You are on page 1of 7

3209jum_Layout 1 8/20/13 11:43 AM Page 1547

ORIGINAL RESEARCH

Clinical Outcomes of Pediatric Patients


With Acute Abdominal Pain and Incidental
Findings of Free Intraperitoneal Fluid on
Diagnostic Imaging
Samantha Matz, DO, Mary Connell, MD, Madhumita Sinha, MD, Christopher S. Goettl, MD, Palak C. Patel, MD,
David Drachman, PhD

Article includes CME test ObjectivesThe presence of free intraperitoneal fluid on diagnostic imaging (sonog-
raphy or computed tomography [CT]) may indicate an acute inflammatory process in
children with abdominal pain in a nontraumatic setting. Although clinical outcomes of
pediatric trauma patients with free fluid on diagnostic examinations without evidence
of solid-organ injury have been studied, similar studies in the absence of trauma are rare.
Our objective was to study clinical outcomes of children with acute abdominal pain of
nontraumatic etiology and free intraperitoneal fluid on diagnostic imaging (abdomi-
nal/pelvic sonography, CT, or both).
MethodsWe conducted a retrospective review of medical records of children aged 0
to 18 years presenting to a pediatric emergency department with acute abdominal pain
(nontraumatic) between April 2008 and March 2009. Patients with intraperitoneal free
fluid on imaging were divided into 2 groups: group I, imaging suggestive of an intra-
abdominal surgical condition such as appendicitis; and group II, no evidence of an acute
surgical condition on imaging, including patients with equivocal studies. Computed
Received November 20, 2012, from the Depart- tomograms and sonograms were reviewed by a board-certified radiologist, and the free
ments of Radiology (S.M., M.C.) and Research fluid volume was quantitated.
(D.D.), Maricopa Medical Center, Phoenix,
Arizona USA; Department of Pediatric Emer- ResultsOf 1613 patients who underwent diagnostic imaging, 407 were eligible for
gency Medicine, Arizona Childrens Center at the study; 134 (33%) had free fluid detected on diagnostic imaging. In patients with
Maricopa Medical Center, Phoenix, Arizona USA both sonography and CT, there was a significant correlation in the free fluid volume
(M.S.); and University of Arizona College of
Medicine, Tucson, Arizona USA (C.S.G., P.C.P.).
(r = 0.79; P < .0005). A significantly greater number of male patients with free fluid had
Revision requested December 7, 2012. Revised a surgical condition identified on imaging (57.4% versus 25%; P < .001). Children with
manuscript accepted for publication January 22, free fluid and an associated condition on imaging were more likely to have surgery
2013. (94.4% versus 6.3%; P < .001).
This study was presented as a podium pres-
entation at the 112th Annual Meeting of the ConclusionsWe found clinical outcomes (surgical versus nonsurgical) to be most
American Roentgen Ray Society; May 2012; correlated with a surgical diagnosis on diagnostic imaging and not with the amount of
Vancouver, British Columbia, Canada. fluid present.
Address correspondence to Madhumita Key Wordsabdomen; children; free fluid; imaging; pain
Sinha, MD, Department of Pediatrics, Maricopa
Integrated Health System, 2601 E Roosevelt St,
Phoenix, AZ 85008 USA.
E-mail: madhumita_sinha@dmgaz.org

Abbreviations
CT, computed tomography; ED, emergency
department
T he presence of free intraperitoneal fluid may be a cardinal
sign of an acute inflammatory process in a patient with
abdominal pain in a nontraumatic setting. Small amounts of
free fluid have been found on diagnostic imaging and are consid-
ered physiologic in asymptomatic children.1,2 Although a finding of
doi:10.7863/ultra.32.9.1547 free intraperitoneal fluid among children with blunt abdominal

2013 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2013; 32:15471553 | 0278-4297 | www.aium.org
3209jum_Layout 1 8/20/13 11:43 AM Page 1548

Matz et alFree Intraperitoneal Fluid in Pediatric Patients With Abdominal Pain

trauma may signal an associated solid or hollow viscous tion such as appendicitis; and group II, including children
injury, recent studies have suggested that isolated free with free intraperitoneal fluid but no evidence of an asso-
intraperitoneal fluid detected on diagnostic imaging in ciated surgical condition on diagnostic imaging and those
pediatric trauma patients, with no associated findings of who had equivocal studies.
solid-organ injury, a normal mental status, and lack of abdom-
inal tenderness, may be managed conservatively without Free Fluid Volume Estimation and Categorization
surgery.35 Our review of published literature yielded few Formal CT and sonographic reports were reviewed by the
studies addressing the same issue in children with acute study team; in addition, images of patients reported to have
abdominal pain of nontraumatic etiology who have inci- free intraperitoneal fluid were reexamined by a board-
dental findings of free fluid without an associated surgical certified pediatric radiologist. Free fluid volume was esti-
condition.6 mated by using the ellipsoid volume formula (4/3 width
Our objective was to study clinical outcomes of chil- height length ; Figure 1).1
dren with acute abdominal pain (nontraumatic etiology)
and isolated findings of free intraperitoneal fluid on diag-
nostic imaging (abdominal/pelvic sonography, computed
tomography [CT], or both). The primary outcome of the
study was the need for surgery; secondary outcomes
included emergency department (ED) disposition, com- Figure 1. Images from a 16-year-old female patient presenting to the
ED with acute abdominal pain. The patient was discharged from the ED
plications, and return visits to the ED for the same condi- with a diagnosis of constipation. A, Sagittal oblique long-axis sonogram
tion after ED discharge. obtained at presentation to the ED showing a pocket of anechoic free
fluid in the pelvis. B, Transverse short-axis pelvic sonogram obtained
Materials and Methods subsequently. Pockets of free fluid were measured in 3 orthogonal
planes, and free fluid volumes were estimated by the ellipsoid volume
formula.
Design
We conducted a retrospective review of medical records A
of children aged 0 to 18 years who presented to the pedi-
atric ED of a tertiary care teaching hospital between April
2008 and March 2009 with acute abdominal pain (non-
traumatic) and who had diagnostic imaging (sonography,
CT, or both) performed.

Setting
This study was conducted in the pediatric ED of a 449-bed
tertiary care teaching hospital. The pediatric ED evaluates
approximately 20,000 pediatric patients annually and also
serves as a pediatric trauma referral center. The Institu-
tional Review Board of the hospital approved the study.
B
Patients
Pediatric patients with acute abdominal pain who under-
went diagnostic imaging while in the ED, were identified
by Current Procedural Terminology codes. Patients with a
history of trauma, patients who had a diagnostic examina-
tion for nonacute abdominal pain (>7 days), and patients
with other chronic abdominal conditions such as inflam-
matory bowel disease were excluded from the study.
Patients who had free intraperitoneal fluid detected on
diagnostic imaging were divided into 2 groups: group I,
including children who had positive sonographic or CT
findings suggestive of an intra-abdominal surgical condi-

1548 J Ultrasound Med 2013; 32:15471553


3209jum_Layout 1 8/20/13 11:43 AM Page 1549

Matz et alFree Intraperitoneal Fluid in Pediatric Patients With Abdominal Pain

Sonographic and CT Scan Techniques detected). If the amount of free fluid detected was too
Sonography (transabdominal and transvaginal in sexually small to quantify, it was designated trace.
active female patients) was performed with LOGIQ 9
ultrasound equipment (GE Healthcare, Waukesha, WI) Statistical Analysis
using real-time 47-MHz linear and curved probes with Statistical analysis was done with SPSS version 19.0 soft-
color Doppler capabilities. For abdominal scans, a survey ware for Windows (SPSS Inc, Chicago, IL). We used
of the abdomen was obtained, including transverse, descriptive statistics to describe sample characteristics and
oblique, and longitudinal images of the upper right, lower t tests and 2 tests to compare demographic (age, sex, and
right, upper left, and lower left quadrants. For pelvic imag- race) and clinical characteristics (type of diagnostic exam-
ing of female patients, the pelvis was surveyed, including ination, surgery done, final diagnosis of appendicitis, dis-
transverse, oblique, and longitudinal images of the bilat- position from the ED, and complications at follow-up)
eral adnexal regions and rectouterine pouch. Free fluid vol- between the groups. P .05 was considered significant.
umes were measured in 3 orthogonal planes, calculated
according to the ellipsoid volume formula, and expressed in Subgroup Analysis
milliliters. A gray scale morphologic evaluation of the fluid Since our cohort of interest was the group II patients, who
collections was performed with attention to echogenicity, had free fluid detected on diagnostic imaging without asso-
septations, and multifocality. According to these parame- ciated surgical conditions and their outcomes, further cat-
ters, fluid collections were classified as free fluid (complex egorization of this group was done to enable a detailed
or simple collections). In patients who had multiple pock- subgroup analysis: subgroup IIa included female patients
ets of free fluid, the 3 largest measureable pockets were older than 10 years, and subgroup IIb included all male
each measured independently in 3 orthogonal planes, and patients in group II as well as female patients 10 years or
the total was summed. younger.
Computed tomography was performed with a Light-
Speed RT 16 CT scanner (GE Healthcare). Patients Results
received a nonionic intravenous contrast agent (Omni-
paque 350 [iohexol]; GE Healthcare) in a single parenteral Overall Results
dose of 1.0 mL/kg, administered by a power injector. In Of the 1613 patients who underwent diagnostic imaging
addition, some patients also received an oral contrast agent for an abdominal/pelvic condition in the ED over a 1-year
(Gastrografin [diatrizoate meglumine and diatrizoate period, 407 fulfilled eligibility requirements and were
sodium]; Bracco Diagnostics, Inc, Monroe Township, included in the study: 250 (61.4%) had sonography, 210
NJ). For patients 0 to 8 years of age, 100 kV (peak) and (51.6%) had CT, and 55 (13.5%) had both. The mean age
Smart mA (automatic exposure control) were used with was 10.6 years (SD, 5.9 years), and 169 (41.5%) were male.
2.5-mm-thick slices and a 1.25-mm interval. For patients Most (79.9%) were Hispanic. In patients with both sonog-
older than 8 years, 120 kV (peak) and Smart mA were used raphy and CT, there was significant correlation in the free
with 5-mm-thick slices and a 4-mm interval. Helical imag- fluid volume (r = 0.79; P < .0005). A total of 134 (33%)
ing of the abdomen and pelvis was obtained with sagittal patients had free fluid detected on diagnostic imaging.
and coronal reformatted images. Similar to the sono- Of those patients who had free fluid on diagnostic
graphic technique, morphologic evaluation of the fluid col- examinations (sonography, CT, or both), 54 (40.3%) were
lections was performed with attention to density, assigned to group I with a definitive surgical condition
septations, and multifocality. According to these parame- identified on imaging. Comparisons of demographic char-
ters, fluid collections were classified as free fluid or locu- acteristics and clinical outcomes between groups I and II
lated collections. In patients who had multiple pockets of are illustrated in Table 1. A significantly greater number of
free fluid, the 3 largest measureable pockets were each male patients with free fluid had a surgical condition iden-
measured independently in 3 orthogonal planes, and the tified on the diagnostic examination (57.4% versus 25.0%;
total was summed (Figure 2). P < .001). There was no difference in the mean age or eth-
On the basis of the free intraperitoneal fluid volume nicity/race among patients with free intraperitoneal fluid
estimation for individual patients, it was decided to divide with and without an associated intra-abdominal surgical
patients into the following categories: small (<15 mL of condition warranting surgery. Children with free fluid and
free fluid detected on sonography or CT), medium (1550 an associated condition on diagnostic examinations were
mL of free fluid detected), and large (>50 mL of free fluid most likely to have surgery (94.4% versus 6.3%; P < .001).

J Ultrasound Med 2013; 32:15471553 1549


3209jum_Layout 1 8/20/13 11:43 AM Page 1550

Matz et alFree Intraperitoneal Fluid in Pediatric Patients With Abdominal Pain

Figure 2. Images from a 9-year-old male patient presenting to the ED with acute abdominal pain who had findings consistent with acute appen-
dicitis with free fluid on imaging. The patient was transferred to the operating room from the ED and was found to have acute gangrenous appendicitis
on pathologic examination. Images represent an irregular fluid collection on CT broken down into 3 separate pockets for the purposes of meas-
urement. The 3 pockets were each measured in 3 orthogonal planes, and the total was summed. A, Anteroposterior and craniocaudal measure-
ments of pocket 1. B, Lateral measurement of pocket 1. C, Anteroposterior and craniocaudal measurements of pocket 2. D, Lateral measurement
of pocket 2. E, Anteroposterior and craniocaudal measurements of pocket 3. F, Lateral measurement of pocket 3.

A B C

D E F

1550 J Ultrasound Med 2013; 32:15471553


3209jum_Layout 1 8/20/13 11:43 AM Page 1551

Matz et alFree Intraperitoneal Fluid in Pediatric Patients With Abdominal Pain

In group II with free intraperitoneal fluid, 5 patients (6.3%) Subgroup Analysis Results
had equivocal examinations for an intra-abdominal surgi- Fifty-one patients (63.8%) were categorized into subgroup
cal condition, and the clinical outcomes of these patients IIa, and 29 (36.2%) were categorized into subgroup IIb
were analyzed: 2 patients were taken to the operating room (Table 3). Children categorized into subgroup IIb were
from the ED on clinical suspicion and found to have significantly more likely to have a surgical outcome than
appendicitis intraoperatively; 1 had surgical repair of an those in group IIa (13.8% versus 2.0%; P = .036). Most
incarcerated inguinal hernia; 1 had an intussusception female patients in subgroup IIa (66.7%), who had free fluid
reduction in the operating room; and 1 was admitted for without an associated condition, were discharged from the
observation and discharged from the inpatient unit with a ED and did not return with complications. Interestingly,
deferred ovarian cyst aspiration performed as an outpatient 100% of patients in subgroup IIb had a large amount of
procedure. Final discharge diagnoses of the remaining 75 free fluid on diagnostic CT, compared to 68.4% in sub-
patients in group II who did not undergo surgery are listed group IIa, and this difference approached significance. Of
in Table 2. the 15 patients with a large amount of free fluid on CT, 1
had a surgical outcome. Of the 4 patients in group IIb who
had surgical outcomes, 2 had large volumes identified on
imaging.

Discussion
Table 1. Comparison of Demographic and Clinical Characteristics
Among Patients With Free Intraperitoneal Fluid and the Presence (Group
I) or Absence (Group II) of Any Associated Surgical Condition Identified
To do or not to do a surgical exploration in children with
on Diagnostic Imaging acute abdominal pain and free intraperitoneal fluid in the
absence of an associated surgical condition clearly identi-
Group I Group II
Characteristic (n = 54) (n = 80) P
fied on a diagnostic imaging often poses a serious man-
agement dilemma for physicians. Although recent studies
Male, n (%) 31 (57.4) 20 (25) <.001a in pediatric trauma patients have shown that nonsurgical
Mean age SD, y 11.79 3.9 12.9 4.9 .12b
Ethnicity/race, n (%) .16c
management is safe even in the presence of small amounts
Hispanic/Latino 45 (83.2) 59 (73.7) of free fluid in children without evidence of an associated
Non-Hispanic white 2 (3.7) 6 (7.5) solid-organ injury,7 the same issue in patients without
African American 1 (1.9) 8 (10.0) trauma has not been well studied. In this retrospective
Asian 0 (0.0) 0 (0.0)
study, we attempted to look at the outcomes of children
Native American 5 (9.3) 6 (7.5)
Other 1 (1.9) 1 (1.3) with acute abdominal pain who have free fluid on sonog-
Diagnostic examination, n (%) .002c raphy or CT; the primary outcome was the need for sur-
CT only 30 (55.6) 37 (46.3)
Sonography only 13 (24.1) 33 (41.3)
Both 11 (20.4) 20 (25) Table 2. Final Discharge Diagnoses for Patients in Group II With Free
Surgery done, n (%) 51 (94.4) 5 (6.3) <.001a Intraperitoneal Fluid, No Surgical Condition on the Diagnostic Exam-
Final diagnosis of 44 (81.5) 2 (2.5) <.001a ination, and a Nonsurgical Final Outcome
appendicitis, n (%)
Disposition from ED, n (%) <.001c Diagnosis n
Admitted and went to surgery 18 (33.3) 2 (2.5) Abdominal pain (unspecified) 28
later Mesenteric adenitis 2
Admitted for inpatient 3 (5.6) 34 (42.5) Acute gastroenteritis 9
observation Colitis 3
Discharged from ED 0 (0.0) 41 (51.2) Constipation 3
Transferred to operating room 33 (61.1) 3 (3.8) Ovarian cyst 12
from ED Pelvic inflammatory disease 6
Complications at follow-up, n (%) 4 (7.5) 0 (0.0) .02a Urinary tract infection 6
A significantly greater number of male patients with free fluid had a sur- Intussusception (nonsurgical reduction) 1
gical condition identified on the diagnostic examination. Nephrolithiasis 2
aContinuity-corrected 2 test of a 2 2 table. Spontaneous abortion 1
bIndependent samples t test. Dysfunctional uterine bleeding 1
c2 test of a 2 n contingency table, where 2 is the number of groups, Gastritis 1
and n is the number of categories. Total 75

J Ultrasound Med 2013; 32:15471553 1551


3209jum_Layout 1 8/20/13 11:43 AM Page 1552

Matz et alFree Intraperitoneal Fluid in Pediatric Patients With Abdominal Pain

gery. We found that in pediatric patients who presented to had equivocal sonographic examinations with the presence
the ED with acute abdominal pain (nontraumatic) and in of free fluid and were found to have acute appendicitis
whom free fluid was detected on diagnostic imaging with intraoperatively also had clinical examinations that were
no associated identifiable surgical condition (eg, appen- highly suggestive of acute appendicitis. All patients with
dicitis), the mere presence of fluid and the fluid volume free fluid and no associated condition who were discharged
were not predictive of a surgical outcome. However, in from the ED did not return with adverse outcomes. Acute
patients with free fluid and a concomitant surgical condi- appendicitis is known to be more common in male
tion identified on imaging, the presence of free fluid was patients, and this situation was found to be no different in
strongly correlated with a surgical outcome. Since our our study. Nearly all of the male patients who went to sur-
focus was on children with just free fluid and no other gross gery had proven acute appendicitis on pathologic exami-
surgical pathologic findings such as acute appendicitis on nation (90.9%). A relatively large prospective study done
radiologic examination, we conducted a detailed subgroup by Sivit6 involved 250 children with acute abdominal pain
analysis in this cohort based on age and sex. We grouped and 50 control participants, and graded compression
older female patients separately in our subgroup analysis, sonography was done for detailed assessment of the lower
as physiologic free fluid related to gynecologic issues is abdomen and pelvis. The results of that study were similar
commonly noted on imaging in this age group. Surpris- to our results and found that the presence of free peritoneal
ingly, even when outcomes in this group of older female fluid detected by sonography in children with acute
patients were analyzed separately, we found that the quan- abdominal pain was largely a nonspecific finding. To our
tity of free fluid on imaging in the absence of an associated surprise, while reviewing published literature for our
surgical condition was not significantly related to a surgical research, we did not find a standard well-accepted method
outcome. In our subgroup analysis, the only 2 patients who for free fluid volume quantitation and stratification; hence,

Table 3. Comparison of Female Patients 10 Years or Older (Subgroup IIa) Versus Female Patients Younger Than 10 Years and All Male Patients
(Subgroup IIb)

Subgroup IIa Subgroup IIb


Outcome (n = 51) (n = 29) P
Surgery done, n (%) 1 (2.0) 4 (13.8) .036a
ED disposition, n (%) .002b
Inpatient admission and surgery later 0 (0.0) 2 (6.9)
Inpatient admission for observation only 16 (31.4) 18 (62.1)
Discharged from ED 34 (66.7) 7 (24.1)
Transferred to operating room from ED 1 (2.0) 2 (6.9)
Return to ED after discharge for same condition 0 (0.0) 0 (0.0)

Subgroup IIa Subgroup IIb


Comparison by Fluid Volume Stratification (CT) (n = 19) (n = 15) P
Estimated size of fluid collection on CT, n (%) .056b
Trace 0 (0.0) 0 (0.0)
Small 3 (15.8) 0 (0.0)
Medium 3 (15.8) 0 (0.0)
Large 13 (68.4) 15 (100.0)

Subgroup IIa Subgroup IIb


Comparison by Fluid Volume Stratification (Sonography) (n = 28) (n = 10) P
Estimated size of fluid collection on sonography, n (%) .11b
Trace 0 (0.0) 1 (10.0)
Small 14 (50.0) 2 (20.0)
Medium 5 (17.9) 1 (10.0)
Large 9 (32.1) 6 (60.0)

Most patients in subgroup IIa (66.7%), who had free fluid without an associated condition, were discharged from the ED and did not return with
complications.
aContinuity-corrected 2 test of a 2 2 table.
b 2 test of a 2 n contingency table, where 2 is the number of groups, and n is the number of categories.

1552 J Ultrasound Med 2013; 32:15471553


3209jum_Layout 1 8/20/13 11:43 AM Page 1553

Matz et alFree Intraperitoneal Fluid in Pediatric Patients With Abdominal Pain

we developed our own stratification method based on age 4. Holmes JF, London KL, Brant WE, Kuppermann N. Isolated intraperi-
and sex. Interestingly, all our patients in subgroup IIb toneal fluid on abdominal computed tomography in children with blunt
(female patients <10 years and male patients) had a large trauma. Acad Emerg Med 2000; 7:335341.
amount of free fluid on diagnostic CT, compared to 68.4% 5. Christiano JG, Tummers M, Kennedy A. Clinical significance of isolated
in subgroup IIa (female patients >10 years), and this differ- intraperitoneal fluid on computed tomography in pediatric blunt abdom-
ence approached significance. Even though these patients inal trauma. J Pediatr Surg 2009; 44:12421248.
did not have surgical diagnoses suggested on imaging, many 6. Sivit CJ. Significance of peritoneal fluid identified by ultrasonographic
did have major nonsurgical conditions such as gastroenteri- examination in children with acute abdominal pain. J Ultrasound Med
tis and pyelonephritis, as mentioned in Table 2. 1993; 12:743746.
Secondary outcomes of this study included emer- 7. Venkatesh KR, McQuay N Jr. Outcomes of management in stable chil-
gency ED disposition, complications, and return visits to dren with intra-abdominal free fluid without solid organ injury after blunt
the ED for the same condition after ED discharge. A sig- abdominal injury. Trauma 2007; 62:216220.
nificant number of older female patients were discharged
from the ED (66.7%); the etiology of abdominal pain in
these patients was predominantly gynecologic. Of the 78
patients who were discharged without surgery, none
returned to the ED within a 2-week period requiring sur-
gery for the same condition or with complications related
to delays in treatment. Our study findings indicate that
patients presenting with acute abdominal pain without an
antecedent history of trauma who have isolated free fluid
detected without definitive surgical conditions on imaging
are highly unlikely to have surgical outcomes, especially in
absence of overt signs of an abdominal condition requir-
ing surgery on clinical examinations.
Our study had several limitations. Since it was a ret-
rospective study, images were not reviewed in real time,
which may have influenced the volume quantitation done
by retrospective review of images, especially for sono-
graphic studies. In addition, the imaging studies were done
by different radiology staff members without prestudy
standardization and were conducted at a single center.
Since the literature in this area is so limited, we anticipate
that larger prospective studies involving multiple centers
will be needed in the future for further validation of our
results.

References

1. Simanovsky N, Hiller N, Lubashevsky N, Rozovsky K. Ultrasonographic


evaluation of the free intraperitoneal fluid in asymptomatic children.
Pediatr Radiol 2010; 41:732735.
2. Rathaus V, Grunebaum M, Konen O, et al. Minimal pelvic fluid in asymp-
tomatic children: the value of the sonographic finding. J Ultrasound Med
2003; 22:1317.
3. Rodriguez C, Barone JE, Wilbanks TO, Rha CK, Miller K. Isolated free
fluid on computed tomographic scan in blunt abdominal trauma: a
systematic review of incidence and management. J Trauma 2002; 53:79
85.

J Ultrasound Med 2013; 32:15471553 1553