What are the True Advantages of a Pediatric

Appendicitis Clinical Pathway?

SHERIF EMIL, M.D., C.M., MICHAEL TAYLOR, B.S., FOMBE NDIFORCHU, M.D., NAM NGUYEN, M.D.

From the Division of Pediatric Surgery, Department of Surgery, University of California, Irvine Children’s
Hospital, Orange, California and Miller Children’s Hospital, Long Beach, California

Multiple protocols have been described for pediatric appendicitis, but few have been compared
with off-protocol treatment. We performed such a comparison. Children treated for appendicitis
by three pediatric surgeons over a 28-month period were studied. A protocol of primary wound
closure without drains, standardized use of antibiotics, and patient discharge according to pre-
determined clinical criteria was compared with individualized drain use, antibiotic selection, and
discharge timing. Three hundred ninety-seven children were treated, 43 per cent on pathway
(Group I) and 57 per cent off pathway (Group II). The two groups showed similar incidence of
acute (45% vs 46%), complicated (50% vs 49%), and normal (5%) appendix. Among patients with
simple appendicitis, Group I had less postoperative antibiotic use (16% vs 80% P < 0.001), shorter
hospital stays (1.44 vs 1.89 days, P = 0.001), and decreased hospital charges ($9,289 vs $10,751, P =
0.001). Among patients with complicated appendicitis, Group I had less drain placement (4% vs
27%, P < 0.001), less use of discharge antibiotics (13% vs 39%, P < 0.001), and no readmission (0%
vs 5%, P = 0.05). Infectious complications were similar between the two groups. A clinical path-
way decreases the use of unnecessary antibiotics, hospital stay, and charges for simple appendi-
citis. It decreases the use of unnecessary drains, and eliminates readmissions after complicated
appendicitis.

first clinical pathways for the treatment Patients and Methods

O NE OF THE
of perforated appendicitis in children was insti-
tuted at the Children’s Hospital in Boston in 1976 and In this retrospective study, we reviewed the charts
was reported by Schwartz et al.1
in 1983. For more of all pediatric patients (<18 years old) who were
than a decade, this pathway served as the “gold stan- treated for appendicitis by pediatric surgeons from
dard” against which other treatment strategies were January 1, 2002 through April 31, 2004 at Miller Chil-
compared. In the last 15 years, multiple other proto- dren’s Hospital, Long Beach, California, and the Uni-
cols for the treatment of simple and perforated appen- versity of California, Irvine Children’s Hospital, Or-
dicitis in children have been published.2–7 These pro- ange, California. Patients treated with an interval
tocols all differ with respect to one or more of their appendectomy approach were excluded. The study
components. Although all claim excellent outcomes, was approved by the institutional review boards of
we are only aware of one that was compared with both hospitals.
historical controls at the same institution.4 None have Collected data included patient age, gender, symp-
directly compared outcomes of patients treated accord- tom duration, results of laboratory tests, results of im-
ing to a protocol with those of patients treated off aging studies, antibiotic use, surgical approach, opera-
protocol during the same time period. We performed tive findings, use of drains, wound management,
such a comparison to delineate the true advantages (if pathologic findings, gangrene and perforation rates,
any) of a pediatric appendicitis clinical pathway used duration of operation, postoperative complications (re-
by the senior author. admissions, wound infections, postoperative ab-
scesses, and other complications), length of hospital
stay, and hospital charges. An appendix was consid-
Presented at the 17th Annual Conference of the Southern Cali- ered normal if determined by pathologic examination,
fornia Chapter of the American College of Surgeons, Bacara Re- regardless of the surgeon’s impression. Otherwise, the
sort and Spa, Santa Barbara, CA, January 20–22, 2006. surgeon’s diagnosis of acute, gangrenous, or perfo-
Address correspondence and reprint requests to Sherif Emil,
M.D., C.M., F.A.C.S., F.R.C.S.C., F.A.A.P., Division of Pediatric rated appendix was used for the remainder of the
Surgery, University of California, Irvine Medical Center, 101 The analysis.
City Drive, Building 55, Route 81, Orange, CA 92868. One surgeon (S.E.) treated all patients according to

885
886 THE AMERICAN SURGEON October 2006 Vol. 72

a standard clinical pathway. These patients formed gradually during the patient’s postoperative course.
Group I. The other two surgeons (F.N. and N.N.) tai- Wounds were closed with subcuticular suture, staples,
lored one or more components of treatment to patient or tissue adhesive.
presentation, surgical findings, and patient progress.
These patients formed Group II. Statistical Analysis
SPSS 12.0 for Windows (SPSS, Chicago, IL) was
Clinical Pathway used to create a database and to perform statistical
This pathway was adapted from a previously pub- analyses. Continuous variables are reported as mean ±
lished standard.6 All patients receive preoperative gen- SEM and are compared by independent samples t
tamicin (2 mg/kg) and clindamycin (10 mg/kg) in the tests. Pearson’s Chi-Square or Fisher’s exact test,
operating room. Patients with simple appendicitis re- where appropriate, was used for categorical variables.
ceived no further antibiotics and underwent no further A P value less than 0.05 was considered statistically
laboratory tests. They were discharged as soon as they significant.
tolerated oral intake and oral analgesics. Patients with
Results
gangrenous or perforated appendicitis received post-
operative ampicillin (200 mg/kg/day in four doses), A total of 397 consecutive patients met the inclusion
gentamicin (6 mg/kg/day in three doses), and metro- criteria and formed the study population. One hundred
nidazole (30 mg/kg/day in three doses). Penicillin- ninety-seven patients (43%) were treated on pathway
allergic patients were treated with clindamycin and (Group I) and 228 patients were treated off pathway
gentamicin only. Gentamicin levels were not routinely (Group II). Noncompliance with the clinical pathway
checked. Routine laboratory tests were not requested occurred mainly in the form of inadvertent postopera-
because they did not change the management plan. tive antibiotic use for simple appendicitis (16%) and
When patients demonstrated resolution of ileus and discharge antibiotic use for complicated appendicitis
were afebrile for 24 hours, a complete blood count was (13%). Table 1 shows the demographic, clinical, and
done. If the leukocyte count was normal, the patient therapeutic characteristics of the two groups. There
was discharged without any antibiotics. If the leuko- was a statistically significant, but not clinically sig-
cyte count was high, the patient continued on intrave- nificant, difference between the two groups in age and
nous antibiotics in the hospital or was discharged on leukocyte count. Preoperative antibiotics were admin-
oral antibiotics per the surgeon’s discretion. istered to all patients on pathway, but inadvertently
During the operation, peritoneal fluid or pus was not omitted in 16 (7%) patients off pathway. There was no
cultured. When peritonitis was encountered, copious difference in the distribution of operative findings be-
saline lavage was applied until all return was clear. No tween the two groups: 45 per cent versus 46 per cent
wound or peritoneal drains were used. All wounds simple appendicitis; 13 per cent versus 12 per cent
were primarily closed. gangrenous appendicitis; 37 per cent perforated appen-
dicitis in both groups; and 5 per cent normal appendix
Off-Pathway Treatment in both groups.
Table 2 shows the intervention and outcome data for
The off-pathway treatment differed from the clinical the two groups for simple appendicitis and compli-
pathway treatment in several ways. Antibiotic treat- cated (gangrenous or perforated) appendicitis. In pa-
ment varied, and included “triple therapy,” third- tients with simple appendicitis, postoperative antibiot-
generation cephalosporin with or without metronida-
zole, and piperacillin-tazobactam with or without TABLE 1. Demographic, Clinical, and
metronidazole. Peritoneal fluid was generally cultured, Therapeutic Characteristics
and antibiotics were occasionally changed during the Group I Group II
postoperative period according to results of intraop- (n ⳱ 169) (n ⳱ 228) P
erative cultures. The duration of antibiotic therapy af-
Age (years) 9.2 ± 0.3 8.2 ± 0.2 0.007
ter simple and complicated appendicitis was at the Gender (% male) 63 56 NS
discretion of the surgeon and was decided on a case- Symptom duration
by-case basis. Termination of in-hospital treatment (days) 2.0 ± 0.1 2.1 ± 0.1 NS
and discharge, as well as the decision to treat with White blood cell
count 16.8 ± 0.5 15.5 ± 0.3 0.02
outpatient antibiotics, was also made on a case-by- Percentage
case basis. Whereas one surgeon did not use drains, Laparoscopic 21 27 NS
the other placed Penrose drains in the peritoneal cavity Preoperative
exiting through the wound in most patients with per- antibiotics (%) 100 93 <0.001
forated appendicitis. These drains were advanced NS, not significant.
No. 10 ADVANTAGES OF A PEDIATRIC APPENDICITIS CLINICAL PATHWAY ⭈ Emil et al. 887

TABLE 2. Interventions and Outcomes

Simple Appendicitis Complicated Appendicitis
Group I Group II P Group I Group II P
Drain use (%) 0 2.9 NS 3.6 27 <0.001
Primary wound closure (%) 100 100 NS 100 99 NS
Post-operative IV antibiotics (%) 16 80 <0.001 98 100 NS
Discharge antibiotics (%) 1.3 3.8 NS 13 39 <0.001
Readmission (%) 1.3 0 NS 0 4.5 0.049
Wound infection (%) 0 0 NS 4.8 1.8 NS
Post-operative abscess (%) 0 0 NS 2.4 2.7 NS
Mean hospital stay (days) 1.44 ± 0.07 1.89 ± 0.10 0.001 5.96 ± 0.36 4.89 ± 0.19 0.005
Mean hospital charges ($) 9,289 ± 294 10,751 ± 306 0.001 20,051 ± 1,145 19,221 ± 744 NS
NS, not significant.

ic use, hospital stay, and hospital charges were rithm, major complications were reduced to 7.7 per
significantly lower in Group I. In Group I, 61 per cent cent.1 Lund and Murphy8 reported further reduction in
of patients were discharged within 24 hours of appen- morbidity using this algorithm over a 10-year period.
dectomy versus 40 per cent in Group II (P ⳱ 0.005). However, the average length of stay was 11.4 days, a
One patient in Group I was readmitted for an early relatively long hospitalization. Fishman et al.,5 from
small bowel obstruction, which was treated nonopera- the same institution, were able to shorten this period to
tively. 9.3 days by discharging selected patients on parenteral
In complicated appendicitis, drains were used in antibiotics through percutaneously inserted central
only 3 patients in Group I versus 30 patients in Group catheters. The other components of the algorithm, spe-
II. These three patients all had large pelvic abscesses. cifically, peritoneal drainage and minimum antibiotic
The use of oral antibiotics after discharge was also duration, were unchanged.5
significantly less in Group I. None of the patients with In 1990, Neilson et al.2 described the “Montreal”
complicated appendicitis treated on the clinical path- protocol for the treatment of all forms of appendicitis.
way were readmitted within 60 days for postoperative While maintaining a low incidence of complications,
complications or symptoms versus five patients the Montreal protocol had two major advantages over
(4.5%) treated off-pathway. All five were readmitted the Boston algorithm. Drains were not used and the
for abdominal pain associated with fever, vomiting, or hospitalization period was shortened.2 Over the ensu-
abdominal distention. Three received computed to- ing 10 years, the Montreal protocol was further sim-
mography scans. Two were diagnosed with small ab- plified, and minimal durations for antibiotic therapy
scesses and three were diagnosed with partial small were omitted. The results of this simplified clinical
bowel obstruction. The average hospital stay during pathway, still among the best in the literature, were
the second admission for the five patients was 1.4 reported by Emil et al.6 in 2003. Hallmarks of the
days. There was a higher rate of wound infection in pathway are the omission of postoperative antibiotics
Group I, but this was not statistically significant (P ⳱ for simple appendicitis, the standardized use of time-
0.41). Two patients in Group I and three in Group II honored inexpensive antibiotics for complicated ap-
developed abscesses postoperatively. All were treated pendicitis, the avoidance of peritoneal cultures and
with antibiotics only without operative or percutane- drains, primary wound closure in all cases, and patient
ous drainage. Group I patients had a significantly discharge according to predetermined clinical criteria
longer length of stay, but hospital charges were not without reliance on a minimal period of antibiotic
significantly different between the two groups. treatment.6 This current Montreal protocol is the clini-
cal pathway used in the patient series reported here.
Discussion To our knowledge, none of the pediatric appendici-
tis algorithms have been directly compared with non-
During the last quarter of the 20th century, the per- algorithm treatment with respect to interventions or
forated appendicitis treatment algorithm instituted at outcomes. The aim of this study was to perform such
the Children’s Hospital in Boston served as the gold a comparison to delineate the true advantages, if any,
standard for many pediatric surgeons. This algorithm of the Montreal protocol. The results reported here
included immediate appendectomy, antibiotic irriga- provide valuable lessons for surgeons who treat pedi-
tion of the peritoneal cavity, transperitoneal drainage atric appendicitis.
through the wound, and 10-day treatment with ampi- The Montreal protocol continues to produce excel-
cillin, gentamicin, and clindamycin.1, 8 With this algo- lent outcomes for appendicitis at all stages. Hospital-
888 THE AMERICAN SURGEON October 2006 Vol. 72

ization periods reproduce the results of the previous were factored in, one would find even less difference
report, and postoperative intra-abdominal infections in hospital cost. Nevertheless, the results point to a
have been further reduced.6 This reduction may be due potential area of improvement in the Montreal proto-
to the switch from clindamycin, an antibiotic that is col, namely discharging selected patients earlier on
losing antianaerobic efficacy, to metronidazole. oral antibiotics. Most patients who show clinical reso-
In the current series, the pathway resulted in a 24 lution of peritonitis, fever, and ileus will have a normal
per cent shorter hospital stay and 14 per cent lower leukocyte count. A small minority will continue to
hospital charges for simple appendicitis. This advan- have leukocytosis. Noting the results of this study, the
tage may be largely from the omission of unnecessary senior author now discharges this group on oral anti-
postoperative antibiotics. Administration of these an- biotics, instead of continuing parenteral antibiotics in
tibiotics may prolong the hospitalization of patients the hospital.
who are otherwise ready for discharge. The 16 per cent Like most other studies on this subject, this one
incidence of postoperative antibiotic therapy in Group suffers from a retrospective design. Could any of the
I patients constitutes a protocol violation and results independent variables, other than pathway assignment,
from lack of familiarity on the part of some pediatric account for the outcome differences? Although the
and surgical house staff with the protocol. Full adher- possibility cannot be dismissed, it is highly unlikely.
ence to the protocol may have further shortened the The only statistically significant differences in inde-
hospital stay and cost in these patients. pendent variables occurred in age and preoperative
In complicated appendicitis, the clinical pathway leukocyte count, and these were not clinically signifi-
does not reduce the incidence of wound infections or cant. Surgeon-dependent differences are also unlikely
postoperative intra-abdominal infections, both of because all of the appendectomies were performed by,
which are quite low in this entire series of patients. or under the supervision of, three pediatric surgeons
However, the comparison points out once again that very experienced in the treatment of pediatric appen-
peritoneal drainage is unnecessary in perforated ap- dicitis.
pendicitis. Multiple retrospective reports and a recent Although clinicians are commonly biased toward
randomized prospective trial have all discouraged the treatments they believe are ideal, studies such as this
use of peritoneal drains.2, 6, 9, 10 This approach has present the true evidence-based advantages of algo-
gained ground in the last 15 years and finally appears rithms and clinical pathways. The Montreal pediatric
to be no longer controversial. This is a fortunate oc- appendicitis protocol presented here shortens hospital
currence because the advancement and removal of stay and decreases hospital charges for simple appen-
drains in the postoperative period is often a traumatic dicitis. It eliminates readmissions for complicated ap-
event for the child, the family, and the nursing staff. In pendicitis. It also decreases the use of unnecessary
a recent survey of North American pediatric surgeons, interventions such as postoperative antibiotics for
only 13 per cent reported frequently or always using simple appendicitis and peritoneal drains for perfo-
drains in ruptured appendicitis.11 In the same survey, rated appendicitis. It does not decrease complications
only 70 per cent of surgeons reported performing rou- for any stage of appendicitis. Finally, the data pre-
tine primary wound closure.11 This is puzzling, given sented here for the entire patient population serves as
the low wound infection rate with primary closure an outcome standard for the treatment of pediatric ap-
shown here and in multiple other studies.2, 5, 6, 12, 13 In pendicitis in the 21st century.
a recent large series, Meier et al.7 showed no differ-
ence in the rate of wound infections after complicated
appendicitis between wounds that were primarily ACKNOWLEDGMENTS
closed, partially closed, or left open.
These studies were carried out in part in the General
In this study, none of the patients treated on path- Clinical Research Center, School of Medicine, University of
way required readmission for short-term, postopera- California, Irvine, with funds provided by the National Cen-
tive complications versus 4.5 per cent of patients ter for Research Resources (grant no. 5M01RR 00827-29),
treated off pathway, a statistically significant differ- U.S. Public Health Service. The study was also made pos-
ence. A similar result was observed prospectively by sible through a grant from the Memorial Medical Center
Hoelzer et al.14 This major advantage of the pathway Foundation, Long Beach, California.
is realized at the cost of increased hospital stay. De-
spite the longer hospitalization in these patients, hos- REFERENCES
pital costs were not significantly higher, perhaps be-
cause of the limits placed on unnecessary laboratory 1. Schwartz MZ, Tapper D, Solenberger R. Management of
testing and the use of inexpensive antibiotics. Further- perforated appendicitis in children: the controversy continues. Ann
more, if the hospital charges of the readmitted patients Surg 1983;197:407–11.
No. 10 ADVANTAGES OF A PEDIATRIC APPENDICITIS CLINICAL PATHWAY ⭈ Emil et al. 889

2. Neilson IR, Laberge JM, Nguyen LT, et al. Appendicitis in citis in children: a decade of aggressive treatment. J Pediatr Surg
children: current therapeutic recommendations. J Pediatr Surg 1994;29:1130–4.
1990;25:1113–6. 9. Johnson DA, Kosloske AM, Macarthur C. Perforated appen-
3. Pearl RH, Hale DA, Molloy M, et al. Pediatric appendec- dicitis in children: to drain or not to drain? Pediatr Surg Int 1993;
tomy. J Pediatr Surg 1995;30:173–81. 8:402–5.
4. Warner BW, Kulick RM, Stoops MM, et al. An evidence- 10. Tander B, Pektas O, Bulut M. The utility of peritoneal
based clinical pathway for acute appendicitis decreases hospital drains in children with uncomplicated perforated appendicitis. Pe-
duration and cost. J Pediatr Surg 1998;33:1371–5. diatr Surg Int 2003;19:548–50.
11. Muehlstedt SG, Pham TQ, Schmeling DJ. The management
5. Fishman SJ, Pelosi L, Klavon SL, O’Rourke EJ. Perforated
of pediatric appendicitis: a survey of North American pediatric
appendicitis: prospective outcome analysis for 150 children. J Pe-
surgeons. J Pediatr Surg 2004;39:875–9.
diatr Surg 2000;35:923–6.
12. Surana R, Puri P. Primary wound closure after perforated
6. Emil S, Laberge JM, Mikhail P, et al. Appendicitis in chil- appendicitis in children. Br J Surg 1994;81:440.
dren: a ten-year update of therapeutic recommendations. J Pediatr 13. Serour F, Efrati Y, Klin B, et al. Subcuticular skin closure
Surg 2003;38:236–42. as a standard approach to emergency appendectomy in children:
7. Meier DE, Guzzetta PC, Barber RG, et al. Perforated appen- prospective clinical trial. World J Surg 1996;20:38–42.
dicitis in children: is there a best treatment? J Pediatr Surg 2003; 14. Hoelzer DJ, Zabel DA, Zern JT. Determining duration of
38:1520–4. antibiotic use in children with complicated appendicitis. Pediatr
8. Lund DP, Murphy EU. Management of perforated appendi- Infect Dis J 1999;18:979–82.