Professional Documents
Culture Documents
Chang et al.
Positioning Algorithm During Ultrasound for Appendicitis
Downloaded from www.ajronline.org by 36.70.68.21 on 05/17/16 from IP address 36.70.68.21. Copyright ARRS. For personal use only; all rights reserved
Gastrointestinal Imaging
Original Research
1006
OBJECTIVE. The purpose of this article is to examine the rates of appendiceal visualization by sonography, imaging-based diagnoses of appendicitis, and CT use after appendiceal
sonography, before and after the introduction of a sonographic algorithm involving sequential
changes in patient positioning.
MATERIALS AND METHODS. We used a search engine to retrospectively identify
patients who underwent graded-compression sonography for suspected appendicitis during
6-month periods before (period 1; 419 patients) and after (period 2; 486 patients) implementation of a new three-step positional sonographic algorithm. The new algorithm included initial conventional supine scanning and, as long as the appendix remained nonvisualized, left
posterior oblique scanning and then second-look supine scanning. Abdominal CT within 7
days after sonography was recorded.
RESULTS. Between periods 1 and 2, appendiceal visualization on sonography increased
from 31.0% to 52.5% (p < 0.001), postsonography CT use decreased from 31.3% to 17.7% (p<
0.001), and the proportion of imaging-based diagnoses of appendicitis made by sonography increased from 63.8% to 85.7% (p = 0.002). The incidence of appendicitis diagnosed by imaging (either sonography or CT) remained similar at 16.5% and 17.3%, respectively (p = 0.790).
Sensitivity and overall accuracy were 57.8% (95% CI, 44.870.1%) and 93.0% (95% CI, 90.1
95.3%), respectively, in period 1 and 76.5% (95% CI, 65.885.2%) and 95.4% (95% CI, 93.1
97.1%), respectively, in period 2. Similar findings were observed for adults and children.
CONCLUSION. Implementation of an ultrasound algorithm with sequential positioning significantly improved the appendiceal visualization rate and the proportion of imagingbased diagnoses of appendicitis made by ultrasound, enabling a concomitant decrease in abdominal CT use in both children and adults.
their sample in this fashion. Those investigators achieved an 89% appendiceal visualization rate by scanning in the supine position
alone and a 99% visualization rate (a figure
that has not been achieved in a North American study) by using a variety of additional
operator-dependent maneuvers, including
intermittent scanning in the LPO position.
The sonographic visualization rates in North
American reports have ranged from 31% to
88% [1215] to as low as 2.4% among patients without appendicitis [16].
Because the appendix lies in a retrocecal
position in 2565% of patients [1719], we
postulated that the routine additional use of
scanning in the LPO position may be very
beneficial in our North American patient
population in whom supine-only scanning
provides visualization rates lower than those
reported in the study of Korean individuals
Downloaded from www.ajronline.org by 36.70.68.21 on 05/17/16 from IP address 36.70.68.21. Copyright ARRS. For personal use only; all rights reserved
Fig. 1Photograph
showing scanning
parallel to psoas
muscle with patient
in left posterior
oblique position, as
implemented during
training period and
conducted throughout
period 2.
duced and implemented in our ultrasound laboratory; the training interval began on June 1, 2012.
Both pediatric patients (18 years old or younger)
and adults were included. In period 1, a total of
419 patients were thus identified, with a mean age
of 17 years (range, < 1 to 81 years), including 154
adult patients (28 men and 126 women) and 265
pediatric patients (138 boys and 127 girls). In period 2, a total of 486 patients were identified, with
a mean age of 16 years (range, 187 years), including 156 adult patients (25 men and 131 women) and 330 pediatric patients (156 boys and 174
girls). Review of the radiology PACS identified
which of these patients underwent abdominopelvic CT within 7 days of their graded-compression
sonography examinations.
At our institution, graded-compression sonography is the first-line modality for the evaluation of suspected appendicitis. CT, the secondline modality, is used when graded-compression
sonography does not provide a diagnosis and there
is ongoing clinical concern for appendicitis. Di-
Patient Selection
Fig. 219-year-old man with acute retrocecal appendicitis visualized only on coronal scanning in left posterior oblique (LPO) position.
A, Transverse sonogram of right lower quadrant in initial supine position shows acoustic shadowing from gas in cecum (C) but did not identify appendix.
B, Coronal sonogram acquired through right flank with patient in 45 LPO position shows psoas muscle (P) and noncompressible enlarged 9-mm appendix (arrow)
consistent with appendicitis. Despite sonographic diagnosis of appendicitis, attending surgeon requested CT.
C, Contrast-enhanced axial CT image confirms retrocecal appendicitis (arrow) and shows gas within cecum (C).
Downloaded from www.ajronline.org by 36.70.68.21 on 05/17/16 from IP address 36.70.68.21. Copyright ARRS. For personal use only; all rights reserved
Chang et al.
Fig. 38-year-old girl with normal retrocecal
appendix seen only on coronal imaging in left
posterior oblique (LPO) position.
A, Transverse right lower quadrant sonogram in initial
supine position shows small-bowel loops (SB) and
psoas muscle (P) but no visualization of appendix.
B, Coronal sonogram acquired through right flank
with patient in LPO position shows normal appendix
(arrow) posterior to cecum (C) and anterior to psoas
muscle (P). Calipers mark appendix.
A
5 years of clinical experience performing appendiceal graded-compression sonography. Sonographers were supervised on a case-by-case
basis by radiology residents, body imaging fellows, and attending radiologists in our abdominal imaging and pediatric imaging sections. All
scans were immediately reviewed by supervising
residents, fellows, or attending radiologists, who
performed additional scanning on an as-needed
basis for equivocal cases. Ultrasound equipment
included the Logiq 9 system (GE Healthcare) and
Sequoia System 512 (Siemens Healthcare) using
linear-array transducers ranging in frequency
from 8 to 15 MHz.
During period 1, all sonographers performed
standardized graded-compression sonography,
as described elsewhere [20, 21], with all patients
scanned only in the supine position. Whenever
possible, each patient was initially asked to indicate their site of maximal abdominal pain,
and this region was specifically scanned during
the study [22]. Scanning was initiated using the
highest frequency transducer providing penetration adequate to visualize the right psoas
muscle and external iliac artery [14], beginning
with the right kidney and the hepatorenal fossa
in the right flank, in the transverse plane, and
continuing inferiorly to the ascending colon. An
attempt was then made to visualize the insertion of the terminal ileum into the cecum, with
gradually increasing pressure applied with the
transducer to reveal the caput of the cecum and
the appendiceal origin. Transverse and longitudinal images were obtained of the cecal region,
as well as cine clips, with the patient remaining
in the supine position. In all female patients of
reproductive age in whom the appendix was not
visualized by transabdominal graded-compression sonography, endovaginal scanning of the
uterus and adnexa was performed using a 6- to
8-MHz endovaginal probe, as is routine for patients with suspected appendicitis at our institution, unless such scanning is contraindicated or
declined by the patient.
1008
B
Training Interval
CT Technique
Abdominopelvic CT was performed using 16or 64-row MDCT scanners (LightSpeed VCT, GE
Healthcare). IV contrast medium was used routinely in all patients, without oral or rectal contrast medium. For adult patients, 150 mL of iohexol medium (Omnipaque 350, GE Healthcare)
was injected via power injector at a rate of 3 mL/s.
For pediatric patients, 2 mL of contrast material
per kilogram of body weight was injected at a rate
of 2 mL/s. Portal venous phase images were obtained from the diaphragm through the pelvis with
collimation of 0.6251.25 mm. Axial images were
reconstructed at slice thicknesses of 1.255 mm,
and both coronal and sagittal images were reconstructed at 2-mm slice thickness.
Adults
(n = 154)
Children
(n = 265)
Total
(n = 486)
Adults
(n = 156)
Children
(n = 330)
130 (31.0)
37 (24.0)
93 (35.1)
255 (52.5)
62 (39.7)
193 (58.5)
< 0.001
0.003
< 0.001
84 (17.3)
21 (13.5)
63 (19.1)
0.790
0.866
1.000
72 (85.7)
18 (85.7)
54 (85.7)
0.002
0.148
0.008
86 (17.7)
48 (30.8)
34 (10.3)
< 0.001
0.034
0.005
Observation
Patients with appendix visualized by sonography
No. (% of n)
p
Imaging-based diagnoses of appendicitis made by CT or
graded-compression sonography
No. (% of n)
69 (16.5)
19 (12.3)
50 (18.9)
p
Sonographic imaging-based diagnoses of appendicitis
No. (% of imaging-based diagnoses of appendicitis made by CT or
graded-compression sonography)
44 (63.8)
12 (63.2)
32 (64.0)
p
Patients who underwent abdominal CT within 7 days of sonography
No. (% of total n)
131 (31.3)
66 (42.9)
65 (24.5)
p
NoteThe p values relate findings in period 2 to corresponding findings in period 1.
Results
Patient Sample
Patients in periods 1 and 2 did not differ
significantly with respect to sex (166 male
and 253 female patients in period 1 vs 181
male and 305 female patients in period 2; p=
0.493), adults versus children (154 adults and
265 children in period 1 vs 156 adults and 330
children in period 2; p = 0.160), or mean age
Period 1
January 2013
February 2013
December 2012
October 2012
November 2012
August 2012
Training
Period
September 2012
July 2012
May 2012
0.0
June 2012
15.0
April 2012
30.0
March 2012
The Fisher exact test was used to compare demographic data and results between period 1 and
period 2, the exact binomial test was used to compare performance values with 95% CIs, and the
Student t test was used to compare ages. Analy-
45.0
January 2012
Statistical Analysis
60.0
February 2012
December 2011
All clinical ultrasound examinations with visualization of the appendix during period 2 were
subsequently reviewed by a blinded independent
radiologist with over 25 years of experience in
performing graded-compression sonography to
define whether the appendix was first visualized
on the conventional supine images, the subsequent
LPO images, or the final second-look supine images. These assessments were facilitated by sonographers image annotations and were confirmed using visualized anatomic landmarks, such
as the psoas muscle and the external iliac vessels.
Rate (%)
Downloaded from www.ajronline.org by 36.70.68.21 on 05/17/16 from IP address 36.70.68.21. Copyright ARRS. For personal use only; all rights reserved
Period 2
Period 2
Month
Chang et al.
TABLE 2: Visualization of the Appendix in Period 1 (All Conventional Scanning) and First Visualization of the
Appendix in Period 2 (New Three-Step Algorithm)
Period 1
Downloaded from www.ajronline.org by 36.70.68.21 on 05/17/16 from IP address 36.70.68.21. Copyright ARRS. For personal use only; all rights reserved
Period 2
Total
(n = 419)
Adults
(n = 154)
Children
(n = 265)
Total
(n = 486)
Adults
(n = 156)
Children
(n = 330)
130 (31.0)
37 (24.0)
93 (35.1)
147 (30.2)
39 (25.0)
108 (32.7)
0.828
0.895
0.601
69
13
56
216 (44.4)
52 (33.3)
164 (49.7)
< 0.001
0.079
< 0.001
39
10
29
255 (52.5)
62 (39.7)
193 (58.5)
< 0.001
0.003
< 0.001
Details of the increased appendiceal visualization rates achieved during each step of period 2 are presented in Table 2. The appendix
was first seen in the conventional supine position, the subsequent LPO position, and the
final second-look supine position in 30.2%
(147/486), 14.2% (69/486), and 8.0% (39/486)
of patients, respectively. Cumulative appendiceal visualization rates achieved during each
of the three steps of period 2 are presented in
Table 2 and significantly exceeded those of period 1 for all patients and for children; the LPO
step did not register a statistically significant
increase among the small sample of adults, although the second-look supine step did. No statistically significant differences were present
between visualization of the appendix during
the conventional all-supine scanning of period 1 and the initial conventional supine scanning step of period 2, a finding that was true for
all patients, both adults and children (Table 2).
Figures 2 and 3 provide representative clinical
images. Sonographers typically required 510
additional minutes to perform the new positional maneuvers in period 2 compared with
the conventional scanning in period 1.
Imaging-Based Diagnosis of Appendicitis
The incidence of appendicitis among all
patients, whether diagnosed by graded-compression sonography or CT, did not differ significantly between period 1 (16.5%; 69/419
patients) and period 2 (17.3%; 84/486 patients; p = 0.790) (Table 1). Similarly, no statistically significant differences were found
among adults and children.
1010
Downloaded from www.ajronline.org by 36.70.68.21 on 05/17/16 from IP address 36.70.68.21. Copyright ARRS. For personal use only; all rights reserved
Value
All Patients
Period 2
Adults
Children
All Patients
Adults
Children
Appendix visualized on
graded-compression sonography
(no. of findings)
True positive
37
11
26
62
18
44
True negative
84
25
59
178
42
136
False positive
False negative
True negative
264
110
154
217
92
125
False negative
25
18
14
12
57.8 (44.870.1)
61.1 (35.782.7)
56.5 (41.171.1)
76.5 (65.885.2)
81.8 (59.794.8)
74.6 (61.685.0)
99.4 (98.099.9)
100.0 (97.3100.0)
99.1 (96.799.9)
99.2 (97.899.8)
100.0 (97.3100.0)
98.9 (96.799.8)
94.9 (82.399.4)
100.0 (71.5100.0)
92.9 (76.599.1)
95.4 (87.199.0)
100.0 (81.5100.0)
93.6 (82.598.7)
99.4 (98.099.9)
100.0 (97.3100.0)
99.1 (96.799.9)
98.8 (97.199.6)
98.5 (94.899.8)
98.9 (96.799.8)
93.0 (90.195.3)
95.4 (90.898.1)
91.6 (87.594.6)
95.4 (93.197.1)
97.4 (93.699.3)
94.4 (91.396.7)
false-negatives included one child with an 8to 9-mm structure thought to be the terminal
ileum or possibly the appendix and another
child with a normal 5-mm appendix, both of
whom proved to have appendicitis on pathologic examination. The five false-negatives
in period 2 included two children and two
adults with noncompressible 6- to 7-mm appendixes and one child with a normal 6-mm
compressible appendix, all of whom had appendicitis on pathologic examination.
Discussion
For both adult and pediatric patients with
suspected appendicitis, our data show that a
three-step graded-compression sonography
algorithm of sequential positioning can significantly improve visualization of the appendix compared with graded-compression sonography performed only in the supine position
and is associated with a decreased rate of subsequent CT use. Specifically, the appendiceal
visualization rate increased from 31.0% to
52.5% for all patients, while CT use decreased
from 31.3% to 17.7% (Table 1), and statistically significant changes were evident for both
children and adults. In addition, the proportion of imaging-based diagnoses achieved
by graded-compression sonography rather
than CT increased significantly among children; this proportion also increased among
the smaller contingent of adults but did not
achieve statistical significance. At our insti-
Downloaded from www.ajronline.org by 36.70.68.21 on 05/17/16 from IP address 36.70.68.21. Copyright ARRS. For personal use only; all rights reserved
Chang et al.
ary nature of our algorithm and, accordingly,
more accurately reflect the potential value of
these positional changes. Other maneuvers
have also been reported as helpful in visualizing the appendix [25].
During the 3-month training interval between periods 1 and 2, our sonographers reported increasing confidence in visualizing
the appendix in general, and small normal
appendixes in particular. Once familiar with
the new three-position algorithm, the sonographers typically required approximately
510 additional minutes to perform the additional LPO and second-look supine steps.
Although we cannot rigorously exclude a
contribution of the Hawthorne effect, in which
individuals perform better while under observation [26], the similar rates of appendiceal visualization during the all-supine scanning in
period 1 and the initial supine scanning step
in period 2 argue against this. Furthermore, although a significantly greater proportion of appendixes were visualized in period 2 compared
with period 1, the sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of graded-compression sonography did not differ significantly.
A number of limitations deserve mention
with respect to this investigation. First, we systematically evaluated the LPO scanning and
then second-look supine scanning after conventional supine scanning proved to be unrevealing, but it was not feasible to evaluate the
second and third steps alone or in the reverse
order. Next, our sonographers required 510
additional minutes for the new positional maneuvers, but this value may differ at other institutions. Although we cannot entirely exclude
the possibility that some of the improved performance with the additional positional maneuvers may simply reflect longer periods of
scanning, our long-standing experience has
been that prolonged supine scanning rarely, if
ever, provides visualization of an appendix that
was not evident during the conventional supine
examination. Although the phase of scanning
providing the first visualization of the appendix was generally clear on review of gradedcompression sonography images, the degree to
which images were annotated was slightly variable. Finally, this investigation is retrospective,
and prospective studies may help further define
the utility of the three-position algorithm.
1012