Column Editor
Weill Cornell Medicine
New York City
Executive Director
Surgical Infection Society
Foundation for Research and Education
Antibiotic Prophylaxis in Breast Cancer Surgery
Stallard S, Savioli F, McConnachie A, et al. Antibiotic prophylaxis in breast cancer surgery (PAUS trial): Randomised clinical double-blind parallel-group multicentre superiority trial. Br J Surg. 2022;109(12):1224-1231. doi:10.1093/bjs/znac280
Summary: Patients with invasive breast cancer undergoing primary surgery were randomized to either a single bolus dose of amoxicillin-clavulanic acid 1.2 g IV or placebo, after induction of anesthesia. The hypothesis was that a single dose of amoxicillin-clavulanic acid would reduce the incidence of surgical site infection (ssI) at 30 days postoperatively. Of 871 evaluable patients, 438 received antibiotic prophylaxis. Seventy-one patients (16.2%) in the antibiotic group developed an ssI by 30 days, whereas 83 controls (19.2%) developed an ssI (odds ratio [OR), 0.82; 95% CI, 0.58-1.15; P=0.25). Stratifying by surgery type (mastectomy [P=0.921] vs. local excision [P=0.152]) showed no effect. In secondary analyses, a body mass index (BMI) of 30 kg/m2 or greater was associated with developing an ssI (OR, 1.67; 95% CI, 1.04-2.66; P=0.038). The risk for ssI increased 29% for every 5 kg/m2 increase in BMI (OR, 1.29; 95% CI, 1.10-1.52; P=0.003). Carriers of Staphylococcus aureus also had an increased risk for ssI (OR, 2.43; 95% CI, 1.54-3.80; P<0.001). However, there was no benefit of preoperative antibiotics for patients with a high BMI or who were carriers of S. aureus.
Commentary: Breast surgery is considered clean (Class I incision), even if reoperative (e.g., reexcision for a positive margin). Despite this, some data do suggest that antibiotic prophylaxis before breast surgery is protective against developing an ssI, but not herein. In this study, there was no significant or clinically meaningful reduction of ssI at 30 days following breast cancer surgery, regardless of extent, after a single preoperative dose of amoxicillin-clavulanic acid. That this was powered as a superiority trial lends credence. Moreover, patients at high risk (overweight/obese, or carriers of S. aureus) gained no benefit. There are several caveats. Foremost, the ssI rate in both groups is higher than expected (standard CDC diagnostic criteria were used). Furthermore, the pitfall of taking an action based on a secondary study end point is underscored: Just because a risk factor has been identified, it does not follow necessarily that intervention will be mitigating.
Oral Antibiotic Prophylaxis for Elective Colorectal Surgery
Futier E, Jaber S, Garot M, et al.; COMBINE study group. Effect of oral antimicrobial prophylaxis on surgical site infection after elective colorectal surgery: Multicentre, randomised, double blind, placebo controlled trial. BMJ 2022;379:e071476. doi:10.1136/bmj-2022-071476
Summary: In a multicenter, double-blind, placebo-controlled trial, 926 adults (at 11 French hospitals) undergoing elective colorectal surgery were randomized to either a single dose of ornidazole 1 g (n=463; similar to metronidazole) or placebo orally 12 hours before surgery, in addition to IV antimicrobial prophylaxis before surgical incision. The primary outcome was the proportion of patients with surgical site infection (ssI) within 30 days afterward. Secondary outcomes were stratified by type of ssI and major postoperative complications (Clavien-Dindo grade >3) within 30 days. The mean age of participants was 63 years and 554 (60%) were male. An ssI within 30 days occurred in 60 of the 463 patients (13%) after ornidazole and in 100 of the 463 patients (22%) after placebo (absolute difference, –8.6%; 95% CI, –13.5% to –3.8%; relative risk, 0.60; 95% CI, 0.45-0.80). In the ornidazole group, 4.8% of patients had a deep incisional ssI compared with 8% in the placebo group (absolute difference, –3.2%; 95% CI, –6.4% to –0.1%). The proportion of patients with organ-space ssI was 5% after ornidazole compared with 8.4% after placebo (absolute difference, –3.4%; 95% CI, –6.7% to –0.2%). Major complications affected 9.1% of patients in the ornidazole group compared with 13.6% after placebo (absolute difference, –4.5%; 95% CI, –8.6% to –0.5%). Each of the confidence intervals was significant.
Commentary: October 2023 will mark the 50th anniversary of what came to be called the “Nichols-Condon” oral antibiotic bowel prep. Countless lives were saved by this protocol, yet enthusiasm waned congruent with the rise of same-day major surgery, the thought being that with outpatient mechanical bowel preparation, patients would also have to ingest two antibiotics on a precise multidose schedule, possibly while having diarrhea from the mechanical prep that might render the regimen ineffective. So, the pendulum swung away, but ssI rates soon increased. Now the pendulum has swung firmly back, thanks to confirmatory studies such as this one. Among adults undergoing elective colorectal surgery, IV antibiotic and a single dose of ornidazole 1 g significantly reduced ssI before surgery, compared with placebo. Reported ssI rates are consistent with the literature. This is a well-done, credible study. To adopt this simple single-dose protocol, one would substitute metronidazole in the United States (where ornidazole is not marketed), so dosing might need to be closer (e.g., 6-8 hours) before surgery, owing to a shorter half-life. Ornidazole is also tolerated better from a GI perspective.
Septic Shock in Advanced Cancer
Cuenca JA, Manjappachar NK, Ramirez CM, et al. Outcomes and predictors of 28-day mortality in patients with solid tumors and septic shock defined by Third International Consensus Definitions for sepsis and septic shock criteria. Chest. 2022;162(5):1063-1073. doi:10.1016/j.chest.2022.05.017
Summary: A total of 271 solid tumor patients admitted to the ICU with septic shock were studied retrospectively. A reduced multivariable logistic regression model was built to identify independent predictors of 28-day mortality. Patients’ median age was 62 years; 57% were male and 54% were white. The most common primary tumors were of lung (19%), breast (8%), pancreatic (8%) or colorectal (7%) origin; most patients (85%) harbored metastatic disease. Mortality was 69% at 28 days. Non-survivors showed a higher rate of advanced cancer, longer pre-ICU hospital stays, and greater organ dysfunction at admission and throughout the ICU stay (P<0.001 for all). Independent predictors of 28-day mortality by multivariable analysis included metastatic disease (odds ratio [OR], 3.17; 95% CI, 1.43-7.03), respiratory failure (OR, 2.34; 95% CI, 1.15-4.74), elevated lactate concentrations (OR, 3.19; 95% CI, 1.90-5.36), and Eastern Cooperative Oncology Group performance scores of 3 or 4 (OR, 2.72; 95% CI, 1.33-5.57). Only 14% of patients were discharged home without medical assistance.
Commentary: The data are retrospective and represent a fairly homogeneous patient population, but the most current definition of sepsis (infection causing organ dysfunction) was used, and there are no similar studies available. The mortality rate is perhaps twice as high as might be expected from a similar degree of organ dysfunction in a general ICU population. Fewer than 25% of the patients were alive at 90 days. At the time of death, almost 90% of the patients had had life support curtailed, and 60% had been transitioned to comfort care. This scenario is common: Imagine a patient with a perforated hollow viscus and septic shock from peritonitis due to a primary GI neoplasm. The implications are huge. These patients do poorly. That may not be surprising, but just how exceptionally poorly might be. This knowledge should inform candid, compassionate goals-of-care discussions when in such a circumstance.
Prolonged Antibiotic Prophylaxis Not Beneficial in Pediatric Surgery
He K, Nayak RB, Allori AC, et al. Correlation between postoperative antimicrobial prophylaxis use and surgical site infection in children undergoing nonemergent surgery. JAMA Surg. 2022;157(12):1142-1151. doi:10.1001/jamasurg.2022.4729
Summary: This multicenter cohort study (40,611 patients [47.3% female; median age, 7 years], 93 centers, from June 2019 to June 2021) analyzed 30-day postoperative surgical site infection (ssI) data from ACS NSQIP-Pediatric, augmented with antibiotic use data obtained through supplemental medical record review. Participants were children (<18 years of age) undergoing nonemergent surgery. Exclusion criteria included allergy, impaired immune function, and preexisting infections requiring IV antibiotics at the time of surgery. The primary outcome was continuation of antimicrobial prophylaxis beyond time of incision closure. Of the patients, 41.6% received postoperative prophylaxis (hospital range, 0-71%). Odds ratios (ORs) for postoperative prophylaxis use ranged 190-fold across hospitals (OR, 0.10-19.30) and ORs for ssI ranged nearly fourfold (OR, 0.55-1.90). No correlations were found between use of postoperative prophylaxis and ssI rates overall (r=0.13; P=0.20), and when stratified by ssI type (superficial/deep incisional ssI: r=0.08; P=0.43; organ-space ssI: r=0.13; P=0.23). Likewise, no correlations were found when stratified by surgical specialty (general surgery, r=0.02; P=0.83; urology, r=0.05; P=0.64; plastic surgery, r=0.11; P=0.35; otolaryngology, r=-0.13; P=0.25; orthopedic surgery, r=0.05; P=0.61; or neurosurgery, r=0.02; P=0.85).
Commentary: Postoperative antimicrobial prophylaxis is commonplace in pediatric surgery despite consensus guidelines recommending discontinuation following incision closure. Unfortunately, evidence is scant compared with adult populations. Although these retrospective data are subject to the potential biases associated with use of any administrative database, they provide insight not available from other sources. No matter the type of operation, antibiotic prophylaxis protects only the incision, and only when it is open. In this circumstance, it turns out that “kids are not different.”
The Epidemiology of Blood Stream Infections
Gouel-Cheron A, Swihart BJ, Warner S, et al. Epidemiology of ICU-onset bloodstream infection: Prevalence, pathogens, and risk factors among 150,948 ICU patients at 85 U.S. hospitals. Crit Care Med. 2022;50(12):1725-1736. doi:10.1097/CCM.0000000000005662
Fakhry SM, MacLeod K, Shen Y, et al. Bacteremia in trauma: A contemporary analysis of blood culture results and outcomes in 158,884 patients. Surg Infect (Larchmt). 2022;23(9):809-816. doi:10.1089/sur.2022.228
Summary: Gouel-Cheron et al examined ICU-related blood stream infections (BSIs) retrospectively, comparing infection present on ICU admission with ICU-onset infection that included 6,906 episodes among 150,948 patients (4.6%). All patients had a minimum 3-day stay in the ICU. Patients with ICU-onset BSI displayed higher crude mortality (38% [vs. 20% when present on admission]; P<0.001) and longer median ICU length of stay (13 vs. 5 days [P<0.001]). ICU-onset BSI displayed more Pseudomonas, Acinetobacter, Enterococcus, Candida, and Staphylococcus species, including more multidrug-resistant (MDR) isolates. Younger age, male sex, Black/Hispanic race, more comorbidities, trauma, acute pulmonary or gastrointestinal presentations, and pre-ICU exposure to antimicrobial agents were associated with greater ICU-onset BSI risk after adjusted analysis. Risk for ICU-onset BSI manifested with any duration of mechanical ventilation, and seven days or more after insertion of central venous or arterial catheters.
By contrast, Fakhry et al. examined 158,884 trauma activation patients at 89 centers; 1,214 were blood culture-positive (0.7%). Adjusted odds ratio (aOR) for the risk for a positive blood culture included age (aOR, 1.004; 95% CI, 1.001-1.007), male gender (aOR, 1.69; 95% CI, 1.48-1.93), comorbidity count (aOR, 1.24; 95% CI, 1.19-1.28), injury severity (aOR, 1.05; 95% CI, 1.04-1.05), blood transfusion within the first four hours (aOR, 2.13; 95% CI, 1.74-2.60), and lacerated bowel (aOR, 3.91; 95% CI, 2.78-5.50). Crude all-cause mortality was 17.2% for bacteremia; bacteremic patients were more likely to die (aOR, 3.78; 95% CI, 3.17-4.51; P<0.001). The most common isolates were S. epidermidis (14%), methicillin-susceptible S. aureus (12%), and Escherichia coli (6%), with highest mortality associated with Pseudomonas aeruginosa (45%).
Commentary: Comparing papers is fraught even when the subject matter is similar, but these papers are of interest for their sheer size. Considering that trauma was identified by Gouel-Cheron et al. as a risk factor for BSI, there is utility in considering both manuscripts. A common theme is that male patients with medical comorbidities are at risk. The biggest shortcoming is that Fakhry et al. considered only bacterial but not fungal infections, producing an underestimate of the overall burden of BSI in trauma, but one that likely does not account for the substantially higher prevalence in the general ICU population. Trauma patients (not all of whom are in the ICU) had a lower crude mortality rate, perhaps due to the lower prevalence of MDR pathogens.
This article is from the April 2023 print issue.
Please log in to post a comment