Journal of Pediatric Surgery (2008) 43, 22422245

www.elsevier.com/locate/jpedsurg

An evidence-based definition for perforated appendicitis

derived from a prospective randomized trial
Shawn D. St. Peter , Susan W. Sharp, George W. Holcomb III, Daniel J. Ostlie
Department of Surgery, The Children's Mercy Hospital, Kansas City, MO 64108, USA

Received 27 August 2008; accepted 29 August 2008

Key words:
Abstract
Definition;
Purpose: Appendicitis is the most common urgent condition in general surgery, and yet there is no
Perforation;
evidence-based definition for perforation. Therefore, all retrospective data published on perforated
Appendicitis;
appendicitis are unreliable because of an ill-defined denominator. For approximately 2 years beginning
Children
in April 2005, we performed a prospective randomized trial investigating 2 different antibiotic regimens
for perforated appendicitis. During this study, we strictly defined perforation as a hole in the appendix
or a fecalith in the abdomen. Before this prospective study, perforation was staff surgeon opinion. We
investigated the abscess rates in both the perforated and nonperforated appendicitis populations before
and during the study to determine if our definition was safe and that there was not an increased risk of
abscess formation in patients treated as nonperforated.
Methods: Records of all patients undergoing laparoscopic appendectomy for appendicitis during the
immediate 2 years before using the definition were compared to those treated in the 2 years after the
definition was implemented. Interval and incidental appendectomies were ruled out. The postoperative
abscess rate (when perforation was not defined) was compared to the abscess rate of those for whom
perforation was strictly defined.
Results: There were 292 patients treated for acute nonperforated appendicitis in the 2 years before the
definition and 388 patients after the definition. There were 131 patients treated for perforated
appendicitis before the definition and 161 after the definition was implemented. The abscess rate in
those with perforated appendicitis increased from 14% to 18% after the definition was used. However,
after the definition began to be used, the abscess rate for those patients treated as nonperforated
decreased from 1.7% to 0.8%.
Conclusions: Defining perforation as a hole in the appendix or a fecalith in the abdomen is effective in
identifying the patients at risk for postoperative abscess formation. Application of these criteria would
allow substantial reduction in therapy for patients with purulent or gangrenous appendicitis who do not
possess the same abscess risk. These data outline the first evidence-based definition of perforation.
2008 Elsevier Inc. All rights reserved.

Appendicitis is the most common urgent condition in

Presented at 41st annual meeting of the Pacific Association of
Pediatric Surgeons, Jackson Lodge, Grand Teton National Park, Wyoming,
June 29-July 3, 2008.
Corresponding author. Tel.: +1 816 234 3575; fax: +1 816 983 6885.
E-mail address: sspeter@cmh.edu (S.D. St. Peter).
general surgery, yet there is no evidence-based definition for
perforation. Therefore, all retrospective data published on
perforated appendicitis are unreliable because of an ill-
defined denominator. In April of 2005, we initiated a

0022-3468/$ see front matter 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jpedsurg.2008.08.051
Perforated appendicitis
prospective, randomized trial investigating 2 different
antibiotic regimens [1]. The primary outcome variable in
this study was postoperative length of hospitalization. The
most import clinical parameter that influences time in the
hospital after appendectomy is the presence or absence of a
postoperative abscess. To assure the study was conducted on
a population of patients with an equal risk of developing a
postoperative abscess, we strictly defined perforation as a
hole in the appendix or a fecalith in the abdomen. Before the
study, patients with gangrenous, necrotic, or purulent
appendicitis were often treated as perforated category,
whereas they were not after the study began. Because we
had data from our appendicitis experience before the
initiation of this strict definition of perforation, we have a
model for investigating the influence of this strict definition.
Developing objective criteria could allow substantial reduc-
tion in therapy for patients with purulent or gangrenous
appendicitis who do not possess the same abscess risk. We
therefore investigated the abscess rates in perforated and
nonperforated populations before and during the study to see
if this definition is safe and that there was not an increased risk
of abscess development in patients treated as nonperforated.
1. Methods
1.1. Definition of perforation
A prospective, randomized trial on perforated appendicitis
was initiated in April 2005. A strict definition of perforation
was used as a hole in the appendix or a fecalith in the
abdomen. The definition has continued to be our criteria for
determination of perforation because this study was initiated.
The complete protocol of the prospective trial, including the
results, have been previously reported [1].
1.2. Postoperative management
During the 2 years before the application of the definition,
all patients treated as perforated appendicitis followed the
same protocol as was used in the study after the definition.
Therefore, all patients treated as perforated appendicitis in
both study arms received a minimum 5 days of intravenous
antibiotics. On postoperative day 5, a white blood cell count
was performed. If this was normal, patients were discharged
home without antibiotics. If the white blood cell count
remained elevated, the patient received a minimum 2 more
days of intravenous antibiotics after which therapy was
directed by the results of further studies such as a computed
tomographic scan.
Both before and after the definition was used, all patients
treated as nonperforated appendicitis were discharged home
without antibiotics when tolerating a regular diet with good
pain control on oral pain medications. This is typically the
day after the laparoscopic appendectomy.
2243
1.3. Study design
After approval from our institutional review board (no.
07-02-033X), a retrospective review was conducted on all
patients treated for appendicitis from April 2003 to April
2007. Data obtained from patients treated in the prospective
study (April 2005 to November 2006) were abstracted from
the existing database.
Patients who underwent laparoscopic appendectomy for
appendicitis during the immediate 2 years before using the
strict definition were compared to those treated in the 2
years after this strict definition was implemented. Those
treated as nonperforated appendicitis before the strict
definition were compared to those treated as nonperforated
after the strict definition. The same comparison was done
for patients managed as perforated before and after the strict
definition. Interval and incidental appendectomies were
ruled out. All patients underwent laparoscopic appendect-
omy. All patients received the same postoperative manage-
ment as described previously.
Descriptive statistics are listed as mean SD.
2. Results
2.1. Patients
During the 2 years before using the strict definition, there
were 292 patients treated for acute appendicitis compared to
388 patients managed after the strict definition. There were
131 patients treated as perforated before the strict definition
and 161 after the strict definition.
2.2. Demographics
In patients treated as nonperforated before the strict
definition, the mean age was 8.2 3.6 years and 57% were
male compared to those treated as nonperforated after the
strict definition who had a mean age of 10.8 3.8 years with
62% being male. In patients treated as perforated appendi-
citis before the strict definition, the mean age was 9.4 4.2
years and 57% were male compared to those treated as
perforated after the strict definition that had a mean age of
9.0 4.1 years with 58% being male.
2.3. Outcomes
The abscess rate before and after the strict definition was
implemented was determined for both perforated and
nonperforated appendicitis and is shown in Table 1. There
was an increase in the abscess rate in those with perforated
appendicitis during the period that the strict criteria for
perforation were used. Importantly, the abscess rate for those
treated as nonperforated did not increase during the latter
2244
Table 1 Impact of strict definition of perforation on abscess
rate
Before definition After definition
Abscess rate (%) Abscess rate (%)
Acute appendicitis 1.7 0.8
Perforated appendicitis 14.0 18.0
period. The corresponding length of postoperative hospita-
lization for these patient groups is listed in Table 2.
3. Discussion
Appendicitis as an entity, and perforated appendicitis
specifically, has been studied extensively in both the
pediatric and adult literatures [2-6]. Unfortunately, these
studies, whether retrospective or prospective, all lack the
single most important clinical factan accurate definition of
perforation. The lack of definition likely accounts for the
large variability in reported perforation rates. A recent review
on this topic documented a dramatic range for perforation
from 20% to 76% [7]. The authors of this study emphasized
that there is no clear explanation for this extreme variability.
However, we feel this variability in diagnosis is based on an
unreliable definition for perforation. The members of the
American Pediatric Surgical Association were surveyed in
2000, and most members reported that they based their
practice approaches on their individual preferences [4].
Without an accurate and proven definition for perforation,
the data that are currently available to practicing surgeons
and physicians regarding management and treatment of
perforated appendicitis are not consistent. The resultant
variability in care has direct effects on quality of care and
subsequent complications, hospital costs, and lengths of
hospitalizations [7,8]. An accurate and proven definition
equips the surgeon with information for treatment, produces
reliable data for counseling patients and families, and allows
for institutions to compare results in a meaningful fashion.
In this study, we compared our initial experience with
individual surgeon-dependent diagnosis of perforation to our
experience with absolute criteria in place. Before instituting
the criteria of either seeing a fecalith in the abdomen or
physically documenting a hole in the appendix, the operating
surgeon would determine perforation based on his experi-
ence and opinion. This meant that the entire spectrum of
appendicitis was susceptible to a diagnosis of perforation
including gangrenous, purulent, or near perforated. In
addition, patients with purulent peritoneal fluid (but no
perforation) were often placed into the perforated category as
well. We have clinically found that pathologic evaluation is
less useful as we see patients with a fecalith in the abdomen
or an obvious hole detected by placing an instrument inside
the appendix and visualizing the instrument through a hole in
S.D. St. Peter et al.
which the pathologic report makes no comment on
perforation. Therefore, we feel the issue of perforation is
greatly simplified when the surgeon either finds an
identifiable hole in the appendix or removes a fecalith
from the peritoneal cavity.
More important than identifying a common term for
perforation, we need to identify the patients who hold a
high risk of postoperative abscess. This is why the
evolution of our working definition of perforation is
based entirely on abscess rates before and after the
definition was used. The trends in the abscess rates strongly
support the accuracy of this definition for perforated
appendicitis. If the definition were too strict, the abscess
rate for the nonperforated appendicitis group should have
increased because of inadequate treatment of contaminated
peritoneal cavities. This did not happen as the 1.7% abscess
rate in the former group dropped to 0.8% after the
institution of the strict definition. Second, assuming
equivocal antibiotic treatment, the abscess rate in the
perforated appendicitis group after the institution of the
strict definition should stay the same or increase. This
would be the effect of eliminating patients who previously
were included in the perforated cohort but were not really
perforated and therefore did not carry the elevated abscess
risk. This affect likely accounts for the increase in abscess
rate in the perforated group from 14% to 18% with the
application of the strict definition.
Regarding statistical evaluation, the shifts in abscess rates
depicted in Table 1 were the outcomes used to determine if
the applied definition was leading to undertreatment for
patients with necrotic, gangrenous, or severely purulent
disease who would have been treated as perforated by most
surgeons without this definition. Therefore, the most
important number in this article is the abscess rate in patients
treated as nonperforated. We were simply concerned about
the presence or absence of a rise in this rate after the
definition was applied. If there had there been a rise in
abscesses seen in nonperforated patients, then a P value
would be applicable to assess the significance of such a rise.
This is because a significant rise would mean we need to treat
patients with necrotic, gangrenous, and purulent appendicitis
as perforated with our standard prolonged course of
intravenous antibiotics. However, in the absence of a rise,
or as occurred in this study, a drop in abscess rate, this means
that nonperforated patients are not being undertreated with
Table 2 Impact of strict definition of perforation on length of
postappendectomy hospital stay
Before definition After definition
Length of stay, Length of stay,
d (mean SD) d (mean SD)
Perforated appendicitis 9.4 4.2 7.4 8.8
Acute appendicitis 1.9 1.3 1.5 1.5
Perforated appendicitis
this definition. We are not concerned if the drop is
statistically significant because a P value holds no meaning
relevant to this article, and it could be misleading. We do not
want to suggest that this strict definition will cause a
significant drop in abscess rates for nonperforated patients
because we know this is not true because patients with more
severe disease are treated as nonperforated under this
definition. More likely, the abscess rate in this group dropped
with advancing laparoscopic experience at our institution.
The contribution from this work is simply to show that a
definition that will treat many patients with advanced
appendicitis as nonperforated does not have a negative
impact on those patients and in fact, avoids overtreatment.
The observed increase in abscess rate in patients treated as
perforated under this definition is expected but not important
to this study. Therefore, the quantity of increase is not
meaningful. Finally, the percentage of patients who were
considered perforated both before and after the definition
was not analyzed because there is no way we can know how
many of those termed as perforated before definition actually
had a perforation. In addition, as a referral center for multiple
facilities over a large area, there are year-to-year changes in
diagnostic systems and surgical coverage in the community
that affects the number of perforated cases seen between 2
separate time frames.
We did not further evaluate the overall length of antibiotic
treatment in this study because it has been previously
published [1]. We did include the length of hospitalization to
point out that regardless of the status of appendicitis
(perforated or acute) the length of postoperative hospitaliza-
tion has decreased. We attribute this to the institution of an
algorithmic approach to the management of appendicitis that
includes a strict definition for perforation.
The importance of this definition for literature discussing
appendicitis cannot be overstated. Outcomes research
focusing on perforated appendicitis is not clinically mean-
ingful when the degree of disease comprising the denomi-
2245
nator remains unknown and variable. We are fortunate to
have a prospective data set collected with absolute criteria in
place for the definition. Against a historical control group,
these data demonstrate that our working definition can be
safely used without a detectable risk of undertreating the
patients defined as nonperforated. This definition can
improve the overall care for children with appendicitis as it
can eliminate unnecessary antibiotic treatment, improve cost
management, simplify treatment protocols, improve the
integrity of data used in family/patient counseling, and
allow for ongoing clinical research with results that are
useful between institutions.
References
[1] St. Peter SD, Tsao K, Spilde TL, et al. Single daily dosing ceftriaxone
and Flagyl versus standard triple antibiotic regimen for perforated
appendicitis in children: a prospective, randomized trial. J Pediatr Surg
2008;43:981-5.
[2] Henry MCW, Walker A, Silverman BL, et al. Risk factors for the
development of abdominal abscess following operation for perforated
appendicitis in children. Arch Surg 2007;142:236-41.
[3] Newman KD, Morrow SE. Current management of appendicitis. Semin
Pediatr Surg 2007;16:34-40.
[4] Chen C, Botelho C, Cooper A, et al. Current practice patterns in the
treatment of perforated appendicitis in children. J Am Coll Surg 2003;
196:212-21.
[5] Kokoska ER, Silen ML, Tracy TF, et al. Perforated appendicitis in
children: risk factors for the development of complications. Surgery
1998;124:619-26.
[6] Azzie G, Salloum A, Beasley S, et al. The complication rate and
outcomes of laparoscopic appendectomy in children with perforated
appendicitis. Pediatr Endosurg Innovative Tech 2004;8:19-23.
[7] Newman K, Ponsky T, Kittle K, et al. Appendicitis 2000: variability in
practice, outcomes, and resource utilization at thirty pediatric hospitals.
J Pediatr Surg 2003:372-9.
[8] Tan EC, Severijnen RS, Rosman C, et al. Diagnosis and treatment of
acute appendicitis in children: a survey among Dutch surgeons and
comparison with evidence based practice. World J Surg 2006;30:512-8.