Surg Endosc (2008) 22:1200–1205
DOI 10.1007/s00464-007-9583-3
Treatment of the intraabdominal abscesses through percutaneous
ultrasound-guided drainage in oncological patients: Clinical and
microbiological data
Guido Cerullo Æ Daniele Marrelli Æ Franco Roviello Æ Bernardino Rampone Æ
Francesco Saverio Ferrari Æ Francesco Vigni Æ Marianna Di Martino Æ
Enrico Pinto
Received: 2 December 2006 / Accepted: 13 June 2007 / Published online: 18 October 2007
Ó Springer Science+Business Media, LLC 2007
Abstract
Aim of the study Oncological patients are particularly
prone to the onset of septic complications such as abdominal
abscesses. The aim of our study was to analyze clinical and
microbiological data in a population of oncological patients,
submitted to percutaneous ultrasound-guided drainage
(PUD) for postoperative abdominal abscesses.
Patients and methods Data from 24 patients operated on
for neoplastic pathologies and treated with PUD for
abdominal abscesses during the postoperative period were
reviewed. In all cases cultural examination with antibio-
gram was performed.
Results In 5 out of 24 patients (20.8%), the abdominal
abscesses appeared after the discharge, with a mean hos-
pital stay of 34.2 ± 24.9 days. In six out of 24 patients
(25%) there were multiple abscesses localizations. The
cultural examination was positive in 23 patients and neg-
ative only in one patient. Abscesses localized only in the
upper abdominal regions had a significant prevalence of
monomicrobial cultural examinations (57.1%) with respect
to the results for abscesses placed in the lower abdominal
G. Cerullo
D. Marrelli
F. Roviello
B. Rampone

M. Di Martino
E. Pinto
Department of General Surgery and Surgical Oncology,
University of Siena, Policlinico S. Maria alle Scotte, Siena, Italy
F. S. Ferrari
F. Vigni
Department of Radiology, University of Siena,
Policlinico S. Maria alle Scotte, Siena, Italy
F. Roviello (&)
Dipartimento di Chirurgia Generale ed Oncologica,
Università degli studi di Siena,
Policlinico S. Maria alle Scotte,
Viale Bracci 53100, Siena, Italy
e-mail: Roviello@unisi.it
123
regions, that were polymicrobial in 88.8% of cases
(p = 0.027). An antibiogram demonstrated a stronger
activity of beta-lactamines, chinolones, and glycopeptides
with respect to aminogycosides, cephalosporins, and
metronidazole.
Conclusions In oncological patients, the planning of the
empiric antibiotic therapy should be based on the ana-
tomotopographic localization of the abdominal abscess and
on the typology of the operation performed giving prefer-
ence to beta-lactamines, chinolones and glycopeptides.
Keywords Surgical oncology
abdominal abscesses

Percutaneous drainage
Antibiotic therapy
The introduction of percutaneous ultrasound-guided
drainage (PUD) for the treatment of abdominal abscesses
has completely modified the approach to this kind of
complication that frequently occurs during the postopera-
tive period following digestive surgery [1–16]. PUD is an
interventistic and mini-invasive procedure [2] that, in
selected patients, is able to solve the infection, purifying
the abscess. Moreover, it allows the planning of correct
antimicrobial therapy through the evaluation of cultural
examination and antibiogram [12]. Surgical and, in par-
ticular, oncological patients, during the postoperative
period, are particularly prone to septic complications such
as abdominal abscesses. A correct evaluation of clinical
data and abdominal location of the abscesses can guide the
physician in choosing the empirical antibiotic therapy,
which is often fundamental for the complete resolution of
the pathology while awaiting antibiogram results [17].
The aim of our study was to analyze clinical and
microbiological data of an oncological patient population,
Surg Endosc (2008) 22:1200–1205
that underwent PUD for postoperative abdominal abscess,
in order to establish the best empirical therapy.
Patients and Methods
We have retrospectively analyzed the clinical and instru-
mental data of patients hospitalized in our department and
treated by PUD for abdominal abscesses between January
1996 and March 2005. In all patients who underwent the
surgical procedure, a prophylactic antibiotic therapy was
performed with a third-generation cephalosporin, generally
ceftazidime (1 g), three times per day. In our study we
included patients operated on for malignant tumors who
developed one or more abscesses in the abdomen as a
complication during the postoperative period. We excluded
patients with abscesses caused by pancreatitis, colecystitis,
appendicitis, gastroduodenal perforation or other benign
diseases. We also excluded patients surgically drained by
means of a new laparotomy in order to better characterize
the data from cultural examinations and antibiograms of the
drained fluid. As such, from 1167 patients operated on for
malignant abdominal and pelvic tumors, we selected 24
cases suffering from abdominal abscess (2.05%) and treated
by PUD. In every patient, the abscess was previously
diagnosed by means of a computed tomography (CT) or
ultrasound scan, and a surgeon and a radiologist evaluated
the possibility to perform the PUD, considering the clinical
and instrumental data of selected patients. Usually the
abscesses drained by PUD were larger than 3 cm [1, 12] and
the procedure were performed by specialized interventistic
radiologists. The use of ultrasound scans was preferred
because it is easy to perform, repeatable, and, furthermore,
it provides live images of the abdominal anatomic structures
during the procedure [18]. The fluid obtained by PUD of the
abdominal abscess was submitted to cultural examination
and antibiogram. The antibiogram was used to evaluate
microbial sensitivity to antibiotics and to other major anti-
microbial components used in clinical practice. We
analyzed the demographic, clinical and laboratory data
from each patient along with the mean length of hospital
stay, the mean postoperative day on which the abscess
developed, and the possible presence of comorbidity. To
better characterize the data obtained, patients were divided
into two categories based on the abdominal localization of
the abscess: upper abdominal regions (UAR), hepatic, and
splenic compartment, epigastrium and renal compartment;
or lower abdominal regions (LAR), pelvis, right and left
iliac fossa, sulci paracolici. The localization of the absces-
ses was correlated with the clinical data and laboratory
results, and with the duration of the hospitalization. Simi-
larly, we also explored potential correlations between
abscess localization and antibiogram characteristics.
1201
Statistical analysis
For the comparison between groups, we used the v2 test;
differences between continuous variables were analyzed by
means of t test or analysis of variance (ANOVA) test.
Results
There were 15 males and 9 females (ratio male/female,
1.67), with a mean age of 60.25 ± 13.2 years. Each patient
underwent surgery for a malignant tumour then, within 5
months, the abdominal abscess appeared. Twelve patients
were treated for gastric adenocarcinoma, seven for colon-
rectum adenocarcinoma, two for pancreatic adenocarci-
noma, two for cancer in the pelvis (one uterine carcinoma,
one sarcoma) and one for renal carcinoma.
In eight of 24 patients (33.3%) there were some com-
orbidities such as ictus cerebri, high blood pressure, and
diabetes mellitus. In 19 of 24 patients (79.1%) the
abdominal abscess appeared during the postoperative per-
iod, with a mean interval of 14.6 ± 10.7 days. In five of 24
patients (20.8%) the abdominal abscess appeared far from
the discharge.
After drainage of the abdominal abscess, the mean
hospital stay was 34.2 ± 24.9 days. The mortality strictly
related to the abscess was 4.16% (1/24). The incidence of
complications related to the PUD of the abdominal abscess
was 8.3% (2/24) and in both situations the complications
came from the occlusion of the drainage tube. In one
patient (4.16%) the PUD was notsuccessful so we had to
perform a new laparatomy to surgically drain the abdom-
inal abscess (Table 1).
Table 2 summarizes the localization of the abdominal
abscesses. Six patients (25%) had multiple abdominal
abscesses, mostly in the upper abdominal region. We found
that the abscesses were mostly in the splenic compartment
and right sulcus paracolici.
Table 3 shows the results of cultural examination and
antibiogram of the drained fluids. The cultural examina-
tion was positive in 23 patients and negative only in one
patient. Of these positive patients, 9 (39.1%) had a single
pathogen present. In 14 patients (60.8%), the microbio-
logic analysis highlighted more than one pathogen: 12
cases had two pathogens, two cases had three pathogens,
and one case had four pathogens present. In 23 cultural
examinations we detected 32 microorganisms: 50%
Gram-positive and 50 % Gram-negative. In 87.5% of
cases, the pathogens were aerobic/anaerobic optional
bacteria; whereas in 12.5% of cases, the pathogens were
anaerobic bacteria. Candida albicans was present in 8 of
23 cultural examinations (34.7%) and only once (4.34%)
was it the only pathogen. Overall, we observed a major
123
1202 Surg Endosc (2008) 22:1200–1205

Table 1 Clinical data from 24 patients operated on for abdominal Table 3 Microbiological data related to 24 patients who underwent
malignancy and submitted to PUD for postoperative abdominal PUD (*) and to 32 pathogens isolated in the cultural examinations (§)
abscesses
No growth 4.16% (1/24)*
Age (mean ± SD) 60.25 ± 13.2 Monomicrobial growth 39.13% (9/23)*
Sex ratio (male/female) 1.67 Polymicrobial growth 60.86% (14/23)*
Primary tumor location Candida 34.78% (8/23)*
Stomach 12 Single microbe 4.34% (1/23)*
Colon-rectum 7 Synchronous microbes 30.43% (7/23)*
Uterus 2 Gram-positive 50% (16/32)§
Pancreatic 2 Gram-negative 50% (16/32)§
Kidney 1 Anaerobic 12.5% (4/32)§
Postoperative drainage 79.16% (19/24) Facultative 87.5% (28/32)§
Day of onset (mean ± SD) 14.63 ± 10.7 Aerobic 0%§
Postdischarge drainage 20.83% (5/24)
Comorbidity 33.3% (8/24)
Morbidity 8.33% (2/24) Table 4 Microbes isolated in the 23 positive cultural examinations
Mortality 4.16% (1/24) from 24 patients submitted to PUD for postoperative abdominal
Relaparotomy 4.16% (1/24) abscesses
Length of hospital stay (mean days ± SD) 34.2 ± 24.9 Enterococcus faecalis 5
PUD, percutaneous ultrasound-guided drainage; SD, standard Citrobacter freudii 4
deviation Escherichia coli 4
Bacteroides fragilis 3
Enterococcus faecium 3
Table 2 The localization of abdominal abscesses treated by PUD (24 St. Aureus 3
patients). The results are divided into two topographic groups: upper Bacteroides ovatus 2
abdominal regions (*) and lower abdominal regions (**) Corynebacterium spp 2
Single 75% (18/24) St. epidermidis 1
Multiple 25% (6/24) Enterobacter aerogenes 1
Hepatic compartment* 13 Morganella morganii 1
Splenic compartment* 5 Bifidobacterium spp 1
Right sulcus paracolici** 5 Enterococcus casselliflavus 1
Left sulcus paracolici** 2 Enterobacter cloacae 1
Epigastrium* 2
PUD, percutaneous ultrasound-guided drainage
Pelvis** 2
Right iliac fossa** 1
Left iliac fossa** 1 difference resulted statistically significant (Pearson v2
Left renal compartment* 1 test; p = 0.027). The leukocyte counts before and after
Upper abdominal regions 21 treatment did not show significant differences, though we
Lower abdominal regions 11 noticed a trend of minor leukocytic responses in patients
PUD, percutaneous ultrasound-guided drainage with abscesses in the lower abdominal regions. The mean
hospital stay was shorter for patients with abscesses
located in the lower regions (29.4 ± 21.8) in comparison
prevalence of Enterococcus faecalis and then, in with those with abscess in the upper regions (35.5 ± 24.7)
decreasing order, Citrobacter freudii, Escherichia coli, (p = 0.360). In Table 5b we compared the different sur-
Bacteroides fragilis and Enterococcus faecium (Table 4). gical procedures with topographical localization of the
In Table 5a we compared the group of patients with abscesses (UAR vs. LAR), and microbiological and
abscesses exclusively in the upper abdominal regions clinical data. After gastric resection, abscesses occurred
(14 patients) with the group of those with abscess local- more frequently in the upper abdominal regions, while
ization in the lower abdominal regions (nine patients). patients who underwent a colectomy primarily developed
Abscesses localized in the lower abdominal regions were infections in the lower abdominal regions; however, this
polymicrobial in 88.8% of cases with respect to 42.8% of difference was not statistically significant (p = 0.233). No
abscess located only in the upper abdominal region. The significant difference was found for microbiological status

123
Surg Endosc (2008) 22:1200–1205 1203

Table 5(a) Abdominal abscess localization in 23 patients with positive cultural examinations: upper abdominal regions (UAR) versus lower
abdominal regions (LAR)
UAR (n = 14) LAR (n = 9) P

Monomicrobial 8 (57.1%) 1 (11.1%) 0.027

Polymicrobial 6 (42.8%) 8 (88.8%)
Leukocyte count (migl/mmc) (mean ± SD) Before PUD 11.9 ± 5.4 8.8 ± 4.9 N.S.
After PUD 9.6 ± 4.9 8.9 ± 6
Length of hospital stay (mean ± SD) 35.8 ± 24.7 29.4 ± 21.8 0.360

Table 5(b) Data related to the surgical procedures in 23 patients with positive cultural examinations as compared with abscess localization, and
clinical and microbiological data
Procedure performed
Gastric resection (n = 12) Colectomy (n = 6) Other procedures (n = 5)§ P

UAR 9 (75%) 2 (33.3%) 3 (60%) 0.233

LAR 3 (25%) 4 (66.7%) 2 (40%)
Monomicrobial 4 (33.3%) 3 (50%) 2 (40%) 0.791
Polymicrobial 8 (66.6%) 3 (50%) 3 (60%)
Leukocyte count (migl/mmc) (mean ± SD)
Before PUD 10.4 ± 3.5 13.8 ± 6.3 11.5 ± 8 N.S.*
After PUD 10.2 ± 4.9 7.7 ± 6.5 8.3 ± 3.1 N.S.*
Length of hospital stay (mean ± SD) 47.1 ± 24.5 24.1 ± 25.8 21 ± 3.4 0.014**
SD, standard deviation; UAR, upper abdominal regions; LAR, lower abdominal region
§ Two pancreatic resection, one nephrectomy, one isterectomy and sarcoma’s resection
* ANOVA Test (Scheffé)
** Two independent samples t test; Gastric resection vs. colectomy and other procedures

and leukocyte count between surgical procedures. Com- Table 6 Antibiogram results of 32 microbes isolated in 23 positive
paring hospitalization for different surgical procedures, cultural examinations: analysis according to the antibiotic groups
we noticed that patients who underwent gastric resection tested
had a significantly higher mean hospital stay than for Antibiotic Group Sensitivity
colectomy and other procedures (p = 0.014).
Beta-lactamines 75% (24/32)
Table 6 reports data from antibiograms performed on
Chinolones 53.1% (17/32)
the 32 microorganisms isolated in the 23 cultural examin-
ations with positive growth. The analysis of the specific Glycopeptides 46.8% (15/32)
response of every pathogen for each antibiotic class Aminoglycosides 43.7% (14/32)
showed a significant sensitivity to beta-lactamines in 75% Cephalosporins 40.6% (13/32)
of the cultures tested. We also noted that chinolones and Tetracyclines 28.1% (9/32)
glycopeptides were effective, as their global sensitivity was Lincosamides 15.6% (5/32)
53.1% and 46.8%, respectively. The bacteria were less Sulfa antibiotics 15.6% (5/32)
sensitive to aminoglycosides (43.7%) and cephalosporins Metronidazole 12.5% (4/32)
(40.6%) and even less sensitive to metronidazole (12.5%). Macrolide antibiotics 12.5% (4/32)
Further analyzing each antibiotic (Table 7), we observed Rifamycins 9.3% (3/32)
that, of the beta-lactamines, imipenem was most effective
(19 of 32 sensitive pathogens; 59.3%); and ciprofloxacin
(50%) was the most effective chinolone. Of the glyco- Discussion
peptides, teicoplanin (46.8%) was most effective; the 10
vancomycin-sensitive pathogens were cosensitive to The presence of an abdominal abscess in an oncological
teicoplanin. patient during the postoperative period should be evaluated

123
1204 Surg Endosc (2008) 22:1200–1205

Table 7 Antibiogram results of 32 microbes isolated in the 23 coagent in one third of cases. This situation could be of
positive cultural examinations: analysis according to the sensitivity of interest when deciding about the empirical antibiotic
the 32 microbes towards the antibiotics tested
therapy, and it may suggest the use of an antimycotic agent.
Antibiotic group Active antibacterial principle Sensitivity The abscess localization should also guide us in
choosing the best empirical antibiotic therapy, while
Beta-lactamines Imipenem 59.3% (19/32)
waiting for the definitive antibiogram. In our experience
Piperacillin/tazobactam 25% (8/32)
the abscesses in the lower abdominal regions more often
Amoxicillin/clavulanic acid 25% (8/32)
present a polymicrobial growth (88.8%). We also noticed
Ampicillin 18.7% (6/32)
that after gastric resection patients frequently developed
Chinolones Ciprofloxacin 50% (16/32)
abscesses in the upper abdominal regions, as well as after
Levofloxacin 25% (8/32)
colectomy more often in the lower regions. Even if the
Glycopeptides Vancomycin 31.2% (10/32)
surgical procedure seems to influence the localization of
Teicoplanin 46.8% (15/32)
the abscess, its relationship with microbiological growth is
Cephalosporins Cefepime 18.7% (6/32) unclear and data are not significant. The low number of
Cefotaxime 12.5% (4/32) patients did not allow a multivariate analysis of factors
Ceftazidime 12.5% (4/32) associated with microbiological growth. We suppose that
Tetracyclines Tetracycline 21.8% (7/32) surgical procedure influences firstly the localization of the
Aminoglycosides Gentamicin 37.5% (12/32) abscess and then its polymicrobial or monomicrobial
Amikacin 34.3% (11/32) growth. Further larger studies are necessary in order to
Tobramycin 18.7% (6/32) demonstrate a correlation between the operation performed
Oxazolindinone Linezolid 18.7% (6/32) and the microbial growth. Regarding clinical data, gastric
Metronidazole Metronidazole 12.5% (4/32) resection is also significantly correlated with a longer
Sulfa antibiotics Sulfamethoxazole/trimethoprim 15.6% (5/32) hospital stay, maybe because this kind of surgical proce-
dure is associated with more severe complications. The
weak leukocyte response of these patients could be the
very carefully in order to establish the best treatment and effect of a suppressed immune system, which could also
minimize the risk of further complications. The conserva- cause the higher susceptibility to infection by multiple
tive management through an empirical antibiotic therapy or pathogenic agents. Moreover, the rich bacterial flora of the
the PUD practice very often is able to solve the abdominal colon could play an important role in determining
abscess thus avoiding a new laparotomy. In the clinical polymicrobic infections. In such situations, the choice of
practice, the onset of abdominal abscesses during the the empirical antibiotic therapy should include wide-
postoperative period is not an unusual event, and it is often spectrum antibiotics and associated compounds. In light of
connected to phlogistic pathologies such as diverticulitis our data, we suggest the use of beta-lactamines and chi-
and appendicitis. In the last years, with the rise of opera- nolones or, as an alternative, glycopeptides. In particular,
bility and more aggressive procedures, we have observed a imipenem, ciprofloxacin, and teicoplanin should be used,
higher number of postoperative abscesses, mostly follow- as they showed the highest efficacy in this study and also in
ing oncological operations. Advanced age, surgical stress, other reports [19, 20]. On the contrary, cephalosporins
and metabolic and immune factors linked to the neoplastic showed a lower antimicrobial efficacy in comparison with
pathology play an important role in determining the onset imipenem and ciprofloxacin. The low sensitivity found in
of postoperative abdominal abscesses. These aspects ceftazidime may be explained by the fact that in most
motivated us to evaluate the problem of abdominal cases, it was used postoperatively as prophylaxis.
abscesses complicating oncological surgery, and to con- Regarding pharmacological associations, Solomkin et al.
sider favorably the use of PUD and antimicrobial therapy [21] have recently compared the efficacy of ciprofloxacin/
based on clinical, pathological and microbiological data. A metronidazole and imipenem/cilastatin combinations, both
number of studies on postoperative abscess have been with good results and overlapping efficacy. In our experi-
published, but very few studies concerned oncological ence, on the basis of the antimicrobial sensitivity, the
patients. Furthermore, analyzing the data related to our metronidazole alone does not show a good therapeutic
hospital setting was interesting. In nearly every patient in efficacy. The positive results obtained by other authors
our study the abdominal abscess occurred during the could be explained by the use of metronidazole with other
postoperative period and rarely after the discharge. The antibiotics such as chinolones.
positive cultural examination, present in the majority of In conclusion the onset of abdominal abscesses during
cases, very often showed a monomicrobial growth with the the postoperative period is a real problem, particularly in
concomitant presence of Candida albicans as a pathogen patients treated for oncological diseases. The planning of

123
Surg Endosc (2008) 22:1200–1205
the empirical antibiotic therapy should be mostly based on
the anatomotopographic localization of the abdominal
abscess, and also on the type of operation performed.
Patients with abscesses in the lower abdominal regions
show more frequently a polymicrobial etiology and a weak
immune response. In our hospital setting, this kind of
patients should be treated with a wide-spectrum antibiotic
therapy which should include at least two antibiotics, with
preference to beta-lactamines and chinolones or glyco-
peptides, in particular: imipenem, ciprofloxacin and
teicoplanin.
Acknowledgements This work was financially supported by the
University of Siena.
References
1. Jaffe T, Nelson R, DeLong D, Paulson E (2004) Practice patterns
in percutaneous image-guided intraabdominal abscess drainage:
survey of academic and private practice centers. Radiology
233(3):750–756
2. American College of Radiology (1999) Percutaneous catheter
drainage of infected intraabdominal fluid collections. In: ACR
Appropriateness Criteria. Reston, VA: American College of
Radiology
3. Gazelle GS, Mueller PR (1994) Abdominal abscess. Imaging and
intervention. Radiol Clin North Am 32:913–932
4. vanSonnenberg E, Ferrucci JT, Mueller PR, Wittenberg J, Sim-
eone JF (1982) Percutaneous drainage of abscesses and fluid
collections: technique, results, and applications. Radiology
142:1–10
5. Bouali K, Magotteaux P, Jadot A, Saive C, Lombard R, Weerts J,
Dallemagne B, Jehaes C, Delforge M, Fontaine F (1993) Percu-
taneous catheter drainage of abdominal abscess after abdominal
surgery: results in 121 cases. J Belg Radiol 76:11–14
6. vanSonnenberg E, Wing VW, Casola G, Coons HG, Nakamoto
SK, Mueller PR, Ferrucci JT Jr, Halasz NA, Simeone JF (1984)
Temporizing effect of percutaneous drainage of complicated
abscesses in critically ill patients. AJR Am J Roentgenol
142:821–826
7. Bufalari A, Giustozzi G, Moggi L (1996) Postoperative intraab-
dominal abscesses: percutaneous versus surgical treatment. Acta
Chir Belg 96:197–200
1205
8. Lang EK, Springer RM, Glorioso LW 3rd, Cammarata CA (1986)
Abdominal abscess drainage under radiologic guidance: causes of
failure. Radiology 159:329–336
9. vanSonnenberg E, Mueller PR, Ferrucci JT Jr (1984) Percuta-
neous drainage of 250 abdominal abscesses and fluid collection. I.
Results, failures, and complications. Radiology 151:337–341
10. Rothlin MA, Schob O, Klotz H, Candinas D, Largiader F (1998)
Percutaneous drainage of abdominal abscesses: are large-bore
catheters necessary? Eur J Surg 164:419–424
11. Lambiase RE, Deyoe L, Cronan JJ, Dorfman GS (1992) Percu-
taneous drainage of 335 consecutive abscesses: results of primary
drainage with 1-year follow-up. Radiology 184:167–179
12. Cinat ME, Wilson SE, Din AM (2002) Determinants for suc-
cessful percutaneous image-guided drainage of intra-abdominal
abscess. Arch Surg 137:845–849
13. Shuler FW, Newman CN, Angood PB, Tucker JG, Lucas GW
(1996) Nonoperative Management for intra-Abdominal Absces-
ses. Am Surg March 62:218–222
14. Hemming A, Davis NL, Robins RE (1991) Surgical versus per-
cutaneous drainage of intra-abdominal abscess. Am J Surg
161:593–595
15. Malangoni MA, Shumate CR, Thomas HA, Richardson JD.
(1990) Factors influencing the treatment of intra-abdominal
abscesses. Am J Surg 159:167–171
16. Gerzof SG, Robbins AH, Johnson WC, Birkett DH, Nabseth DC
(1981) Percutaneous catheter drainage of abdominal abscesses: A
five-year experience. N Engl J Med 305:653–657
17. Christou NV, Turgeon P, Wassef R, Rotstein O, Bohnen J, Potvin
M, the Canadian Intra-abdominal Infection Study Group (1996)
Management of intra-abdominal infections. The case for intra-
operative cultures and comprehensive broad-spectrum antibiotic
coverage. Arch Surg 131:1193–1201
18. Men S, Akhan O, Koroglu M (2002) Percutaneous drainage of
abdominal abscess. Eur J Radiol 43:204–218
19. Solomkin JS, Dellinger EP, Christou NV, Busuttil RW (1990)
Results of a multicenter trial comparing imipenem/cilastatin to
tobramycin/clindamycin for intra-abdominal infections. Ann Surg
212:581–591
20. Poenaru D, De Santis M, Christou NV (1990) Imipenem versus
tobramycin-antianaerobe antibiotic therapy in intra-abdominal
infectons. Can J Surg 33:415–422
21. Solomkin JS, Reinhart HH, Dellinger EP, Bohnen JM, Rotstein
OD, Vogel SB, Simms HH, Hill CS, Bjornson HS, Haverstock
DC, Coulter HO, Echols RM (1996) Results of a randomized trial
comparing sequential intravenous/oral treatment with ciproflox-
acin plus metronidazole to imipenem/cilastatin for intra-
abdominal infections. Ann Surg 223:303–315
123