Langenbeck's Archives of Surgery (2023) 408:345 https://doi.org/10.1007/s00423-023-03063-4 RESEARCH Impact of empiric antibiotic therapy on the clinical outcome of acute calculous cholecystitis Maite de Miguel‑Palacio1,2 · Ana‑María González‑Castillo2,3 · Estela Membrilla‑Fernández2,3 · María‑José Pons‑Fragero2 · Amalia Pelegrina‑Manzano2,3 · Luis Grande‑Posa1,2 · Ricard Morera‑Casaponsa3 · Juan‑José Sancho‑Insenser1,2 Received: 17 June 2023 / Accepted: 14 August 2023 © The Author(s) 2023 Abstract Purpose Although mortality and morbidity of severe acute calculous cholecystitis (ACC) are still a matter of concern, the impact of inadequate empirical antibiotic therapy has been poorly studied as a risk factor. The objective was to assess the impact of the adequacy of empirical antibiotic therapy on complication and mortality rates in ACC. Methods This observational retrospective cohort chart-based single-center study was conducted between 2012 and 2016. A total of 963 consecutive patients were included, and pure ACC was selected. General, clinical, postoperative, and micro- biological variables were collected, and risk factors and consequences of inadequate treatment were analyzed. Results Bile, blood, and/or exudate cultures were obtained in 76.3% of patients, more often in old, male, and severely ill patients (P < 0.001). Patients who were cultured had a higher overall rate of postoperative complications (47.4% vs. 29.7%; P < 0.001), as well as of severe complications (11.6% vs. 4.7%; P = 0.008). Patients with positive cultures had more overall complications (54.8% vs. 39.6%; P = 0.001), more severe complications (16.3% vs. 6.7%; P = 0.001), and higher mortality rates (6% vs. 1.9%; P = 0.012). Patients who received inadequate empirical antibiotic therapy had a fourfold higher mortality rate than those receiving adequate therapy (n = 283; 12.8% vs. 3.4%; P = 0.003). This association was especially marked in severe ACC TG–III patients (n = 132; 18.2 vs. 5.1%; P = 0.018) and remained a predictor of mortality in a binary logistic regression (OR 4.4; 95% CI 1.3–15.3). Conclusion Patients with positive cultures developed more complications and faced higher mortality. Adequate empirical antibiotic therapy appears to be of paramount importance in ACC, particularly in severely ill patients. Keywords Cholecystitis · Acute calculous cholecystitis · High-risk patients · Empiric antibiotic treatment · Tokyo Guidelines · Antibiotic adequacy Introduction Acute calculous cholecystitis (ACC) is the second most frequent cause of surgical emergency admission in the Western world. There is a general agreement on laparoscopic cholecystectomy * Maite de Miguel‑Palacio as the preferred initial treatment of choice for patients with mild mdemiguel@psmar.cat or moderate cholecystitis, as proposed by the Tokyo Guidelines 1 2018 (TG18) [1, 2]. For severe cases, however, there is a greater Department of Surgery, Autonomous University of Barcelona, Barcelona, Spain variability of therapeutic recommendations, including antibiotic 2 therapy and/or cholecystostomy. Empiric antibiotic treatment Department of General Surgery, Hospital del Mar, Hospital del Mar Medical Research Institute (IMIM), Barcelona, (EAT) is central to every therapeutic option [3, 4]. Spain As far ago as 1991, Mosdell et al. demonstrated that the 3 Department of Medicine and Life Sciences, Pompeu Fabra adequacy of empiric antibiotic treatment is paramount to University, Barcelona, Spain obtaining good clinical results in abdominal infections [4]. 13 Vol.:(0123456789) 345 Page 2 of 9 Langenbeck's Archives of Surgery (2023) 408:345 Our group updated Mosdell’s findings and demonstrated that or gallbladder carcinoma, were excluded from the analysis the adequacy of antibiotic treatment in secondary peritoni- (Fig. 1). tis yielded a better prognosis, lowering both morbidity and mortality [5]. Variables Regrettably, there is a lack of similar hard data about the consequences of the inadequacy of EAT on ACC. In the last Data were extracted and coded from the electronic medical decade, only a few studies have analyzed the relationship records. A total of 211 primary variables were recorded, between microbiology, EAT, and clinical results in ACC. including demography, laboratory and imaging tests, treat- Coccolini et al. [6] and, more recently, Suh S. et al. [7] ments, length of stay, complications, and mortality. An addi- described the antibiotic resistance pattern in a small group tional file shows this in more detail (see Additional file 1. of patients with acute cholecystitis and signaled the impor- Variable list). Charlson’s Comorbidity Index and the ASA tance of EAT in mild or moderate acute cholecystitis. Both classification were used as preoperatory risk assessment studies addressed the inadequacy of EAT only marginally measures. In addition, the type of initially received treat- and with some limiting methodological flaws. Moreover, ment, surgical or nonsurgical, type of empiric antibiotic these studies could not draw a clear relationship between administered, and duration of the antibiotic treatment (both the inadequacy of EAT and poor clinical results; however pre-and postoperative days) were analyzed. According to logical, this causality may seem. Most likely, the clue is to the microbiological variables, genera and species of iso- stratify the analysis by the degree of severity of the ACC. lated germs were recorded, as well as subgroup classifica- What may be irrelevant in a mild case can become a critical tion according to Gram stain, oxygen requirements (aerobic/ issue in a frail patient with ACC and severe sepsis. Recently, anaerobic), and morphology (cocci, bacilli, coccobacilli). our group identified a concise set of major risk factors for The antibiotic sensitivity was determined for each isolated mortality in ACC and even proposed a new score (ACME) germ (sensitive, resistant, intermediate, or unknown), and focused on severe ACC, where the Tokyo Guidelines lack extended-spectrum beta-lactamase bacteria (ESBLB) clear indications [8]. expression was recorded. Variables were collected in File Therefore, there is a need for evidence in the ACC assess- Maker v.13 (Mountainview, CA, USA). ing the relevance of EAT on clinical results. Since every strategy to treat ACC includes EAT, this analysis is vital to Primary and secondary outcomes improving clinical results. The main objective of the study was to investigate the impact of inadequate EAT on mortality. This primary outcome was Material and methods assessed in the first 30 days after the diagnosis of ACC or occurring at any time during the same admission. Secondary Study type and setting outcomes were the impact of inadequacy on the prevalence and severity of complications during the same period as This is a retrospective single-center study performed in the mortality, as well as the description of comorbidities, use of Emergency Department of General Surgery of a University vasopressors, and Tokyo Guidelines grade analysis. Hospital in Barcelona, Spain. Adequacy estimation Patients, inclusion, and exclusion criteria Therapy was considered adequate when at least one intra- A total of 963 consecutive patients with an initial diagnosis venous antimicrobial was given within 24 h of admission. of acute cholecystitis between January 2012 and December Antimicrobial treatment was considered adequate when used 2016 were included. Data were collected by 2018, and analy- at acceptable doses and intervals, with the optimal route of sis was completed by 2021. administration and following local guidelines and recom- All patients with a clinical diagnosis of acute cholecystitis mendations. The adequacy of the antibiotic was determined according to the Tokyo Guidelines of 2018 (TG18) were by the in vitro susceptibility of the specific bacterial isolate, included, as well as those with a histopathologic diagnosis and in the case of polymicrobial infections, adequacy was of ACC [9]. The study focused on patients with “pure” acute determined for all isolates. calculous cholecystitis [8]. Therefore, patients with other A germ was considered sensitive to the antibiotic interrelated diagnoses, such as postoperative cholecystitis, when the minimum inhibitory concentration was less acute cholangitis, acute pancreatitis, acalculous cholecys- than the internationally accepted threshold [10]. When titis, persistent biliary colic, and cholecystitis following all germs isolated in a patient’s sample(s) were sensitive endoscopic retrograde cholangiopancreatography (ERCP) to the empiric antibiotics received, antibiotic therapy was 13 Langenbeck's Archives of Surgery (2023) 408:345 Page 3 of 9 345 Fig. 1  Patient selection and patient flow diagram considered adequate. The germs with intermediate sen- Interventions sitivity to the administered antibiotic were coded jointly with the resistant ones. Every treatment of a patient The national guidelines for antibiotic treatment in with ≥ 1 resistant germ or with intermediate sensitivities intra-abdominal infection were followed, mainly to the antibiotics received was labeled as inadequate. using cefotaxime associated with metronidazole or 13 345 Page 4 of 9 Langenbeck's Archives of Surgery (2023) 408:345 piperacillin-tazobactam [11]. In patients with severe Surgery infections and risk factors for a poor outcome, empiri- cal treatment with carbapenems was indicated, with or Most of the cohort (95%) was operated on, with percuta- without association to linezolid or antifungals. Surgical neous cholecystostomy and/or antibiotic therapy used in a patients underwent laparoscopic or open cholecystectomy minority of the frailest patients (n = 13). Cholecystectomy depending on the surgeon’s choice. Intraoperative cultures was completed laparoscopically in 74% of patients. were taken from peritoneal fluid and intravesicular bile. Veress’ needle puncture of the gallbladder was used in laparoscopic procedures, and direct needle aspiration was Cultures obtained used in open surgery. Blood cultures were taken from both peripheral blood and/or central venous catheters. Nonsur- Some 553 (76%) patients had a pre- or intraoperative culture gical patients were treated with either antibiotic therapy obtained from bile (n = 495; 68%), peritoneal fluid (n = 148; alone or ultrasound-guided cholecystostomy using an 8Fr 20%), and/or blood (n = 138; 19%). Patients in whom cul- catheter (SKATER ™, Argon Medical Devices, Rochester, tures were obtained were more frequently males, older, had NY, USA) by transhepatic or transperitoneal insertion. a higher rate of comorbidities, required more vasoactive drugs, and had a higher severity degree according to the TG18. Surgical initial treatment was chosen significantly Statistical analysis more often in cultured patients. Those patients suffered a higher complication rate compared to those in whom cul- Univariate analysis of the association among qualitative tures were not obtained, including more infectious compli- variables was carried out using the chi-square test or Fish- cations, surgical site infections, and severe complications er’s test as needed, with results given as odds ratios with (grade > IIIa according to Clavien‒Dindo classification) 95% confidence intervals (CIs). [11]. Mortality was almost fourfold in the group where the The normality of the distribution of quantitative vari- cultures were taken, although statistical significance was not ables was estimated by the Kolmogorov‒Smirnov test reached (Table 1). and none followed a normal distribution. Therefore, these Germs were isolated in 283 (51%) patients in which cul- variables were summarized by their median and interquar- tures were obtained. The positivity of the culture varied tile range (IQR), and the comparison between nonpaired depending on the sample obtained. Germs were isolated in means was carried out using Mann‒Whitney’s U test. 50% of bile cultures, 33% of blood cultures, and 30% of The multivariate analysis was completed with a forward peritoneal fluid. conditional (F to enter 0.05, to remove 0.10) binary logis- Patients in whom an extra biliary culture was obtained tic regression, entering the variables with a significant (peritoneal fluid cultures and/or blood cultures) presented association in the univariate analysis. IBM SPSS Statis- worse results than those in whom only a bile culture sam- tics v. 25 (Rochester, MN, USA) was used in all analyses. ple was obtained. Accordingly, patients with extra biliary cultures experienced a greater severity according to the TG scale (TG-III 54.2% vs. 33.7%; OR, 2.33; 95% CI 1.7–3.3), higher rate of complications (63.6% vs. 33.7%; OR, 3.45; Results 95% CI 2.4–4.9), more severe complications (19.8% vs. 5%; OR, 4.68; 95% CI 2.6–8.6), and eight times higher mortality The median age of the study population was 69 (IQR (8.3% vs. 1%; OR, 8.96; 95% CI 2.6–30.4). 53–80) years old, with 23% being over 80 years old. Of the 781 samples, 420 (54%) grew identifiable germs. The median Charlson Comorbidity Index score was 1.0 The bacteria most frequently isolated were Escherichia coli (from 0 to 10) points (IQR 0–2). Among the most preva- in 26% of total positive cultures (109/420 positive sam- lent comorbidities, a tenth of the patients suffered from ples, or 37% in 107/283 patients), Enterococcus spp. in chronic obstructive pulmonary disease (COPD), and 5% 16% (66/420 positive samples, or 23% in 65/283 patients), had a previous diagnosis of dementia [8]. An additional Klebsiella spp. in 16% (66/420 positive samples, or 22% in file shows this in more detail (see Additional file 2. Main 62/283 patients), Streptococcus spp. in 11% (48/420 positive values population). samples, or 16% in 46/283 patients), Enterobacter spp. in All patients received at least one empiric intravenous 11% (47/420 positive samples, or 16% in 44/283 patients), antibiotic therapy dose. The overall mean duration of ther- Citrobacter spp. in 4% (16/420 positive samples, or 6% in apy was 6.7 days, 4 days in patients without postoperative 16/283 patients), and Clostridium spp. in 3% (14/420 posi- complications, and 10 days on average in patients with tive samples, or 5% in 14/283 patients). Candida spp. were complications. isolated in 4/420 (0.9%). 13 Langenbeck's Archives of Surgery (2023) 408:345 Page 5 of 9 345 Table 1  Differences between groups with cultures performed vs. no ­culturesa Global Culture No culture Pb OR (95% CI) (N = 725) (n = 553) (n = 172) Gender (male %) 381 (53) 309 (55.9) 72 (41.9) 0.001 1.759 (1.244–2.487) Median age (IQR) 69 (53–80) 71 (57–80) 60 (44.25–78) < 0.001c Charlson’s Comorbidity Index Median (IQR) 1 (0–2) 1 (0–2) 0 (0–1) < 0.001c ACME > 80 y/o 166 (23) 131(23.7) 35 (20.3) 0.363 1.215 (0.799–1.849) COPD 71 (10) 60 (10.8) 11 (6.4) 0.086 1.781 (0.914–3.471) Dementia 33 (5) 29 (5.2) 4 (2.3) 0.142 2.324 (0.806–6.707) Vasopressors 30 (4) 28 (5.1) 2 (1.2) 0.026 4.533 (1.069–19.228) Initially surgical treatment 689 (95) 533 (96.4) 156 (90.7) 0.003 2.733 (1.383–5.402) TG I 149 (21) 96 (17.4) 53 (30.8) 0.001d 1.651d (1.148–2.374) TG II 284 (39) 219 (39.6) 65 (37.8) TG III 292 (40) 238 (43) 54 (31.4) Overall complications 313 (43) 262 (47.4) 51 (29.7) 0.001 2.136 (1.470–3.084) Infectious complications 133 (18) 114 (20.6) 19 (11) 0.005 2.091 (1.244–3.515) SSI complications 80 (11) 71 (12.8) 9 (5.2) 0.005 2.668 (1.304–5.458) Severe complications > ­IIIae 74 (10) 65 (11.8) 9 (5.2) 0.014 2.412 (1.175–4.952) Mortality 26 (3.6) 24 (4.3) 2 (1.2) 0.050 3.856 (0.902–16.486) FTR 26/313 (8.3) 24/262 (9.2) 2/51 (3.9) 0.277 2.471 (0.565–10.798) OR odds ratio, CI confidence interval, IQR interquartile range, y/o years old, ACME acute cholecystitis mortality estimation, chronic obstructive pulmonary disease, TG Tokyo Guidelines, SSI surgical site infection, FTR failure to rescue a N (%) b Chi-square test c Mann-Whitney’s U test d Tokyo Guidelines grade I + II vs. Tokyo Guidelines grade III e According to Clavien-Dindo Classification Positive cultures cluster in the frequency of inadequacy was Enterococcus spp. involved in one third of cases (n = 30). As shown in Table 2, patients with positive cultures were Patients with adequate EAT had a considerably lower older and had greater comorbidity. They presented a higher mortality rate than those who received inadequate treatment rate of overall complications and infectious complications (3.2% vs. 12.8%; OR, 0.22; 95% CI 0.1–0.6), both in cases of compared to those in which culture(s) had been taken but no mild ACC-moderate (mortality 2% vs. 7.4%; OR, 0.26; 95% germs were isolated. Severe complications were 2.5 times CI 0.05–1.5) and more strikingly in severe TG-III cases of higher in the culture-positive group, and mortality almost ACC (mortality 4.4% vs. 20%; OR, 0.18; 95% CI 0.1–0.7). tripled in these cases (mortality 6.4% vs. 2.2%; OR, 2.99; As shown in Table 3, global complications, severe complica- 95% CI 1.2–7.7). tions (Clavien-Dindo > IIIa), infectious complications, and surgical site infections were more frequent in patients with inadequate EAT, although statistical significance was not Impact of adequacy reached. Nonadequacy was an independent variable predicting In those patients with positive cultures, the adequacy or mortality in a binary logistic regression model that also inadequacy of the empiric antibiotic treatment received was included the four variables of ACME [8] with a significant analyzed (Table 3). association with mortality in the univariate analysis: age Gram-negative bacilli from the Enterobacteriaceae fam- over 80 years old, the requirement of vasopressor therapy, ily were the germs most frequently involved in inadequacy and the preoperative diagnosis of dementia and/or chronic (n = 55), especially those producing extended-spectrum obstructive pulmonary disease (COPD). Thus, as shown in beta-lactamases (ESBLs), which were found in 36 patients Table 4, patients with inadequate EAT were 4.4 times more (12.7% of patients with positive cultures). They were mainly likely to die than those with adequate EAT (OR, 4.4; 95% E. coli, Enterobacter spp., and Citrobacter spp. The second CI 1.3–15.3). 13 345 Page 6 of 9 Langenbeck's Archives of Surgery (2023) 408:345 Table 2  Differences between groups with isolated germs vs. no germ ­isolationa Culture Positive cultures Negative cultures Pb OR (95% CI) (n = 553) (n = 283) (n = 270) Gender (male %) 309 (55.9) 160 (56.5) 149 (55.2) 0.749 1.056 (0.755–1.478) Median age (IQR) 71 (57–80) 75 (66–82) 65.5 (49–79) < 0.001c Charlson’s Comorbidity Index Median (IQR) 1 (0–2) 1 (0–3) 0 (0–2) < 0.001c ACME > 80 y/o 131(23.7) 78 (27.6) 53 (19.6) 0.028 1.558 (1.047–2.319) COPD 60 (10.8) 35 (12.4) 25 (9.3) 0.240 1.383 (0.804–2.380) Dementia 29 (5.2) 18 (6.4) 11 (4.1) 0.228 1.599 (0.741–3.452) Vasopressors 28 (5.1) 19 (6.7) 9 (3.3) 0.070 2.087 (0.927–6.698) Initially surgical treatment 533 (96.4) 270 (95.4) 263 (97.4) 0.208 0.553 (0.217–1.407) TG I 96 (1.4) 42 (14.8) 54 (20) 0.114d 1.313d (0.937–1.840) TG II 219 (39.6) 110 (38.9) 109 (40.4) TG III 238 (43) 131 (46.3) 107 (39.6) Overall complications 262 (47.4) 155 (54.8) 107 (39.6) 0.001 1.845 (1.316–2.587) Infectious complications 114 (20.6) 75 (26.5) 39 (14.4) 0.001 2.136 (1.389–3.283) SSI complications 71 (12.8) 43 (15.2) 28 (10.4) 0.090 1.549 (0.931–2.574) Severe complications > ­IIIae 65 (11.8) 46 (16.3) 19 (7) 0.001 2.564 (1.460–4.503) Mortality 24 (4.3) 18 (6.4) 6 (2.2) 0.017 2.989 (1.168–7.647) FTR 24/262 (9.2) 18/155 (11.6) 6/107 (5.6) 0.098 2.212 (0.848–5.771) OR odds ratio, CI confidence interval, IQR interquartile range, y/o years old, ACME acute cholecystitis mortality estimation, COPD chronic obstructive pulmonary disease, TG Tokyo Guidelines, SSI surgical site infection, FTR failure to rescue a N (%) b Chi-square test c Mann-Whitney’s U test d Tokyo Guidelines grade I + II vs. Tokyo Guidelines grade III e According to Clavien-Dindo Classification Discussion Empirical antibiotic treatment of acute cholecystitis The current clinical study investigates the impact of not pre- In recent years, few studies have been published on the scribing adequate empirical therapy to patients treated for microbiology of acute calculous cholecystitis and empirical acute cholecystitis. Most of the cohort (95%) was operated antibiotic therapy. In a registry-based (CIAO and CIAOW) on, with percutaneous cholecystostomy and/or antibiotic multicentric observational study, the group of Coccolini therapy used in a minority of the frailest patients (n = 13). et al. accurately described the antibiotic resistance pattern in an acute cholecystitis subpopulation (n = 567). They Selection of the patient and the sample for culture also found a significant association between mortality and and determination of adequacy inadequacy in the univariate analysis. Their population was decidedly heterogeneous, and nevertheless, their main con- As stated in the TG18, in our department, it was not com- clusion agrees with the current study, although inadequacy pulsory to culture every sample (bile, blood and peritoneal was not identified as an independent factor in their multivari- fluid) from every patient [3]. However, surgical teams tended ate analysis [6]. both to obtain cultures more often from severe patients and Suh et al. [7] described the trends in microbial etiology to obtain blood and peritoneal fluid cultures (in addition to and antibiotic resistance over the years focusing solely on bile cultures) from the frailest patients. This obvious obser- bile cultures and the importance of the adequacy of anti- vational bias guided our subgroup analysis, assessing the biotic therapy in a selected and well-analyzed population. effect of inadequacy in mild/moderate or severe groups of Unfortunately, unlike the current study, their group included ACC separately. only mild and moderate acute cholecystitis, excluding Although there is no international standard to define EAT patients with severe acute cholecystitis (TG18 grade III). In adequacy, we used the most widely employed definition of addition, patients with acalculous cholecystitis or with con- EAT adequacy in the English scientific literature [12, 13]. current pancreatitis were included, increasing heterogeneity. 13 Langenbeck's Archives of Surgery (2023) 408:345 Page 7 of 9 345 Table 3  Differences between groups who received adequate vs. inadequate empiric ­treatmenta Positive cultures Adequate treatment Non-adequate treatment Pb OR (95% CI) (n = 283) (n = 189) (n = 94) Gender (male %) 160 (56.5) 106 (56.1) 54 (57.4) 0.828 0.946 (0.574–1.559) Median age (IQR) 75 (66–82) 72 (63.5–80) 78 (71.75–85) 0.001c Charlson’s Comorbidity Index Median 1 (0–3) 1 (0–2) 1 (0–3) 0.209c (IQR) ACME > 80 y/o 78 (27.6) 45 (23.8) 33 (35.1) 0.045 0.578 (0.337–0.991) COPD 35 (12.4) 19 (10.1) 16 (17) 0.094 0.545 (0.266–1.116) Dementia 18 (6.4) 10 (5.3) 8 (8.5) 0.296 0.601 (0.229–1.576) Vasopressors 19 (6.7) 13 (6.9) 6 (6.4) 0.875 1.083 (0.398–2.947) Initially surgical treatment 270 (95.4) 181 (95.8) 89 (94.7) 0.681 1.271 (0.404–3.997) TG I 42 (14.8) 26 (13.8) 16 (17) 0.618d 1.254d (0.761–2.064) TG II 110 (38.9) 72 (38.1) 38 (40.4) TG III 131 (46.3) 91 (48.1) 40 (42.6) Overall complications 155 (54.8) 98 (51.9) 57 (60.6) 0.162 0.699 (0.423–1.156) Infectious complications 75 (26.5) 46 (24.3) 29 (30.9) 0.242 0.721 (0.416–1.249) SSI complications 43 (15.2) 24 (12.7) 19 (20.2) 0.097 0.574 (0.296–1.112) Severe complications > ­IIIae 46 (16.3) 27 (14.3) 19 (20.2) 0.203 0.658 (0.344–1.257) Mortality 18 (6.4) 6 (3.2) 12 (12.8) 0.002 0.224 (0.081–0.618) FTR 18/155 (11.6) 6/98 (6.1) 12/57 (21.1) 0.005 0.245 (0.086–0.694) OR odds ratio, CI confidence interval, IQR interquartile range, y/o: years old, ACME acute cholecystitis mortality estimation, COPD chronic obstructive pulmonary disease, TG Tokyo Guidelines, SSI surgical site infection, FTR failure to rescue a N (%) b Chi-square test c Mann-Whitney’s U test d Tokyo Guidelines grade I + II vs. Tokyo Guidelines grade III e According to Clavien-Dindo Classification Table 4  Binary logistic regression on mortality between adequate vs. in urgent cholecystectomies for uncomplicated mild or non-adequate EAT groups moderate ACC [14]. They observed a higher complication B (SE) P OR 95% CI rate, especially infectious complications. Thus, although surgery plays a key role in the control of the infectious EAT ­adequacya 1.479 (0.638) 0.020 4.391 1.258–15.322 focus in mild and moderate phases, we cannot disregard Age > 80 y/o 1.867 (0.644) 0.004 6.466 1.830–22.855 the central role of antibiotics, even in these early stages Vasopressors use 2.372 (0.758) 0.002 10.716 2.426–47.323 of the disease. Dementia 1.693 (0.860) 0.049 5.436 1.008–29.316 To our knowledge, there is a notorious lack of relevant COPD 1.707 (0.673) 0.011 5.512 1.475–20.603 information on the role of antibiotic adequacy in severe B(SE) coefficient for the constant (standard error), OR odds ratio, CI acute cholecystitis. In the current study, we wanted to fill confidence interval, EAT empirical antibiotic therapy, y/o years old, this gap, including in the analysis the most severe patients, COPD chronic obstructive pulmonary disease since they bear the greatest margin for improvement both a Inadequate or unknown/unclear treatment vs. adequate treatment in mortality and morbidity. On the other hand, in recent years, frailty has acquired They could not draw a relationship between inappropriate indisputable relevance when making therapeutic deci- empirical antibiotics and clinical results, probably because sions in elderly patients. Several frailty scales have been severe cases were not included. Despite these limitations described to classify patients trying to facilitate the sur- and coinciding with our results, they conclude that surgical geon’s decision in selecting which patients, despite their control of the infectious focus is essential. advanced age, could benefit from surgical treatment. In In PEANUTS II, a multicenter randomized clinical this subgroup of elderly patients, perhaps less likely to trial of noninferiority, clinical results were worse in the receive a final surgical treatment, the role of antibiotic group without preoperative antibiotics compared with the therapy appears essential to achieve control of the infec- group who did receive an empiric prophylactic antibiotic tious focus [15–17]. 13 345 Page 8 of 9 Langenbeck's Archives of Surgery (2023) 408:345 Multiresistant bacteria empirical broad-spectrum antibiotic therapy securing cov- erage of ESBL and Enterococcus spp. The worrisome growth rate of multidrug-resistant bacteria in Supplementary Information The online version contains supplemen- our environment is well known. Other groups have already tary material available at https://d​ oi.o​ rg/1​ 0.1​ 007/s​ 00423-0​ 23-0​ 3063-4. described the presence of ESBLB in patients with acute cholecystitis, and the current study confirms these findings. Authors’ contributions All authors contributed to the study conception The presence of Enterococcus spp. is also noteworthy in and design. Material preparation, data collection, and analysis were performed by Maite de Miguel-Palacio, Ana-María González-Castillo, a third of the patients who did not receive adequate initial María-José Pons-Fragero, and Ricard-Josep Morera-Casaponsa. The EAT [6, 7, 18–21]. first draft of the manuscript was written by Maite de Miguel-Palacio, and all authors commented on previous versions of the manuscript. All Clinical guidelines authors read and approved the final manuscript. Funding Open Access Funding provided by Universitat Autonoma de The Spanish guidelines with recommendations for intra- Barcelona. abdominal infection [22] and the revised Guidelines of the Surgical Infection Society [23] emphasize limiting the dura- Data availability The datasets used and/or analyzed during the cur- rent study are available from the corresponding author on reasonable tion of antibiotic treatment but fail to offer differential thera- request. peutic options for severe ACC patients. It is possible that if the EAT protocols are modified to empir- Declarations ically include broad-spectrum antibiotics in unselected patients, we will witness a surge of new antibiotic resistance. However, Ethics approval and consent to participate The study met the stand- ards outlined in the Declaration of Helsinki and Good Epidemiological we believe that by widening the spectrum in severe patients, Practices. The study was conducted with the approval of the Ethics morbidity and mortality will improve. On the other hand, the Committee of Hospital del Mar in Barcelona (Spain). Committee’s duration of EAT could be drastically reduced, following the Reference Number: 2020/9312. Code: MMP-ANT-2020–01. recommendations for short-term therapy and de-escalating Consent for publication Not applicable. treatments as soon as germ isolation is confirmed [24]. Conflict of interest The authors declare no competing interests. Limitations of the study Competing interests The authors declare no competing interests. The current study presents the limitations of retrospective cohort studies, such as the possible selection bias of cases, Open Access This article is licensed under a Creative Commons Attri- since it is a single-center study that covers a specific popula- bution 4.0 International License, which permits use, sharing, adapta- tion of the city of Barcelona, and the lack of data on some tion, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, variables. However, the data were collected and reviewed provide a link to the Creative Commons licence, and indicate if changes by two authors, which reduces the risk of misclassification, were made. The images or other third party material in this article are increases reliability and reduces interobserver differences. included in the article's Creative Commons licence, unless indicated The decision to obtain cultures and from where the samples otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not were taken was entirely left to the surgical team’s preference. permitted by statutory regulation or exceeds the permitted use, you will The results imply, however, that the patient and site selection need to obtain permission directly from the copyright holder. To view a to obtain cultures were not random, as cultured patients had copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. worse clinical results, suggesting a positive selection bias. Conclusion References There was a greater tendency to perform cultures in older 1. Yokoe M, Hata J, Takada T, Strasberg SM, Asbun HJ, Waka- bayashi G et al (2018) Tokyo Guidelines 2018: diagnostic criteria and male patients, with more severe ACC. Patients with pos- and severity grading of acute cholecystitis (with videos). J Hepa- itive cultures developed more complications and higher mor- tobiliary Pancreat Sci 25(1):41–54 tality than patients without germ isolation. Non-adequate 2. Okamoto K, Suzuki K, Takada T, Strasberg SM, Asbun HJ, Endo I empirical antibiotic therapy was associated with a higher et al (2018) Tokyo Guidelines 2018: flowchart for the management of acute cholecystitis. J Hepatobiliary Pancreat Sci 25(1):55–72 mortality rate, especially in severe ACC TG18 grade III. 3. Gomi H, Solomkin JS, Schlossberg D, Okamoto K, Takada T, Therefore, adequate empirical antibiotic therapy appears Strasberg SM et al (2018) Tokyo Guidelines 2018: antimicrobial to be of paramount importance in ACC, and those patients therapy for acute cholangitis and cholecystitis. J Hepatobiliary with severe ACC will probably benefit from a wider Pancreat Sci 25(1):3–16 13 Langenbeck's Archives of Surgery (2023) 408:345 Page 9 of 9 345 4. Mosdell DM, Morris DM, Voltura A, Pitcher DE, Twiest MW, results from the UK Observational Emergency Laparotomy and Milne RL et al (1991) Antibiotic treatment for surgical peritonitis. Frailty (ELF) Study. Ann Surg 273(4):709–718 Ann Surg 214(5):543–549 16. Castillo-Angeles M, Cooper Z, Jarman MP, Sturgeon D, Salim 5. Membrilla-Fernández E, Sancho-Insenser JJ, Girvent-Montllor M, A, Havens JM (2021) Association of frailty with morbidity and Álvarez-Lerma F, Sitges-Serra A (2014) Effect of initial empiric mortality in emergency general surgery by procedural risk level. antibiotic therapy combined with control of the infection focus on JAMA Surg 156(1):68–74 the prognosis of patients with secondary peritonitis. Surg Infect 17. Rosa F, Covino M, Cozza V, Fransvea P, Quero G, Fiorillo C et al (Larchmt) 15(6):806–814 (2021) Management of acute cholecystitis in elderly patients: a 6. Coccolini F, Sartelli M, Catena F, Montori G, di Saverio S, Sugrue propensity score-matched analysis of surgical vs. medical treat- M et al (2015) Antibiotic resistance pattern and clinical outcomes ment. Dig Liver Dis 53(12):1620–6 in acute cholecystitis: 567 consecutive worldwide patients in a 18. Darkahi B, Sandblom G, Liljeholm H, Videhult P, Melhus Å, prospective cohort study. Int J Surg 1(21):32–37 Rasmussen IC (2014) Biliary microflora in patients undergoing 7. Suh SW, Choi YS, Choi SH, Do JH, Oh HC, Kim HJ et al (2021) cholecystectomy. Surg Infect (Larchmt) 15(3):262–265 Antibiotic selection based on microbiology and resistance profiles 19. Gil Fortuño M, Granel Villach L, Sabater Vidal S, Soria Martín of bile from gallbladder of patients with acute cholecystitis. Sci R, Martínez Ramos D, Escrig Sos J et al (2019) Microbiota biliar Rep 11(1):2969 en pacientes colecistectomizados: Revisión de la antibioterapia 8. González-Castillo AM, Sancho-Insenser J, de Miguel-Palacio M, empírica. Rev Esp Quimioter 32(5):426–31 Morera-Casaponsa JR, Membrilla-Fernández E, Pons-Fragero MJ 20. Hadavand F, Moradi H 2021 Evaluation of bile microbiology and et al (2021) Mortality risk estimation in acute calculous cholecys- its pattern of resistance in patients with acute cholecystitis. Arch titis: beyond the Tokyo Guidelines. World J Emerg Surg 16(1):24 Clin Infect Dis. 16(6) 9. Hirota M, Takada T, Kawarada Y, Nimura Y, Miura F, Hirata K 21. Nitzan O, Brodsky Y, Edelstein H, Hershko D, Saliba W, Keness et al (2007) Diagnostic criteria and severity assessment of acute Y et al (2017) Microbiologic data in acute cholecystitis: ten years’ cholecystitis: Tokyo Guidelines. J Hepatobiliary Pancreat Surg experience from bile cultures obtained during percutaneous chol- 14(1):78–82 ecystostomy. Surg Infect (Larchmt) 18(3):345–349 10. EUCAST Breakpoint Tables. The European Committee on Antimi- 22. Guirao X, Arias J, Badía JM, García-Rodríguez JA, Mensa J, crobial Susceptibility Testing. Breakpoint tables for interpretation of Álvarez-Lerma F et al (2009) Recommendations in the empiric MICs and zone diameters. Version 8.1, 2018. Available from http://​ anti-infective agents of intra-abdominal infection. Rev Esp Qui- www.​eucast.​org mioter 22(3):151–72 11. Dindo D, Demartines N, Clavien PA (2004) Classification of sur- 23. Mazuski JE, Tessier JM, May AK, Sawyer RG, Nadler EP, Rosen- gical complications: a new proposal with evaluation in a cohort of gart MR et al (2017) The surgical infection society revised guide- 6336 patients and results of a survey. Ann Surg 240(2):205–213 lines on the management of intra-abdominal infection. Surg Infect 12. Kollef MH (2000) Inadequate antimicrobial treatment: an impor- (Larchmt) 18(1):1–76 tant determinant of outcome for hospitalized patients. Clin Infect 24. Membrilla-Fernández E, Gómez-Zorrilla S, González-Castillo Dis 31(SUPPL. 4):131–138 AM, Pelegrina-Manzano A, Guzmán-Ahumada J, Prim N et al 13. Kollef MH, Sherman G, Ward S, Fraser VJ (1999) Inadequate (2022) Scientific evidence of the duration of antibiotic treatment antimicrobial treatment of infections: a risk factor for hospital in intra-abdominal infections with surgical focus control. Cir Esp mortality among critically III patients. Chest 115(2):462–474 100(10):608–613 14. van Braak WG, Ponten JEH, Loozen CS, Schots JPM, van Geloven AAW, Donkervoort SC et al (2022) Antibiotic Publisher's note Springer Nature remains neutral with regard to prophylaxis for acute cholecystectomy: PEANUTS II mul- jurisdictional claims in published maps and institutional affiliations. ticentre randomized non-inferiority clinical trial. Br J Surg 109(3):267–273 15. Parmar KL, Law J, Carter B, Hewitt J, Boyle JM, Casey P et al (2021) Frailty in older patients undergoing emergency laparotomy: 13