Mortality Rate and Antibiotic Resistance in Complicated Diverticulitis: Report of 272 Consecutive Patients Worldwide: A Prospective Cohort Study

Abstract

Background:

Complicated diverticulitis (CD) is a common problem for surgeons. In treating it, as a general principle, every verified source of infection should be controlled. Supplementary antimicrobial management involves a delicate balance of optimizing empirical therapy while reducing unnecessary antibiotic use. The necessity to know the most frequent infecting pathogens and their spectra of resistance is becoming pivotal. The aim of this study was to determine the microbiologic profile of complicated intra-abdominal infections (IAIs) secondary to CD, to analyze the role of empirical antimicrobial therapy, and to describe the clinical aspects of CD worldwide.

Material and Methods:

The study derives from two multicenter prospective observational studies: CIAO (Complicated Intra-Abdominal infection Observational study) and CIAOW (Complicated Intra-Abdominal infection Observational World” study). The aim of the study was to analyze the intra-abdominal bacteriology in complicated diverticulitis and its relation to the clinical outcome.

Results:

The 272 patients had a mean/median age of 66.3 ± 14.9 (standard deviation; SD) and 69 (range 18–99). Patients >70 years old totaled 122 (44.9%). Conditions at admission were sepsis in 113 patients (41.5%) and severe sepsis and septic shock in 37 (13.6%) and 21 (7.7%), respectively; and localized peritonitis in 148 patients (54.4%), whereas in 124 (45.6%), the condition was generalized. Some 94 patients (34.6%) experienced a delay in initial intervention (>24 h). The mean and median duration of antimicrobial therapy were 12.3 ± 8.7 (standard deviation) and 10 (range 1–59) days. A total of 162 patients (59.6%) obtained adequate empirical antibiotic therapy, and 96 patients (35.3%) were admitted to the intensive care unit (ICU). The 30-day mortality rate was 12.1%. A total of 311 bacteria were isolated: 136 (43.7%) gram-negative, 76 (24.4%) gram-positive, 22 (7%) fungi, and 77 (24.7%) anaerobes. Of the 363 bacteria isolated, 22 (7%) were drug resistant. Four of these infections (22.2%) were health-care-associated and 18 (5.7%) community-acquired. By univariable analysis, the only statistically significant factor associated with resistant bacteria was inadequacy of the empirical antimicrobial therapy (p = 0.004). The factors associated with death were delay in initial intervention (p = 0.006) and ICU admission because of severe sepsis on admission (p = 0.004).

Conclusion:

Early source control is mandatory to reduce the mortality rate in complicated diverticulitis. Effective empirical antimicrobial agent therapy is necessary to reduce resistance and improve the clinical outcome.

In Western countries, acute left colonic diverticulosis (ALCD) is a common disease, and its prevalence is increasing, likely because of changes in lifestyle [1]. Although ALCD is more common among elderly patients, an important rise of its incidence has been seen in the younger age groups in recent years [2]. As many as one fifth of patients with acute diverticulitis under the age of 50 years are affected by ALCD according to data from Western populations [3 –5]. Recent evidence suggests that the lifetime risk of ALCD is about 4% among patients with diverticulitis [6].

Complicated diverticulitis (CD) is a common problem encountered by surgeons in the acute setting. It encompasses a variety of conditions, ranging from localized diverticular inflammation to perforation and fecal peritonitis. In uncomplicated cases of intra-abdominal infection (IAI), the infection involves a single organ and does not spread to the peritoneum. In complicated IAI, the infectious process proceeds beyond a single organ, causing either localized or diffuse peritonitis [7 –9].

As a general principle, every verified source of infection should be controlled as soon as possible, and adequate antimicrobial agent therapy should be given. The degree of urgency of treatment is determined by the affected organ(s), the relative speed at which symptoms worsen, and the underlying physiological stability of the patient [7 –9]. Antimicrobial management of severe IAIs involves a delicate balance of optimizing empirical therapy, which improves outcomes, while simultaneously reducing unnecessary antibiotic use [10]. Intra-operative intra-peritoneal cultures are common whenever a fluid collection or pus is present. The spectrum of infecting organisms and their antibiotic sensitivity may help in choosing the best antibiotic therapy.

Two sets of guidelines for IAI management with updates have been published. In 2010 and 2017, the Surgical Infection Society and the Infectious Diseases Society of America (SIS-IDSA) created guidelines for the diagnosis and management of complicated IAIs, replacing those published in 2002–2003. The second set of guidelines was published by the World Society of Emergency Surgery (WSES) in 2013 and 2017. These guidelines represent additional contributions, made by specialists worldwide, to the debate regarding proper antimicrobial drug choice [7 –11].

Systematic evaluation of infecting pathogens in CD is frequently conducted. The necessity to know the more frequent infecting pathogens and their spectrum of resistance is becoming of pivotal, especially in the presence of the increasing trend in treating CD conservatively [12]. Even more important, however, is the knowledge of the pathogens and their resistance spectra in choosing empirical antimicrobial therapy for a complicated IAI caused by CD in critically ill patients. The aim of this study was to describe the microbiologic profile of complicated IAIs secondary to CD, to analyze the role of empirical antimicrobial therapy, and to describe the clinical aspects of CD worldwide.

Materials and Methods

The study derived from two prospective multi-center observational cohort studies involving 90 medical institutions worldwide (CIAO: “Complicated Intra-Abdominal infection Observational” study and CIAOW: “Complicated Intra-Abdominal infection Observational World” study) encompassing two six-month periods (January–June 2012 and October 2012–March 2013) [13,14]. They prospectively enrolled consecutive patients older than 17 years with IAIs who were undergoing surgery, interventional drainage, or conservative treatment. From these studies, patients affected by IAIs attributable to CD were extracted. Medical institutions from each continent participated in the study. As the present study aimed principally to evaluate the microbiologic profile of IAIs attributable to CD, neither informed consent nor formal approval by an ethics committee was required.

The study met the standards outlined in the Declaration of Helsinki and Good Epidemiological Practices. It was monitored by the coordination center, which investigates and verifies missing or unclear data submitted to the central database. In each center, the coordinator collected and compiled data in an online case report system. These data included: (1) patient and disease characteristics; i.e., demographic data, type of infection (community- or healthcare-acquired), severity (defined according to the current Surviving Sepsis campaign definition during the study period as systemic inflammatory response syndrome (SIRS0, sepsis, severe sepsis, and septic shock) with curative antibiotic therapy administered in the seven days preceding the treatment; (2) the origin of the infection and the surgical procedures performed along with the effectiveness of empirical antibiotic therapy, defined as clinical and laboratory improvement after the beginning of the therapy without introduction of other drugs; and (3) microbiologic data (i.e., identification of bacteria and microbial pathogens within the peritoneal fluid, the presence of yeasts [if applicable], and the antibiotic susceptibilities of the bacterial isolates).

Descriptive statistical analysis was conducted for categorical and continuous data. Results are expressed as mean/median ± standard deviation (SD) and range. Univariable analysis was conducted to investigate factors related to death, intensive care unit (ICU) admission, and the presence of resistant bacteria. Statistical significance was defined as p < 0.05. All analyses were performed using SPSS for Windows, version 20.0 (SPSS, Chicago, IL, USA).

Results

The patients' characteristics and clinical data are shown in Table 1. The total number of patients was 272, of whom 126 were male (46.3%). Patients older than 70 years accounted for 122 of the total (44.9%). Clinical conditions at admission were sepsis in 113 patients (41.5%) and severe sepsis and septic shock in 37 (13.6%) and 21 (7.7%), respectively. Unfortunately, in 101 patients (37.1%), the condition on admission was not reported. A series of 124 (45.6%) patients was admitted with generalized peritonitis. The source control was adequate in 247 patients (90.8%). In 180 patients (66.2%), the adequacy of empirical antimicrobial therapy was reported: in 18 (6.6%), it was not adequate, and resistant bacteria were present in five of those patients. A total of 96 patients (35.3%) were admitted to the ICU. The 30-day mortality rate was 12.1%.

Table 1.

Patient Characteristics (N = 272)

	N (%)
Male/female	126 (46.3)/146 (53.7)
Age (y)
Mean (SD)	66.3 (14.9)
Median (range)	69 (17–99)
>70	122 (44.9)
Clinical condition at admission (171 pts)
Sepsis	113 (41.5)
Severe sepsis	37 (13.6)
Septic shock	21 (7.7)
Type of infection
Community acquired	255 (93.7)
Healthcare associated	17 (6.2)
Type of peritonitis
Localized	148 (54.4)
Generalized	124 (45.6)
Technique for source control (229 pts)
Open colorectal resection ± anastomosis ± stoma creation	172 (63.2)
Laparoscopic colorectal resection ± anastomosis ± stoma creation	6 (2.2)
Laparoscopic/open/percutaneous drainage	43 (15.8)
Other procedures	8 (2.9)
Source control
Adequate	247 (90.8)
Inadequate	25 (9.2)
Delay in source control >24 h	94 (36.6)
Intra-abdominal microbiologic evaluation	180 (66.2)
Duration of antimicrobial therapy (255 pts)(d)
Mean (SD)	12.3 (8.7)
Median (range)	10 (1–59)
≤7	82 (30.1)
8–14	106 (39)
≥14	67 (24.6)
Clinical adequacy of empirical antimicrobial therapy (180 pts)
Yes	162 (59.6)
No	18 (6.6)
Antimicrobial therapy ≤7 d before admission (265 pts)
Yes	60 (22.0)
No	205 (75.3)
Admitted to ICU	96 (35.3)
Co-morbidities
Malignancy	24 (8.8)
Immunosuppressive or corticosteroid therapy	52 (19.1)
Cardiovascular disease	63 (23.1)
Hospital stay (d) (265 pts)
Mean (SD)	15.6 (14.4)
Median (range)	10 (1–103)
≤7	137 (50.4)
8–14	69 (25.4)
≥14	66 (24.3)
Death rate	12.1%

ICU = intensive care unit.

In 180 patients (66.2%), intra-abdominal microbiological evaluation was performed. The bacteria isolated are listed in Table 2. There were 311 organisms isolated from 158 patients, of which 136 (43.7%) were gram-negative bacteria, 76 (24.4%) were gram-positive bacteria, 77 (24.7%) were anaerobic bacteria, and 22 were fungi (7%). Resistant bacteria were found in 15 patients (Table 3). A total of 22 bacteria (7%) of the 311 were drug resistant. Four of these (18.1%) were health-care-associated infections, and 18 (81.8%) were community-acquired infections. Glycopeptide-resistant enterococci were found in three patients, but antimicrobial therapy was adequate for all of them. No drug-resistant fungi were isolated.

Table 2.

Bacteria Isolated (N = 362) (158 patients)

	N (%)^a	(%)
Gram negative N = 171 (47.2%)
Escherichia coli	104 (60.8)	28.7
Enterobacter	9 (5,3)	2.5
Klebsiella pneumoniae	11 (6.4)	3
Proteus	7 (4.1)	2
Pseudomonas aeruginosa	15 (8.8)	4.1
Acinobacter baumannii	2 (1.2)	0.6
Others	8 (4.7)	2,2
Resistant	15 (8.8)	4.1
Gram positive N = 80 (22.1%)
Enterococcus faecalis	24 (30)	6.6
Ent. faecium	18 (22.5)	5
Streptococcus spp.	19 (23.7)	5.2
Staphylococcus aureus	4 (5)	1
Others	12 (15)	3.3
Resistant	3 (3.7)	0.8
Fungi N = 22 (6,1%)
Candida albicans	18 (81,8)	5
Non-albicans Candida	4 (18,2)	1.1
Anaerobe N = 89 (24.6%)
Bacteroides	40 (44.9)	11
Clostridium	11 (12.3)	3
Resistant bacteria	3 (3.4)	0.8
Others	35 (39.3)	9.7

Subgroup percentage.

Table 3.

Microbiologic Characteristics of Resistant Bacteria

	N (%)^a	%
Resistant bacteria isolated in health-care-associated infections N = 4 (18%)
Gram-negative
Escherichia coli ESBL+ resistant to third-generation cephalosporins	2 (50)	9.0
KPC ESBL+	1 (25)	4.5
Enterbacter faecium VRE resistant to glycopeptides	1 (25)	4.5
Resistant bacteria isolated in community-acquired infections N = 18 (82%)
Gram-negative
Escherichia coli ESBL+ resistant to third-generation cephalosporins	7 (38.8)	32
KPC ESBL+	3 (16.6)	13.6
KPC resistant to carbapenems	1 (5.5)	4.5
Pseudomonas aeruginosa resistant to carbapenems	1 (5.5)	4.5
Gram-positive
Enterococcus faecium VRE resistant to glycopeptides	3 (16.6)	13.6
Anaerobes
Bacteriodes resistant to metronidazole	2 (11.1)	9.0
Clostridium resistant to metronidazole	1 (5.5)	4.5

Subgroup percentages.

ESBL = extended-spectrum β-lactamase-producing; KPC = Klebsiella pneumoniae resistant to carbapenems; VRE = vancomycin-resistant enterococci.

Univariable analysis findings are reported in Table 4. This analysis demonstrated that the only statistically significant factor associated with the presence of drug-resistant bacteria was the failure of empirical antimicrobial therapy (p = 0.004). The statistically significant factor associated with death was a delay in the initial intervention for source control (p = 0.006), and the statistically significant factor associated with ICU admission was severe sepsis at admission (p = 0.004).

Table 4.

Univariable Analysis for Resistant Bacteria, Death, and Intensive Care Unit Admission

Risk Factors for	P Value
Death
Generalized peritonitis	0.252
Delay in initial intervention >24 h	0.006
Recent antibiotic therapy	0.916
Age >60 y	0.572
Malignancy	0.925
Immunosuppression/corticosteroid therapy	0.174
Cardiovascular disease	0.686
Clinical condition at admission	0.068
Source control	0.518
ICU admission	0.056
Failure of empirical antibiotic therapy	0.438
Reoperation	0.080
Clinical condition at admission	0.226
Resistant bacteria
Age	0.580
Clinical condition at admission	0.146
Failure of empirical antibiotic therapy	0.004
Duration of antibiotic therapy	0.379
Hospital stay	0.524
ICU admission	0.501
Recent antimicrobial therapy	0.076
Malignancy	0.931
Immunosuppression/corticosteroid therapy	0.133
Cardiovascular disease	0.919
ICU admission
Age	0.189
Clinical condition at admission	0.004
Clinical condition during stay	0.166
Generalized peritonitis	0.574
Source control failure	0.714
Failure of antibiotic therapy	0.699
Reoperation	0.114
Malignancy	0.326
Immunosuppression/corticosteroid therapy	0.855
Cardiovascular disease	0.561
Delayed initial intervention (>24 h)	0.226

Discussion

Intra-abdominal infections occur in a variety of pathological conditions, ranging from uncomplicated diverticulitis to fecal peritonitis. The mainstay in the treatment of all these conditions encompasses both source control and antibiotic therapy in order to prevent local and hematogenous spread and to reduce late complications. Source control includes all measures undertaken to eliminate the origin of the infection. This generally involves drainage of an abscess or infected fluid collection, debridement of necrotic or infected tissues, and definitive removal of the infected organ.

Antimicrobial agent therapy also plays a pivotal role in the management of IAIs, especially in patients with severe sepsis and septic shock, who require immediate empirical antibiotic therapy. The decision algorithm depends mainly on the presumed pathogens, their risk factors for major resistance patters, the place of acquisition (community acquired or healthcare associated), and the patient's clinical condition (stable or with severe sepsis or septic shock). Initial antimicrobial therapy typically is empirical, because the patients need immediate treatment, and culture results and the antibiogram may require as long as 24–72 h to become available. An insufficient or otherwise unsuccessful antimicrobial regimen is one of the variables strongly associated with unfavorable outcomes in critically ill patients [11,15] Moreover, previous studies have demonstrated a higher risk of death as patients transition from sepsis to severe sepsis or septic shock [11].

An alarming problem that has become more and more prevalent over recent years is drug-resistant bacteria. In its recent annual report on global risk, the World Economic Forum (WEF) concluded that “arguably the greatest risk to human health comes in the form of antibiotic-resistant bacteria” [16]. Moreover, other key organizations, including the Infectious Diseases Society of America (IDSA), the U.S. Centers for Disease Control and Prevention (CDC), and the World Health Organization (WHO) have made antibiotic resistance the focus of highly visible reports, conferences, and actions. This problem has posed a serious challenge to clinicians worldwide as they attempt to administer efficacious antimicrobial therapy.

Previous data demonstrated that factors influencing the presence of resistant bacteria are recent hospitalization (health-care associated bacteria), urinary catheterization, age >70 years, antimicrobial therapy, and immunosuppression [17 –19]. In our study, the most common organisms isolated from patients with CD were Escherichia coli, Bacteriodes, Enterococcus faecalis, Streptococcus spp., Enterococcus faecium, Candida albicans, Pseudomonas aeruginosa, and Klebsiella. The 62.8% of patients CD were admitted with sepsis, severe sepsis, or septic shock, so more half of the patients were in critical condition. Moreover, as demonstrated in the univariable analysis, not surprisingly, the patient's condition at admission is an independent risk factor for ICU admission (p = 0.004). These data confirm the necessity to initiate promptly an empirical therapy that should cover the pathogens presumed to be involved, taking into consideration also the risk factors indicative of major resistance patterns locally and the clinical condition of the patient. Unsuccessful empirical therapy was seen in 6.6% of the patients in our series. The 30-day mortality rate was 12.1%.

To the best of our knowledge, no previous reports exist of microbiologic investigation in a large cohort of patients. The present paper describes the first worldwide group of patients with acute diverticulitis for which there has been detailed microbiologic evaluation. This examination showed that 7% of all bacteria isolated were drug resistant. Of these, 81.8% were community-acquired infections, and 18.1% were health-care related. Particularly problematic resistant bacteria are E. coli and Klebsiella pneumoniae expressing extended-spectrum β lactamase (ESBL), organisms resistant to third-generation cephalosporins, Pseudomonas and Klebsiella resistant to carbapenems (KPC), Ent. faecium resistant to glycopeptides, and Bacteroides and Clostridium resistant to metronidazole.

Adequate source control was performed in 247 of our patients (90.8%). It is well known that inadequate source control at the time of the initial operation is associated with a higher mortality rate in patients with severe IAIs [11].

As previously said, accurate risk stratification is of fundamental importance in critically ill patients. Despite the relatively low prevalence of resistant bacteria, high-risk patients must be started on an aggressive broad-spectrum empirical agent that should have efficacy against ESBL- and KPC-positive bacteria, taking into account geographic differences and local epidemiologic data. Also, the increasing number of patients who present with a community-acquired multi-resistant infection must be emphasized. It is important to underline, as shown in recently published studies [20, 21], that the recognized factors for E. coli ESBL infections are changing (age >70 y, recent antibiotic therapy [within the previous three mos], urinary catheterization, and recovery in healthcare centers). The rise of resistant bacteria has been greater than expected, and the majority of resistant and multi-resistant organisms is now being isolated from community-acquired infections. These unexpected findings suggest that there are causes of resistance beyond the known ones. More extensive and detailed epidemiologic studies should be performed.

As already seen in such common diseases as appendicitis and cholecystitis, the paradigm of treatment is shifting [20,21]. In fact, the necessity to consider the provenance of the patient, his/her co-morbidities, and the change in the community bacterial flora increases the difficulty of selecting effective empirical antimicrobial therapy. As a counterpart to the antimicrobial paradigm, shifting the necessity to reduce antimicrobial usage must be taken into consideration. The data also showed another important factor: The effectiveness of antimicrobial therapy is not influenced solely by the presence of resistant bacteria, as a number of patients received unsuccessful therapy even though they were infected with common pathogens. Unfortunately, the research team did not record all the drug regimens, although it is of great importance that every surgeon know and adopt the therapy guidelines [7 –9].

Present and previous data [20,21] demonstrate that correct antimicrobial management is a key component in preventing the emergence of drug resistance. Our univariable analysis showed that the only factor associated with resistant bacteria was inappropriate empirical antimicrobial therapy (p = 0.004). Of course, the presence of resistant bacteria is both a cause and a consequence of improper antibiotic use. On the one hand, the presence of multi-resistant species is attributable to irresponsible use of antibiotics; on the other hand, the presence of these bacteria, with or without known risk factors, makes the standard antibiotic therapy for intra-abdominal infections ineffective.

The balance between the adequacy of empirical therapy and the reduction of antimicrobial use with adequate and timely source control definitively improves clinical outcomes. To obtain the best results in the management of complicated IAIs, it is crucial that surgeons, besides mastering the technical aspects of source control, understand the properties of each antibiotic they are using, including not only its spectrum of activity but also the correct dose.

Conclusions

Early source control is mandatory to reduce the mortality rate in CD. Effective and adequate empirical antimicrobial therapy also is necessary to reduce resistance and improve the clinical outcomes.

Footnotes

Author Disclosure Statement

All authors declare no source of funding or conflict of interest in relation to this manuscript.

References

Weizman

, Nguyen

. Diverticular disease: Epidemiology and management. Can J Gastroenterol, 2011; 25:385–389.

Schoetz

. Diverticular disease of the colon: A century-old problem. Dis Colon Rectum, 1999; 42:703–709.

Collins

, Winter

. Modern concepts in diverticular disease. J Clin Gastroenterol, 2015; 49:358–369.

Warner

, Crighton

, Moineddin

, et al. Fourteen-year study of hospital admissions for diverticular disease in Ontario. Can J Gastroenterol, 2007; 21:97–99.

Jamal Talabani

, Lydersen

, Endreseth

, Edna

. Major increase in admission- and incidence rates of acute colonic diverticulitis. Int J Colorectal Dis, 2014; 29:937–945.

Shahedi

. Long-term risk of acute diverticulitis among patients with incidental diverticulosis found during colonoscopy. Clin Gastroenterol Hepatol, 2013; 11:1609–1613.

Sartelli

, Viale

, Catena

, et al. 2013. WSES guidelines for management of intra-abdominal infections. World J Emerg Surg, 2013; 8:3 DOI 10:1186/1749-7922-8-3

Sartelli

, Catena

, Abu-Zidan

, et al. Management of intra-abdominal infections: Recommendations by the WSES 2016 consensus conference. World J Emerg Surg, 2017; 12:22 DOI: 10.1186/s13017-017-0132-7

Mazuski

, Tessier

, May

, et al. The Surgical Infection Society revised guidelines on the management of intra-abdominal infections. Surg Infect, 2017:18:1–76.

10.

Sartelli

, Catena

, Coccolini

, Pinna

. Antimicrobial management of intra-abdominal infections: Literature's guidelines. World J Gastroenterol, 2012; 18:865–871.

11.

Sartelli

, Catena

, Di Saverio

, et al. Current concept of abdominal sepsis: WSES position paper. World J Emerg Surg, 2014; 9:22 DOI 10:1186/s13017-017-0132-7

12.

Ansaloni

, Catena

, Coccolini

, et al. Surgery versus conservative antibiotic treatment in acute appendicitis: A systematic review and meta-analysis of randomized controlled trials. Dig Surg, 2011; 28:210–221.

13.

Sartelli

, Catena

, Ansaloni

, et al. Complicated intra-abdominal infections worldwide: The definitive data of the CIAOW Study. World J Emerg Surg, 2014; 9:37 DOI 10.1186/1749-7922-9-37.

14.

Sartelli

, Catena

, Ansaloni

, et al. Complicated intra-abdominal infections in Europe: A comprehensive review of the CIAO study. World J Emerg Surg, 2012; 7:36 DOI 10.1186/1749-7922-7-36.

15.

Shani

, Muchtar

, Kariv

, et al. Systematic review and meta-analysis of the efficacy of appropriate empiric antibiotic therapy for sepsis. Antimicrob Agents Chemother, 2010, 54:4851–4863.

16.

www3.weforum.org/docs/WEF_GlobalRisks_Report_2013.pdf Accessed October 21, 2015.

17.

Tumbarello

, Trecarichi

, Bassetti

, et al. Identifying patients harboring extended-spectrum-beta-lactamase-producing Enterobacteriaceae on hospital admission: Derivation and validation of a scoring system. Antimicrob Agents Chemother, 2011:3485e3490.

18.

Johnson

, Anderson

, May

, Drew

. Utility of a clinical risk factor scoring model in predicting infection with extended-spectrum β-lactamase-producing Enterobacteriaceae on hospital admission. Infect Control Hosp Epidemiol, 2013; 34: 385e392.

19.

Shitrit

, Reisfeld

, Paitan

, et al. Extended-spectrum beta-lactamase-producing Enterobacteriaceae carriage upon hospital admission: Prevalence and risk factors. J Hosp Infect, 2013; 85:230e232.

20.

Coccolini

, D'Amico

, Sartelli

, et al. Antibiotic resistance evaluation and clinical analysis of acute appendicitis: Report of 1431 consecutive worldwide patients: A cohort study. Int J Surg, 2016; 26:6–11.

21.

Coccolini

, Sartelli

, Catena

, et al; CIAO and CIAOW Study Groups. Antibiotic resistance pattern and clinical outcomes in acute cholecystitis: 567 consecutive worldwide patients in a prospective cohort study. Int J Surg, 2015 Jul 17. pii: S1743-9191(15)00379-9.