized, paralyzed, unintubated subject in the confines of the MRI scanner prevented us from performing such a study.

Although the two-handed technique does not represent usual clinical practice, this technique should theoretically result in less lateral displacement than the single-handed technique for the following reasons: CP was applied in line with the axial skeleton by an anesthesia resident, and the sensation of midline and symmetrical pressure was confirmed with each subject prior to imaging. Despite these measures, CP was asymmetrically applied, with greater pressure on the right in nearly 50% of subjects. CP at 20 N compressed the airway, displaced both the cricoid and the esophagus to the left, and widened the lateral distance between them. Our results suggest that CP is unreliable at producing midline esophageal compression without distorting the airway anatomy and raises the possibility of airway obstruction when applied with the proper amount of pressure (20–30 N). Airway obstruction has been demonstrated in a previous study using 30 N of CP in anesthetized subjects. 14 

Sellick 1originally described CP with the head and neck fully extended. In our study, the head was placed in a neutral position, and possibly, the esophagus was more mobile in this position. However, in our clinical experience, the patient's head is usually placed in either a neutral or a sniffing position to maximize visual alignment of the larynx, even with the application of CP; therefore, the effect of CP on the neck anatomy in the neutral position may be more generalizable to our clinical practice. Furthermore, because the esophagus lies to the left of midline below the cricoid in greater than 90% of people as it begins its course through the left side of the mediastinum, 13stretching of the esophagus with head and neck extension may pull the esophagus laterally away from the midline.

In summary, this study demonstrated that the cricoid, esophagus, and vertebral body were not aligned in more than half of all normal subjects. The application of CP resulted in an increased frequency and degree of both esophageal and airway displacement relative to the vertebral body.

The authors thank Toni Cormier, M.R.T.M.R., and Ann Sisak, M.R.T.M.R. (Department of Radiology, St. Joseph's Healthcare, Hamilton, Ontario, Canada), for their invaluable help with magnetic resonance imaging; Don McCormick (Pulmonary Development Technologist, Firestone Institute for Respiratory Health, St. Joseph's Healthcare) for his technical support in the manuscript preparation; and all of the volunteers for their time, patience, and trust.