The American Journal of Surgery 191 (2006) 497–502

Clinical surgery–American

Classifying patients suspected of appendicitis with regard to likelihood

Robert H. Birkhahn, M.D.*, Matthew Briggs, Ph.D., Paris A. Datillo, R.N.,
Shawn K. Van Deusen, M.D., Theodore J. Gaeta, D.O., M.P.H.
Department of Emergency Medicine, New York Methodist Hospital, 506 6th St. Brooklyn, NY 11215, USA

Manuscript received March 2, 2005; revised manuscript May 30, 2005

Abstract
Background: We sought to develop a clinical predictive model for acute appendicitis and contrast it with current clinical practice.
Methods: A prospective observational study of patients presenting with signs or symptoms consistent with acute appendicitis. Random-
partition modeling was used to develop an appendicitis likelihood model (ALM).
Results: Four hundred thirty-nine patients were enrolled, 101 with appendicitis, and 338 with other diagnoses. The ALM classified patients
as “low likelihood” if they had a white blood cell count ⬍9,500 and either no right lower– quadrant tenderness or a neutrophil count ⬍54%.
Patients were classified as “high likelihood” if they had a white blood cell count ⬎13,000 with rebound tenderness or both voluntary
guarding and neutrophil count ⬎82%. The ALM outperformed actual clinical practice with regard to “missed” appendicitis, negative
laparotomies, and total number of imaging studies.
Conclusion: The ALM may permit more judicious use of advanced radiographic imaging with lower nontherapuetic laparotomy rates.
© 2006 Excerpta Medica Inc. All rights reserved.

Keywords: Appendicitis; Clinical scoring; Diagnosis

Appendicitis is one of the most common reasons for emer-
gent laparotomy, with a nearly 7% lifetime cumulative
incidence [1,2]. Prompt and accurate diagnosis is imperative
to decrease the frequency of complications induced by ap-
pendiceal rupture.
The evaluation of patients with suspected acute appen-
dicitis is clinically complex. A constellation of history,
physical signs, radiographic investigation, and laboratory
analysis is used to balance the risk of delayed operative
intervention against the removal of a normal appendix (non-
therapuetic laparotomy).
The cornerstone of the diagnosis of acute appendicitis
has traditionally been the combination of history and phys-
ical examination. The overall accuracy for the clinical ex-
amination in diagnosing acute appendicitis has been re-
ported to be 70% to 87% (54% to 70% in children and 50%
to 70% in women of childbearing age) [3– 8]. Advanced
radiographic imaging, ultrasound, and abdominal computed
axial tomography (CAT) have been shown to be accurate in
* Corresponding author. Tel.: ⫹1-718-780-5040; fax: ⫹1-718-780-
3153.
E-mail address: rhbirkhahn@pol.net
the evaluation of suspected appendicitis [9 –16]. The use of
advanced imaging for cases of suspected appendicitis has
been reported to be more frequent in recent studies [7,17].
Some investigators have advocated the use of preoperative
imaging (ultrasound or CAT) in all cases of suspected
appendicitis because of the high accuracy, the ability of
advanced imaging to identify alternative diagnoses, and the
utility of imaging in select populations (eg, pediatric, geri-
atric, and women of childbearing age) [18 –21]. Others have
argued for the judicious use of advanced imaging, pointing
out that imaging is not without cost in terms of resource
utilization, operative delay, and potential for increased per-
foration rate [6,17,22–25]. The compilation of signs and
symptoms into algorithms and scoring systems in the eval-
uation of suspected appendicitis is not new, but previous
scoring systems alone have not significantly improved the
overall accuracy of diagnosing acute appendicitis [26 – 40].
The focus of our investigation was to derive an appen-
dicitis likelihood model (ALM) and compare it with actual
clinical practice (ACP) in a population presenting for the
acute evaluation of right lower– quadrant (RLQ) pain. Spe-
cifically, our objective was to streamline the use of ad-
vanced imaging in patients with suspected appendicitis by

0002-9610/06/$ – see front matter © 2006 Excerpta Medica Inc. All rights reserved.
doi:10.1016/j.amjsurg.2005.08.031
498 R.H. Birkhahn et al. / The American Journal of Surgery 191 (2006) 497–502
determining if a simple clinical rule could accurately iden-
tify patients with such a high or low likelihood of appen-
dicitis that imaging was not warranted before operative
intervention or discharge.
Materials and Methods
We conducted a prospective observational study that
enrolled all patients who presented with signs or symptoms
consistent with acute appendicitis. Patients were enrolled
from our urban Emergency Department (ED) based in a
community teaching hospital (with an annual census of
68,000 patients) from February 2003 to December 2003.
Trained research assistants based in the ED for 16 hours
a day, 7 days a week, identified subjects. To obtain data on
eligible patients who may have been missed, the ED charts
were reviewed by chief complaint of abdominal pain and
cross-referenced with a standardized charting system on a
daily basis during the study period.
Patients of all ages were enrolled if they had either signs
or symptoms consistent with acute appendicitis. Patients
were considered to have symptoms consistent with acute
appendicitis if they presented for evaluation with a history
of acute abdominal pain (⬍7 days in duration) localizing to
the periumbilical region, suprapubic region, or RLQ. Pa-
tients were considered to have physical signs consistent
with acute appendicitis if on examination by the attending
physician they were found to have point tenderness in the
RLQ or clinical evidence of peritonitis (voluntary or invol-
untary abdominal guarding or rebound tenderness).
Patients were excluded from the study if they had an
obvious cause of abdominal pain on physical examination
(eg, penetrating trauma, postoperative pain), known human
immunodeficiency virus infection, or history of appendec-
tomy. To make the population more clinically relevant,
patients were also excluded from the study if the diagnosis of
appendicitis was not considered in the clinical differential (ie,
the treating clinician did not order laboratory investigation).
Research assistants completed a standardized data col-
lection instrument on patients who were willing and able to
provide informed consent for participation. This data form
was developed using guidelines from the American College of
Emergency Physicians [41,42]. Elements of the history were
obtained directly from the patient. The research assistant ob-
tained elements of the physical examination from the attending
physician before laboratory or radiographic evaluation.
Guarding was a subjective assessment specific to the
clinician if the abdominal wall was felt to be tense on
examination and was classified as either voluntary or invol-
untary based on whether or not the clinician believed that
the tension could be relieved by mental distraction of the
patient. Rebound tenderness (evidence of peritoneal irrita-
tion) was defined as the elicitation of pain on removal of a
hand from the palpated abdomen. Although different clini-
cians have their own manners of eliciting rebound tender-
ness; if the evaluating clinician believed there was evidence
on physical examination of peritoneal irritation, the patient
was coded as having rebound tenderness. Vital signs (ie,
temperature, heart rate, and blood pressure) were obtained
at ED triage.
Enrolled patients had evaluation and treatment plans
determined by the discretion of the treating physician with-
out regard to study enrollment. Data regarding radiographic
imaging use and results, diagnostic pathology, total white
blood cell (WBC) count, urine human chorionic gonadotro-
pin, ketonuria, and neutrophil count were collected.
Patients were contacted at 7 days after ED presentation.
Contact was attempted by way of telephone or email for up
to 2 months after the index ED visit. Patients were specif-
ically asked about the resolution of their presenting symp-
toms, the need to revisit the ED, hospital readmission, and
the need for subsequent operation. Patients who could not
be contacted within this time frame were considered lost to
follow-up.
For the purposes of data analysis, patients were divided
into two groups: appendicitis present and appendicitis ab-
sent. Patients who underwent operative exploration for sus-
pected appendicitis were categorized based on the results of the
surgical pathology. Patients with evidence of acute appendici-
tis at pathology were considered “appendicitis present.” Pa-
tients with a normal appendix at surgical pathology were con-
sidered “appendicitis absent.” Patients who underwent surgical
exploration for reasons other than for presumed appendicitis
(eg, ectopic pregnancy, colon cancer) had this documented
in the database but were considered “appendicitis absent”
for the primary analysis.
Patients who did not undergo operative intervention were
followed-up longitudinally to determine resolution of symp-
toms without subsequent operative intervention at either our
institution or another. These patients were classified as “ap-
pendicitis absent” if symptomatology resolved after the
7-day period of observation after first presentation. If a
patient required subsequent hospital admission but their
symptoms resolved without operative intervention, it was
presumed that they did not have acute appendicitis, and they
were classified as “appendicitis absent.” This definition
mimics the clinical practice of admission with serial obser-
vation for resolution or worsening of symptoms.
The actual use of advanced imaging and operative
choices is reported as ACP with 95% confidence intervals
(CIs) for proportions used as reference. At our institution,
patients are initially evaluated by an ED trained physician,
and if a suspicion of acute appendicitis exists, a surgical
consult is obtained. Senior surgical house staff members
provide surgical consultation 24 hours a day, 7 days a week,
at the direction of a surgical attending. Advanced radio-
graphic imaging in patients with suspected appendicitis is
typically obtained after surgical consultation, but it may be
ordered by the ED physician alone. Patients were consid-
ered “missed” in ACP if they had an operative delay of ⬎24
hours from initial presentation.
 R.H. Birkhahn et al. / The American Journal of Surgery 191 (2006) 497–502 499

The ALM was developed specifically to identify (1) Table 1

patients who had such a high likelihood of acute appendi- Patient characteristics
citis that they should receive exploratory laparotomy with- Finding Appendicitis Non appendicitis
out delay and (2) patients who had such low likelihood of (n ⫽ 101) (n ⫽ 338)
acute appendicitis that they did not require abdominal CAT. Mean age (in y range) 31 (3–87) 36 (3–93)
By default, our decision rule would classify all patients not Age ⬍18 y (%) 12 16
meeting criteria for either high or low probability as an Male (%) 61 27
Race (%)
intermediate-probability cohort that should receive ad-
White 52 32
vanced imaging before disposition. Black 12 31
The ALM was derived with all collected patient charac- Hispanic 26 27
teristics (history, physical examination, and laboratory), us- Asian 4 4
ing statistical random-partition (tree) models to develop a Other* 6 6
Mean WBC count (K/uL) 14.8 9.2
clinical decision rule [43]. Random-partition models grow a
95% CI (13.9–15.8) (9.0–9.4)
decision tree by finding, for each decision point, the optimal Mean neutrophil count (%) 82 68
variable and value of that variable that maximizes the cor- 95% CI 80–84 66–70
rect number of classifications subject to a user-provided loss % Pregnant (qualitative urine HCG) 0 14
matrix. Once the tree is fully grown, it is “pruned” to avoid Abdominal CT obtained (%) 65 55
Abdominal sonogram obtained (%) 3 32
overfitting. It does this by a cross-validation procedure
Operation performed (%) 100 9†
designed to decrease error in independent data sets. Every
attempt was made, by way of the cost-complexity pruning CAT ⫽ computed axial tomography; CI ⫽ confidence interval; HCG ⫽
human chorionic gonadotropin; WBC ⫽ white blood cell.
of the random partition model, to avoid overfitting on our
* “Other” includes American Indian, pacific islander, and nonresponsive.
particular data set. † 32 patients without appendicitis underwent operative intervention (15
The primary analysis was the performance of the each with other pathology and 17 with nontherapeutic laparotomy).
scoring system with regard to the number of patients who
would have received a nontherapeutic laparotomy in the
high-likelihood category (false positives), the total number appendicitis; 43 were excluded (8 trauma, 21 previous ap-
of patients who would have received delayed appendectomy pendectomy, and 14 with known human immunodeficiency
(miss rate) in the low-likelihood group (false negatives), and virus); and 42 patients were missed or refused to consent for
the total number of imaging studies recommended in the enrollment; the result was 439 patients (91%) available for
intermediate-probability group. analysis.
The nontherapuetic laparotomy rate was calculated by Follow-up was completed successfully for 426 patients
dividing the number of patients with no pathologic evidence (97%); 34 patients had a repeat ED visit within 7 days (8
of acute appendicitis by the total number of patients under- patients visited another institution); 10 were admitted to the
going surgery specifically to identify acute appendicitis (in- hospital; and none underwent surgery. There were 101 pa-
cluding 1 patient with unanticipated appendicitis found at tients with pathologically confirmed appendicitis, and 338
exploration). Patients with mesenteric adenitis were consid- patients had no evidence of acute appendicitis on follow-up;
ered to have no appendiceal pathology. The 15 patients who full characteristics are listed in Table 1. The distribution of
underwent operative exploration for a condition other than final diagnoses seen in the study cohort is shown in Fig. 1.
acute appendicitis (eg, ectopic pregnancy, colon cancer, When using the software to develop a rule to identify
tubo-ovarian abscess) were not included in the calculation high-likelihood patients, we instructed the random-partition
of the nontherapuetic laparotomy rate. model that a false-positive outcome (nontherapuetic lapa-
It was calculated that a sample size of 98 patients with rotomy) was to be weighted twice as bad as a false-negative
confirmed acute appendicitis was needed to attain a 95% CI outcome (appendectomy delayed by imaging). Varying the
width of 10%. With an anticipated prevalence of 25% in our weighting beyond a 2:1 ratio resulted in a lower nonthera-
population, we estimated that 392 patients would be re- puetic laparotomy rate at the cost of dramatically increased
quired to develop the model. This protocol was reviewed and imaging in patients with true appendicitis. When using the
approved by the Institutional Review Board. All statistical software to develop a rule for the low-likelihood patients,
calculations, including the random-partition model, were done where the decision point is to image or discharge, we found
using R: A Language and Environment for Statistical Com- that weighting a false-positive diagnosis (discharged true
puting, v1.8.1. (R Development Core Team, Vienna, Austria). appendicitis) as 20 times worse than a false-negative diag-
nosis (imaging a normal appendix) decreased the missed
appendicitis rate to nearly zero.
Results The random-partition model found that total WBC was
the single most important contributor to the overall likeli-
During the study period, 524 patients were screened for hood of acute appendicitis. By weighting to minimize false-
enrollment with signs or symptoms consistent with acute positive diagnoses, patients who had a high likelihood of
500 R.H. Birkhahn et al. / The American Journal of Surgery 191 (2006) 497–502

Fig. 2. Appendicitis likelihood model (ALM).

intervention for acute appendicitis ⬎24 hours after ED

Fig. 1. Final diagnosis observed in study cohort.
appendicitis were identified by a WBC count ⬎13,000 with
either rebound tenderness on physical examination or a
neutrophil count ⬎82% and voluntary guarding on physical
examination. The low-likelihood population could be sum-
marized by those patients with a WBC count ⬍9,500 with
either a neutrophil count ⬍54% or no localized tenderness
in the RLQ (Fig. 2).
The classification of patients by the ALM and in ACP is
listed in Table 2. The one patient with acute appendicitis
that the ALM classified as low likelihood was an obese
female with a right inguinal hernia and tenderness below the
inguinal ligament on examination but no RLQ tenderness.
This patient went to the operating room for a presumed
incarcerated hernia and was found to have a necrotic ap-
pendix.
Evaluation of the ALM showed that it performed better
than ACP with regard to both nontherapuetic laparotomy
and missed appendicitis rates (Table 3). Ten patients were
considered “missed” in this analysis, defined as operative
presentation. All of these patients had surgical consultation
in the ED, and 8 were believed to be nonsurgical at presen-
tation and were admitted for serial observation. The 2 re-
maining patients were imaged and suspected of acute ap-
pendicitis in the ED, but delays in imaging and access to the
operative suite resulted in a total delay ⬎24 hours.
Fifteen patients required operative intervention for rea-
sons other than acute appendicitis (5 cases of ectopic preg-
nancy, 1 case of ovarian torsion, 3 cases of acute cholecys-
titis, 1 case of surgical diverticulitis, 1 case of colon cancer
with perforation, and 4 cases of nonperforated colon can-
cer). All 15 these patients would have been classified by the
ALM as intermediate likelihood, suggesting that advanced
radiographic imaging be obtained before disposition.
Comments
No major medical association or professional organiza-
tion currently endorses a standardized pathway for the eval-
uation of patients with suspected appendicitis. The role of
advanced radiography (ultrasound and abdominal CAT) and
its impact on the nontherapuetic laparotomy rate and per-
foration rate remain open questions. One recent position
article suggested that patients with a “classic” history and
physical examination undergo explorative laparotomy with-
out delay, whereas patients with an “equivocal” presenta-

Table 2
ALM compared with ACP
ALM versus Patients with acute appendicitis Patients with normal appendix
ACP
To OR Image first Missed To OR Imaged No imaging
(n ⫽ 101) (n ⫽ 101) (n ⫽ 101) (n ⫽ 338) (n ⫽ 338) (n ⫽ 338)

ALM (%) 47 52 1 2 36 62
95% CI 37–56 43–62 0–5 1–4 31–41 57–67
ACP (%) 34 56 10* 5 73* 28
95% CI 24–44 46–66 5–17 3–8 68–78 23–33

ACP ⫽ actual clinical practice; ALM ⫽ appendicitis likelihood model; OR ⫽ operating room.
* Ten patients underwent surgery ⬎24 hours after presentation.
 R.H. Birkhahn et al. / The American Journal of Surgery 191 (2006) 497–502
Table 3
Performance of ALM compared with ACP
ALM versus No missed (%) No. negative Total no.
ACP laparotomy imaging
rate (%) obtained (%)
ALM 1/101 (1) 6/107 (5.6) 274 (62)
95% CI 0–5 2–12 58–67
ACP* 10/101 (10%) 17/118 (14%) 312 (71%)
95% CI 5–17 9–22 67–75
ACP ⫽ actual clinical practice; ALM ⫽ appendicitis likelihood model;
CI ⫽ confidence interval.
* A missed case of appendicitis in ACP was consider to be any patient
who went to the operating room ⬎24 hours after presentation.
tion undergo ultrasound if pregnant and abdominal CAT if
nongravid [44]. The use of advanced imaging in all equiv-
ocal patients would represent a significant deviation from
current practice, where imaging is arbitrarily left to the
discretion of the treating physician. Universal imaging of
equivocal patients might prevent operative delay in some
patients, but there is fear that such a policy would create a
dramatic burden on hospital resources and lead to increased
reliance on imaging as a surrogate for sound clinical deci-
sion making.
The ALM, developed out of this dataset, provides a
simple method that improves accuracy in predicting acute
appendicitis because of the hierarchical nature of the ran-
dom-partition model. The random-partition model identifies
variables that, taken alone or in combination, have signifi-
cance in identifying patients with acute appendicitis. We
found that WBC count was the single best predictor in
identifying patients with a high or low likelihood of appen-
dicitis. Several recent studies have questioned the clinical
utility of the WBC count in the evaluation of suspected
appendicitis, most frequently quoting a likelihood ratio that
was too low to alter clinical decision making. Calculating a
receiver-operating characteristic curve for our data showed
that the WBC count had an area under the curve (AUC) of
0.84 (95% CI 0.80 to 0.89), which is at the high end of
previous reports, which reported an AUC of 0.67 to 0.80
[45–50]. Previous reports were skeptical of the utility of the
WBC count because they considered the utility of the test by
itself to alter a previously established clinical probability
[48,49]. The strength of the random-partition model used in
this study is its ability to consider a number of clinical signs
and symptoms together with basic laboratory tests to estab-
lish a hierarchical approach to a given problem. That is, the
WBC count in our strategy was the most important step but
only the first step toward establishing probability of appen-
dicitis. The ALM was developed as a clinical tool that could
help establish a pretest probability of acute appendicitis.
We were unable to determine what influence the use of a
clinical prediction model would have on perforation rates.
Information on the timing of perforation, and the effect
implementation of a classification scheme would have on
the rate of perforation, is not available from this dataset.
501
The cross-validation techniques used to develop the
ALM are well established and should minimize “overfit-
ting” of our model to this independent data set. However,
testing this classification scheme on an independent data set
would provide stronger evidence with regard to the validity
of the ALM. The model should be tested not only on an
independent data set but also with specific regard to popu-
lations that have been historically difficult to diagnose
(young children, elderly, and pregnant women). However, it
is unlikely that this decision aid would be used in pregnant
women because few physicians would make a disposition
without the aid of advanced imaging.
In conclusion, the ALM is an easily implemented deci-
sion instrument that could have utility in classifying patients
with regard to their probability for acute appendicitis. The
ALM could be used to establish pretest probability of acute
appendicitis and guide the evaluation of patients after basic
laboratory investigation and urinalysis. The use of the ALM
has the potential to promote the efficient use of advanced
radiographic while decreasing the number of nontherapeutic
laparotomies performed.
References
[1] Addiss DG, Shaffer N, Fowler BS, et al. The epidemiology of ap-
pendicitis and appendectomy in the United States. Am J Epidemiol
1990;132:910 –25.
[2] Pittman-Waller VA, Myers JG, Stewart RM, et al. Appendicitis: why
so complicated? Analysis of 5755 consecutive appendectomies. Am
Surg 2000;66:548 –54.
[3] John H, Neff U, Kelemen M. Appendicitis diagnosis today: clinical
and ultrasonic deductions. World J Surg 1993;17:243–9.
[4] Wade DS, Marrow SE, Balsara ZN, et al. Accuracy of ultrasound in
the diagnosis of acute appendicitis compared with the surgeon’s
clinical impression. Arch Surg 1993;128:1039 – 46.
[5] Wilson DH, Wilson PD, Walmsley RG, et al. Diagnosis of acute
abdominal pain in the accident and emergency department. Br J Surg
1977;64:250 – 4.
[6] Bergeron E, Richer B, Gharib R, et al. Appendicitis is a place for
clinical judgment. Am J Surg 1999;177:460 –2.
[7] Styrud J, Eriksson S, Segelman J, et al. Diagnostic accuracy in 2,351
patients undergoing appendicectomy for suspected acute appendicitis:
a retrospective study 1986 –1993. Dig Surg 1999;16:39 – 44.
[8] Saidi RF, Ghasemi M. Role of Alvarado score in diagnosis and
treatment of suspected acute appendicitis. Am J Emerg Med 2000;
18:230 –1.
[9] Rao PM, Rhea JT, Rattner DW, et al. Introduction of appendiceal CT:
impact on negative appendectomy and appendiceal perforation rates.
Ann Surg 1999;229:344 –9.
[10] Balthazar EJ, Rofsky NM, Zucker R. Appendicitis: the impact of
computed tomography imaging on negative appendectomy and per-
foration rates. Am J Gastroenterol 1998;93:768 –71.
[11] Rao PM, Rhea JT, Novelline RA, et al. Helical CT technique for the
diagnosis of appendicitis: prospective evaluation of a focused appen-
dix CT examination. Radiology 1997;202:139 – 44.
[12] Schuler JG, Shortsleeve MJ, Goldenson RS, et al. Is there a role for
abdominal computed tomographic scans in appendicitis? Arch Surg
1998;133:373–7.
[13] Rettenbacher T, Hollerweger A, Gritzmann N, et al. Appendicitis:
should diagnostic imaging be performed if the clinical presentation is
highly suggestive of the disease? Gastroenterology 2002;123:992– 8.
502 R.H. Birkhahn et al. / The American Journal of Surgery 191 (2006) 497–502
[14] Sivit CJ, Applegate KE, Stallion A, et al. Imaging evaluation of
suspected appendicitis in a pediatric population: effectiveness of
sonography versus CT. AJR Am J Roentgenol 2000;175:977– 80.
[15] Poortman P, Lohle PN, Schoemaker CM, et al. Comparison of CT
and sonography in the diagnosis of acute appendicitis: a blinded
prospective study. AJR Am J Roentgenol 2003;181:1355–9.
[16] Garcia Pena BM, Mandl KD, Kraus SJ, et al. Ultrasonography and
limited computed tomography in the diagnosis and management of
appendicitis in children. JAMA 1999;282:1041– 6.
[17] Garfield JL, Birkhahn RH, Gaeta TJ, et al. Diagnostic pathways and
delays on route to operative intervention in acute appendicitis. Am
Surg 2004;70:1010 –3.
[18] Rhea J, Halpern E, Ptak T, et al. The status of appendiceal CT in an
urban medical center 5 years after its introduction: Experience with
753 patients. AJR Am J Roentgenol 2005;184:1802–1808.
[19] Rettenbacher T, Hollewerger A, Gritzmann N, et al. Appendicitis:
should diagnostic imaging be performed if the clinical presentation is
highly suggestive of the disease? Gastroenterology 2002;123:992–998.
[20] Brandt MM, Wahl WL. Liberal use of CT scanning helps to diagnose
appendicitis in adults. Am Surg 2003;69:727–32.
[21] Rao PM, Rhea JT, Novelline RA, et al. Effect of computed tomog-
raphy of the appendix on treatment of patients and use of hospital
resources. N Engl J Med 1998;338:141– 6.
[22] Partrick DA, Janik JE, Janik JS, et al. Increased CT scan utilization
does not improve the diagnostic accuracy of appendicitis in children.
J Pediatr Surg 2003;38:659 – 62.
[23] Fuchs JR, Schlamberg JS, Shortsleeve MJ, et al. Impact of abdominal
CT imaging on the management of appendicitis: an update. J Surg
Res 2002;106:131– 6.
[24] Perez J, Barone JE, Wilbanks TO, et al. Liberal use of computed
tomography scanning does not improve diagnostic accuracy in ap-
pendicitis. Am J Surg 2003;185:194 –7.
[25] Torbati SS, Guss DA. Impact of helical computed tomography on the
outcomes of emergency department patients with suspected appendi-
citis. Acad Emerg Med 2003;10:823–9.
[26] Fenyo G, Lindberg G, Blind P, et al. Diagnostic decision support in
suspected acute appendicitis: validation of a simplified scoring sys-
tem. Eur J Surg 1997;163:831– 8.
[27] Fenyo G. Routine use of a scoring system for decision-making in
suspected acute appendicitis in adults. Acta Chir Scand 1987;153:
545–51.
[28] Malik AA, Wani NA. Continuing diagnostic challenge of acute appen-
dicitis: evaluation through modified Alvarado score. Aust N Z J Surg
1998;68:504 –5.
[29] Macklin CP, Radcliffe GS, Merei JM, et al. A prospective evaluation
of the modified Alvarado score for acute appendicitis in children. Ann
R Coll Surg Engl 1997;79:203–5.
[30] Alvarado A. A practical score for the early diagnosis of acute appen-
dicitis. Ann Emerg Med 1986;15:557– 64.
[31] Owen TD, Williams H, Stiff G, et al. Evaluation of the Alvarado
score in acute appendicitis. J R Soc Med 1992;85:87– 8.
[32] Ohmann C, Yang Q, Franke C. Diagnostic scores for acute appendi-
citis. Abdominal Pain Study Group. Eur J Surg 1995;161:273– 81.
[33] Kalan M, Talbot D, Cunliffe WJ, et al. Evaluation of the modified
Alvarado score in the diagnosis of acute appendicitis: a prospective
study. Ann R Coll Surg Engl 1994;76:418 –9.
[34] Kiely N, Williams N. Evaluation of the modified Alvarado score in
the diagnosis of acute appendicitis: a prospective study. Ann R Coll
Surg Engl 1995;77:157.
[35] Tobias M, Samuel E. Ultrasound and the Alvarado score. J R Soc
Med 1992;85:585– 6.
[36] Adedeji OA. Alvarado score and acute appendicitis. J R Soc Med
1992;85:508 –9.
[37] Gwynn LK. The diagnosis of acute appendicitis: clinical assessment
versus computed tomography evaluation. J Emerg Med 2001;21:
119 –23.
[38] Showstack J. Predictive power of the MANTRELS score. Ann Emerg
Med 1991;20:1048 –9.
[39] Graff L. The MANTRELS score. Ann Emerg Med 1991;20:712–3.
[40] Bond GR, Tully SB, Chan LS, et al. Use of the MANTRELS score in
childhood appendicitis: a prospective study of 187 children with
abdominal pain. Ann Emerg Med 1990;19:1014 – 8.
[41] Clinical policy: critical issues for the initial evaluation and manage-
ment of patients presenting with a chief complaint of nontraumatic
acute abdominal pain. Ann Emerg Med 2000;36:406 –15.
[42] Clinical policy for the initial approach to patients presenting with a
chief complaint of nontraumatic acute abdominal pain. American
College of Emergency Physicians. Ann Emerg Med 1994;23:906 –22.
[43] Venables WN, Ripley BD. Modern Applied Statistics With S. 4th ed.
New York, NY: Springer; 2002.
[44] Paulson EK, Kalady MF, Pappas TN. Clinical practice. Suspected
appendicitis. N Engl J Med 2003;348:236 – 42.
[45] Rodriguez-Sanjuan JC, Martin-Parra JI, Seco I, et al. C-reactive
protein and leukocyte count in the diagnosis of acute appendicitis in
children. Dis Colon Rectum 1999;42:1325–9.
[46] Paajanen H, Mansikka A, Laato M, et al. Novel serum inflammatory
markers in acute appendicitis. Scand J Clin Lab Invest 2002;62:579 –
84.
[47] Korner H, Soreide JA, Sondenaa K. Diagnostic accuracy of inflam-
matory markers in patients operated on for suspected acute appendi-
citis: a receiver operating characteristic curve analysis. Eur J Surg
1999;165:679 – 85.
[48] Cardall T, Glasser J, Guss DA. Clinical value of the total white blood
cell count and temperature in the evaluation of patients with sus-
pected appendicitis. Acad Emerg Med 2004;11:1021–1027.
[49] Snyder B, Hayden SR. Accuracy of leukocyte count in the diagnosis
of acute appendicitis. Ann Emerg Med 1999;33:565–574.
[50] Andersson RE, Hugander AP, Ghazi SH, et al. Diagnostic value of
disease history, clinical presentation, and inflammatory parameters of
appendicitis. World J Surg 1999;23:1– 40.