Article, Radiology

Does oral radiocontrast affect image quality of abdominal sonography?

a b s t r a c t

Objective: Emergency Department patients with abdominal pain may require both an ultrasound (US) and com- puted tomography (CT) for an accurate diagnosis. Patients are often asked to drink oral radiocontrast while awaiting ultrasound, in order to better expedite a CT in the case of a non-diagnostic US. The impact of oral radiocontrast on US Image quality has not been studied. We compared the quality of US images obtained before and after the ingestion of oral radiocontrast in healthy adult volunteers.

Methods: This was a prospective study in which adult volunteer subjects underwent sonographic studies of the aorta, the right upper quadrant, the right lower quadrant, and the Focused Assessment with Sonography in Trau- ma (FAST) examination. Initial studies were performed prior to ingestion of oral radiocontrast, with subsequent imaging occurring at 1 and 2 hour post-ingestion. All of the images from the sonographic exams were random- ized and subsequently scored for quality by two emergency ultrasound fellowship trained emergency physicians with extensive experience in performing and interpreting US.

Results: 638 images from 240 exams were obtained from 20 subjects at three time points. Six exams were not scored due to inadequate images. There were no significant differences in image quality for any of the US exam types after the ingestion of oral radiocontrast at 1 and 2 h.

Conclusion: Ingestion of oral radiocontrast did not affect image quality of four common abdominal ultrasound examinations.

(C) 2017

Introduction

Abdominal pain is the most common chief complaint of patients pre- senting to Emergency Departments in the United States [1], and diag- nostic imaging is frequently used in the evaluation of these patients. Ultrasonography is often chosen as the initial diagnostic imaging test for patients with abdominal pain, as it involves no ionizing radiation, is associated with a lower cost when compared to CT and MRI, and is the recommended first line Imaging study in certain clinical situations [2]. However, ultrasonography may not always provide sufficient diag- nostic information due to inadequate visualization of intra-abdominal structures, especially in patients with large amounts of adipose tissue or bowel gas.

A non-diagnostic ultrasound (US) is often followed by a computed tomography (CT) study, for which many centers employ oral radiocontrast, a radio-opaque liquid that can assist in CT interpretation. In order for oral radiocontrast to transit throughout the gastrointestinal

* Corresponding author at: Maimonides Medical Center, 4802 10th Ave, Brooklyn, NY 11219, United States.

E-mail address: [email protected] (L.E. Haines).

tract, a delay of approximately 1 to 2 h is often recommended from the initiation of contrast ingestion until the performance of the CT.

Many emergency physicians employ a time-saving strategy of order- ing oral radiocontrast administration at the same time as ordering an Abdominal ultrasound, if they believe that a subsequent CT scan may be needed. Conversely, patients are asked to be nil per os (NPO) prior to a scheduled ultrasound examination due to concerns that bowel con- tents can adversely affect ultrasound image quality. No previous studies have investigated whether ingestion of oral contrast would have a neg- ative effect on the quality of the abdominal ultrasound images. We hy- pothesized that ingestion of oral radiocontrast would not affect the quality of US images obtained in a group of healthy adult volunteers.

Methods

This was a prospective study conducted in a large, urban, academic hospital with an emergency medicine residency and an emergency ul- trasound fellowship. The study was approved by the hospital’s Institu- tional Review Board. Emergency Department staff were invited to voluntarily participate in the study. Those with underlying gastrointes- tinal disease or previous Allergic reactions to oral radiocontrast were

https://doi.org/10.1016/j.ajem.2017.11.025

0735-6757/(C) 2017

C. Dang et al. / American Journal of Emergency Medicine 36 (2018) 684686 685

Table 1

Comparison of ultrasound images at 1 hour and 2 hour after drinking oral contrast.

Study

Pre-contrast

Post-contrast_1 hour

Post-contrast_2 hours

p-Value

FAST (N = 20)

3.68 +- 0.52

3.80 +- 0.44

3.43 +- 0.59

0.367c; 0.180d

Aorta (N = 20)

3.70 +- 1.03

3.93 +- 1.10

3.78 +- 1.16

0.070; 0.685

RUQ (N = 20)

3.75 +- 0.50

3.90 +- 0.77

3.93 +- 0.73

0.316; 0.246

RLQ (N = 20)

3.13 +- 1.07

2.92 +- 0.96a

2.97 +- 0.83SDb

0.243, 0.799

a N = 18 due to missing images.

b N = 16 due to missing images.

c p-Value for 1 hour contrast versus baseline.

d p-Value for 2 hour contrast versus baseline.

excluded. Informed consent was obtained from all subjects who were then given a standard oral radiocontrast solution to drink (Gastroview,

20 mL of diatrizoate meglumine 66%-diatrizoate sodium 10% (Mallinckrodt Pharmaceuticals; Dublin, Ireland) in 500 mL of sterile water). Each subject underwent sonographic imaging of the aorta, the right upper quadrant, the right lower quadrant, and the Focused Assess- ment with Sonography in Trauma (FAST) exam. All ultrasound exam types were performed just prior to ingestion of oral radiocontrast, and then repeated at one and two hour post-ingestion.

Initial sonogram

Ultrasonography was performed using a Zonare z.one ultra machine (Zonare Medical Systems, Inc. Mountain View, CA) with a 6-2 MHz cur- vilinear transducer. Sonographic examinations were performed by two senior emergency medicine residents and an emergency ultrasound fel- low. All sonologists had previously completed an 8 hour didactic and hands-on course in the use of point-of-care emergency sonography, followed by seven weeks of dedicated point-of-care ultrasound training and at least two years of clinical experience using point of care ultra- sound. Each sonologist had performed a minimum of 25 clinical ultra- sound exams of each of the four exam types used in this study.

The same sonologist performed all of the ultrasound examinations at all three time points for a given subject. For each ultrasound exam type, the sonographers stored still images. The recommended number of stored images for each exam type was the following: For the aorta ex- amination, the sonographers were to have stored transverse images of the proximal, mid, and distal aorta, as well as a longitudinal view of the aorta (4 images). For the right upper quadrant (RUQ) examination, longitudinal and transverse images of the gallbladder were to have been stored (2 images). For the right lower quadrant (RLQ) examination, a transverse image of the right iliac artery, right iliac vein, and psoas mus- cle were to have been obtained and stored (1 image). These Anatomic landmarks functioned as surrogates for the appendix as a normal ap- pendix is often difficult to visualize. For the FAST exam, single images were to be stored from each of the four standard views (subxiphoid, right upper quadrant, left upper quadrant, and suprapubic) (4 images). Thus for each volunteer subject, the study protocol required 11 still im- ages obtained at each of the three time points for a total of 33 images per subject.

Review of sonograms

The exams, which were marked only with a study code, were ran- domized and then evaluated by two expert reviewers who rated image quality using a 5 point scale: 1 – not interpretable, 2 – barely inter- pretable, 3 – adequate for interpretation, but of poor quality, 4 – inter- pretable and of average quality, 5 – interpretable and of superior quality. The reviewers were both fellowship-trained in emergency ul- trasonography, had performed well over 1000 point-of-care ultrasound examinations, and had performed well over 10,000 quality assurance reviews of point-of-care ultrasound examinations. The reviewers were aware of the type of sonographic examination, but were blinded to

the time points of the images, including as to whether the images they were reviewing were obtained before or after ingestion of the oral con- trast. The reviewers were also blinded to the identity of the sonologist who performed the scan, to the study subject being imaged, and to the scores of the other reviewer.

Statistical analysis

The primary outcome was to assess whether there was a difference in image quality after the ingestion of oral radiocontrast. Paired sample t-tests were used to compare the pre-contrast mean image quality score with the mean score at 1 hour post-contrast and the mean score at 2 hour post-contrast for each ultrasound examination.

In addition, we looked at whether there were any significant inter- rater differences in rating of quality of the exams at any time point. A re- peated measures ANOVA design assessed whether there were any changes in the ratings post-contrast (pre-contrast versus one hour post-contrast, and two hour post-contrast) for either of the raters. In order to maximize power, mixed model regression was employed for the analysis in order to allow data from all subjects to be included even in the case of missing values.

A sample size calculation indicated that a minimum of 15 partici- pants was needed for 80% power to detect at least a one point score dif- ference (our minimal clinically significant difference) between baseline and contrast ratings assuming a standard deviation of 1.25 and an effect size of 0.80; a minimum of 19 participants would be necessary for 90% power.

Results

Twenty participants were enrolled into the study and four exams were performed on each of the subjects at three separate times points for a total of 240 possible exams. 638 out of a possible 660 images were available to review allowing us to score 234 of the 240 possible exams. Six right lower quadrant studies not able to be scored due to a lack of images stored for presumed technical archival issues.

As shown in Table 1, there were no significant differences in the image quality rating for any of the exams when comparing the mean pre-contrast rating to the mean one hour and two hour post-contrast

Fig. 1. Raters’ assessment patters over time for FAST exam.

686 C. Dang et al. / American Journal of Emergency Medicine 36 (2018) 684686

Fig. 4. Raters’ assessment patterns over time for RLQ exam.

Fig. 2. Raters’ assessment patters over time for RUQ exam.

time point ratings. p-Values for a comparison of pre-contrast to 1 hour and to 2 hour post-contrast were all >= 0.070. All image quality mean values post-contrast were within 0.5 rating units of the pre-contrast mean, well within our clinically significant difference of 1 point. This held even in the case of the smallest p-value (2 hour post-contrast Aorta compared with Aorta pre-contrast).

There were no significant inter-rater differences of image quality rating over time for any of the four exam types (see Figs. 1-4). There were similar pattern of ratings between the reviewers across the three time points for all four exam types despite some minor differences in agreement based on inspection of the trend lines for each rater (p N 0.16). For example, in the FAST exam, Rater 2 appeared to give a slightly higher rating to the image quality at 1 hour post-contrast. Similarly, Rater 2 gave a higher rating for the RUQ exam at the 2 hour post- contrast time point. In contrast, for the RLQ exam, Rater 1 tended to as- sign a lower image quality at both post-contrast time points. Despite these minor differences, we assumed a clinically significant difference between mean rating scores to be 1.0 rating units or greater. There were no inter-rater discrepancies N 0.5 units. We conclude accordingly that no differences were either statistically or clinically significant.

Discussion

undifferentiated abdominal pain is the most common chief com- plaint encountered in the Emergency Departments (EDs) in the United States [1]. imaging tests are often included in the evaluation of these patients. If an US is ordered as the first line imaging study, but is subsequently non-diagnostic, a CT may be necessary. If a patient begins drinking oral contrast only after the US is performed, this could contrib- ute to a prolonged length of stay in the ED. As EDs in this country are be- coming increasingly crowded, a method to streamline the evaluative process and decrease length of stay would be beneficial.

Our findings demonstrate that there were no differences in sono- graphic image quality when comparing images of the right lower quad- rant, the gallbladder, the aorta, and the FAST exam before and after the ingestion of oral radiocontrast. Therefore, we would advocate giving

patients contrast while they wait for an US if there is anticipation that these patients may subsequently require a CT with oral contrast.

The limitations of the study include that the data was gathered from a single site location from a small number of healthy volunteers, which may not accurately reflect an undifferentiated ED patient population. In addition, the data were obtained by a small number of sonographers. Also, images of the iliac vessels and psoas muscle were chosen as a sur- rogate marker for an appendix ultrasound, as a normal appendix is often not visualized. The image quality of an inflamed appendix may be af- fected by radiocontrast. Furthermore, 22 images (3% of the total) were not stored for expert review preventing interpretation of 6 exams. The reasons for this are unknown, and presumed to be a result of sonologist error in saving images. However, since this number represents only a small fraction of the total images that were to be obtained and we were able to score 234 of the 240 exam types (98%), we feel that these missing images would not have significantly altered our final re- sults. Also, we only had 2 expert raters review all of the images, and both of these raters had similar training in emergency ultrasonography. If there were more reviewers from a variety of specialties, the results may have been different. Lastly, our participants were not asked to be NPO prior to the examinations as we felt this would more closely reflect the reality of an ED environment. This effect is likely mitigated by the fact that each subject served as their own control for the subsequent time points after ingestion of oral contrast.

In conclusion, this study demonstrated that ingestion of oral

radiocontrast did not compromise abdominal ultrasound image quality; therefore, early administration of oral contrast can be utilized if there is a suspicion that a CT will be necessary subsequent to the ultrasound examination.

References

  1. Thompson DA, Eitel D, Fernandes CM, Pines JM, Amsterdam J, Davidson SJ. Coded chief complaints–automated analysis of free-text complaints. Acad Emerg Med 2006;13(7):777-8.
  2. American College of Radiology Appropriateness Criteria. http://www.acr.org/Quality- Safety/Appropriateness-Criteria, Accessed date: 15 February 2016.

    Fig. 3. Raters’ assessment patterns over time for Aorta exam.

Leave a Reply

Your email address will not be published. Required fields are marked *