Little data on the usefulness of microbiological samples in appendicular peritonitis (AP) are available. The objectives of the study were to document the clinical value of systematic peritoneal swabbing in complicated appendicitis, to establish whether they influence postoperative outcome, and to help to optimize empirical preoperative treatment.
Charts of all consecutive patients undergoing appendicectomy for acute perforated appendicitis were analyzed over a 7-year period from 2005 to 2011 in a tertiary hospital in New Caledonia. From prospectively recorded data, microbiological culture results, and sensitivity of isolates were obtained in relation to histopathological findings and infective morbidity.
Overall, 144 cases including 47 (33%) children and 97 (67%) adults with perforated appendicitis were included in the study. Fifty-one (35%) had generalized peritonitis treated laparoscopically in 30 (59%) cases. Peritoneal fluid samples yielded a positive culture in 104 (74%) patients. The most commonly recovered species were Escherichia coli (81%), Streptococcus milleri group (12%), and Pseudomonas aeruginosa (12%). Postoperative infectious complications occurred in 32 (22%) including intra-abdominal abscess (n = 20) and wound infection (n = 12). Probabilist antibiotic regimen was less often suitable in children than in adults (p = 0.003). Infectious complications occurred more often in cases with an unsuitable antibiotic regimen after culture results compared with a suitable antibiotic regimen before culture results (p = 0.01).
Although antibiotic use may be considered as an adjunct to surgical intervention of AP, the appropriate use of preoperative antibiotics and modifications according to culture results are essential to prevent infectious complications.
Acute appendicitis is a common visceral emergency both in children and adults. A dramatic decrease in mortality and morbidity has been established since the systematic use of empiric broad spectrum intravenous antibiotics in acute appendicitis.1 Recommendations for community-acquired peritonitis suggest the use of a large choice of broad-spectrum antibiotics, either administered alone or in combination.2 ; 3
Documenting the intraperitoneal pathogens is important in order to identify the pathogen and adjust the antibiotic regimen. Little data on microbiological samples and antibiotic treatment in appendicular peritonitis (AP) in the Pacific region are available.4 ; 5 This study was performed to assess the clinical value of systematic peritoneal swabbing in complicated appendicitis admitted in New Caledonia and to establish whether they influence postoperative outcome. In addition, we intended to characterize the current bacterial epidemiology of AP in New Caledonia in order to develop a local consensus concerning empirical antibiotic regimen in patients undergoing surgery for acute AP.
Clinical and bacteriologic data for all cases of AP admitted from January 1, 2005 to December 31, 2011 to the general surgery department of the Centre Hospitalier Territorial, Noumea, New Caledonia were retrospectively collected. Cases were identified from the electronic records of the hospital system. AP was defined by the presence of peritoneal purulent liquid noticed during surgery and the presence of perforated appendicitis on histology. Peritonitis was localized when infection was limited by adjacent organs and not spread to the entire peritoneal cavity. Clinical data collected included age at surgery, sex, antibiotic administration before diagnosis, surgical technique (laparotomy or laparoscopy), macroscopic findings at surgery, type and duration of postoperative antibiotherapy, and postoperative infectious complications. Infectious complications were defined as those occurring within 30 days of surgery, and included intra-abdominal abscess (defined by a combination of clinical and radiological results) and/or wound infection (defined by clinical findings). Children were defined by patients aged < 15 years.
In the Department of General Surgery at our hospital, two different empirical antibiotic regimens for AP are used according to surgeon or anesthetist preference. Patients received either amoxicillin-clavulanate (amoxi/clav) + gentamycin or cefotaxime/ceftriaxone + gentamycin + metronidazole. Aminoglycoside was given for 2 days and other antibiotics were administered for 5–7 days unless continued sepsis warranted further treatment.
Peritoneal fluid or pus specimens were collected from the peritoneal cavity at surgery. The liquid was aspirated into a syringe immediately after opening of the abdominal cavity. A fluid specimen was injected into an anaerobic transport vial (Portagerm, bioMérieux). In the microbiological laboratory, 5% horse blood agar, PVX agar, and Schaedler agar were inoculated and incubated during 48 hours under aerobic and anaerobic atmospheres. TCS and Schaedler medium were incubated during 4 days. All organisms were identified using conventional methods (biochemical tests). Susceptibility to commonly used antibiotics was tested by automated liquid medium culture (Vitek2C, bioMérieux) or semi-automated method for anaerobic bacteria (miniAPI, bioMérieux).
The study was approved by the Institutional Review Board of Centre Hospitalier Territorial. Informed consents were not obtained from patients as this was a retrospective study. All data were anonymized.
All analyses were performed using STATA 12 (College Station, TX, USA). Quantitative and qualitative variables were compared by using paired Student t test and Chi-square test, respectively. With a frequency of events < 5 or values not following normal distributions, Fishers Exact and Mann-Whitney tests were applied. A p value < 0.05 was defined as statistically significant.
During the study period, a total of 144 consecutive patients were identified. Peritoneal fluid samples were taken in all cases, yielding positive culture in 107 (74%), including 47 (33%) children and 97 (67%) adults. The median age of patients was 19 years (Interquartile range, 13–41). Seventy-five (52%) were female (Table 1). A total of 93 (65%) had localized peritonitis while 51 (35%) had generalized peritonitis. Among patients presenting generalized peritonitis, 30 (59%) underwent laparoscopy and 21 (41%) underwent laparotomy (Table 2). No deaths were observed.
|Patient data||Children n = 47||Adults n = 97||Total n = 144|
|Age, mean (IQR)||10 (8–12)||30 (18–46)||19 (13–41)|
|Gender n (%)|
|Male||25 (53)||44 (45)||69 (48)|
|Female||22 (47)||53 (55)||75 (52)|
IQR = interquartile range.
|Macroscopic findings at surgery by surgical technique||Children n = 47||Adults n = 97||Total n = 144||p|
|Localized peritonitis||29 (62)||64 (66)||93 (65)||0.67|
|Laparoscopy||15 (51)||45 (71)||60 (66)||0.03|
|Laparotomy||14 (49)||19 (29)||33 (34)||—|
|Generalized peritonitis||18 (38)||33 (34)||51 (35)||—|
|Laparoscopy||10 (58)||20 (60)||30 (59)||0.71|
|Laparotomy||8 (42)||13 (40)||21 (41)||—|
Data are presented as n (%).
Ten patients with negative bacteria culture had received antibiotics preoperatively. No yeast was observed on peritoneal cultures. There were one, two, and three bacterial species per specimen in 178, 31, and seven cases, respectively. Escherichia coli, Streptococcus group milleri, and Pseudomonas aeruginosa were the most commonly recovered species. Among the 87 E. coli of the study, the culture was monomicrobial in 71% of cases. Anaerobic cultures were positive for 10 cases of PA: four anaerobic mixed cultures (≥ 3 species), and six Bacteroides spp. The susceptibilities of aerobic bacteria to the empirical antibiotics mostly used are shown in Table 3. There were no differences in terms of susceptibility between children and adults. After culture results, the antibiotic regimen was altered accordingly in 10/31 (32%) patients receiving unsuitable presumptive antibiotics.
|Bacteria||n (% positive cultures)||Susceptibility to antibiotics|
|Escherichia coli||87/107 (81)||53/87 (61)||82/87 (94)||87/87 (100)||87/87 (100)|
|Streptococcus milleri group||13/107 (12)||13/13 (100)||13/13 (100)||13/13 (100)||13/13 (100)|
|Pseudomonas aeruginosa||13/107 (12)||0/13 (0)||0/13 (0)||13/13 (100)||13/13 (100)|
|Enterococcus spp.||5/107 (5)||5/5 (100)||0/5 (0)||3/5 (60)||5/5 (100)|
|Other Enterobacteriacae||7/107 (6)||3/7 (43)||7/7 (100)||7/7 (100)||7/7 (100)|
Thirty-two (22%) patients developed a postoperative infection, including intra-abdominal abscess in 20 cases and wound infection in 12 cases. Children received suitable treatment less often (p = 0.003) and developed infectious complications more often than adults (p = 0.02; Table 4). Patients with an unsuitable antibiotic regimen after culture results developed infectious complications more often than others (p = 0.001; Table 5).
|Treatment and complications||Children n = 47||Adults n = 97||Total n = 144||p|
|Positive culture||38 (81)||69 (71)||107 (74)||0.11|
|Suitable antibiotic regimen||28 (75)||62 (90)||90 (86)||0.003|
|Unsuitable antibiotic regimen||10 (25)||7 (10)||17 (14)||—|
|Infectious complications||14 (30)||18 (19)||32 (22)||0.02|
|Intra-abdominal abscess||12 (86)||8 (44)||20 (62)||—|
|Wound infection||2 (14)||10 (56)||12 (38)||—|
Data are presented as n (%).
|Complications||Antibiotic regimen||Total n = 144||p|
|Suitable before culture results n = 112||Unsuitable before culture results n = 32|
|Infectious complications||21 (19)||11 (34)||32 (22)||0.03|
|Intra-abdominal abscess||12 (57)||8 (73)||20 (62)||0.02|
|Wound infection||9 (43)||3 (27)||12 (38)||0.08|
Data are presented as n (%).
In 25 cases, the infectious complication was cultured (13 intra-abdominal abscesses and 12 wound infections). Twenty cultures were positive, with bacteria resistant to the empirical antibiotic treatment in five cases (1 E. coli producing extended-spectrum β-lactamase not detected in the previous sample, 2 P. aeruginosa, and 2 Bacteroides fragilis resistant to amox/clav).
Antibiotics used preemptively are not adapted to the sensitivity of bacteria most commonly found in the peritoneal fluid in 22% of patients treated for PA in New Caledonia. It is largely conceded that all potential pathogens should be included in the spectrum of the empiric antibiotic regimen to reduce the subsequent morbidity of PA.6 ; 7 The traditional practice of obtaining intraoperative cultures has been challenged for several reasons, including often negative results, different pathogens involved in postoperative infectious complications, or absence of antibiotic changes on the swab results.8; 9; 10 ; 11 In line with another report,12 our results suggest a significant association between resistance to empirical antimicrobials and postoperative infection. Therefore, systematic cultures are of great value in detecting epidemiologic changes in the resistance patterns of pathogens associated with peritonitis. Routine culture of fluids from PA should be performed to allow subsequent adjustment of the antibiotic regimen to the sensitivity of the species identified.
A majority of peritoneal fluids yielded positive culture, consistent with previous reports.11 ; 13 A wide variation of positive culture rate has previously been described, ranging from 24% to 93%,10; 14 ; 15 most likely explained by the various techniques employed for collection and transfer of specimens and culture.
With the exception of E. coli, the most frequently cultured organisms are not dominant in the normal intestinal microbiota, particularly in patients coming directly from the community. 16 Our findings suggest that amoxi/clav is not a satisfying treatment for perforated appendicitis, as 40% of all recovered aerobic isolates showed resistance to this association. A third-generation cephalosporin associated with metronidazole is an acceptable alternative and this combination is included in the choice of empirical therapies proposed by guidelines for management of intra-abdominal infection both in the United States3 and France.2
P. aeruginosa was the third most frequently cultured organism identified in the peritoneal fluid and the most frequently resistant to empirical therapy. According to previous reports, the incidence of P. aeruginosa in cases of perforated appendicitis is highly variable and can be as high as 35% of cases with peritoneal positive cultures. 15 The bacterial spectrum of the combination of third-generation cephalosporin associated with metronidazole does not include P. aeruginosa. If this combination is chosen as empirical therapy, adding an aminoglycoside for 2 days to cover P. aeruginosa should be considered, according to the results of culture. This combination has been described as cost-effective. 17
Another important finding of our study shows that fewer children receive suitable treatment than adults, and they consequently develop infectious complications more frequently. As previously shown, children should be treated in the same way as adults until there is no further clinical evidence of infection.18 Ideally, empirical antibiotic therapy should be administered before surgery.
Although antibiotic use can be considered as an adjunct to surgical intervention, their appropriate use is an important aspect of the care of patients with PA. Therefore, intraoperative cultures should be systematically performed to determine antibiotic susceptibility patterns. Further, the ecologic evolution should be regularly re-evaluated to optimize the empirical antibiotic regimens for such patients.
The authors thank Lea Sjögren for her critical review of the paper.