July 2020: Antimicrobial Stewardship and Infection Control

Antimicrobial guidelines adherence was associated with shorter length of stay and lower 30-day mortality in patients admitted with cellulitis and CAP. The authors randomized their 4 hospitals with a crossover design to use of a weighted incidence syndromic combination antibiogram (WISCA – essentially you take antibiogram data stratified by infectious syndrome and including patient variables that predict resistance like age, prior culture data, and prior antibiotic exposure, then use that to create a model that predicts the efficacy of various empiric regimens for a given patient with given infection and risk factors) versus standard care.  Specifically, an ASP team member called primary teams with the WISCA intervention arm and did nothing in the standard of care arm. The WISCA was developed for adults with UTI, intraabdominal infection, pneumonia (which they subdivided as CAP, aspiration, and “Nursing home-associated”), and nonpurulent cellulitis- in the cases of cellulitis and CAP, recommendations were based on IDSA guidelines and the consensus group of ID physicians.  Patients with an unclear source of infection, multiple possible foci of infection, or no infection in the opinion of the ASP physician were excluded.  The primary outcome was length of stay, and secondary outcomes included 30-day mortality, readmission, CDI, acquisition of MDROs, and cost. O

The authors enrolled 6,849 patients with one of the above infections between 2015 and 2018.  The ASP physician did or would have recommended a change in antibiotics in 31% of cases, and an ID consult in 1.4% of cases.  Interestingly, the antibiotics recommended often looked a lot like those actually used (pip-tazo and amp-sulbactam frequently appeared in both categories), which may explain the lack of benefit of the intervention. There were no differences in LOS or any of the secondary outcomes between the study arms, and this was also true of the patients in whom ASP recommendations were made and followed vs not.  HOWEVER, patients recommended guidelines-based therapy for nonpurulent cellulitis had shorter lengths of stay (coeff estimate 0.53; p=0.02) and patients recommended guieliens-based therapy for CAP had lower 30-day mortality (aOR 0.58, p=0.02) versus controls.

So, this WISCA proved not to shift antibiotic usage in a way that meaningfully affected patient outcomes, but adherence to existing guidelines was beneficial in cellulitis and CAP. PMID: 32712674

A similar study this month used a pre-/post-intervention design to evaluate the addition of a multivariate model prediction tool for antibiotic resistance of gram-negative bacteremia to standard care.  A total 383 patients (n=182 before the intervention and n=201 after) were included.  Patients who received therapy informed by the prediction tool had higher rates of antimicrobial de-escalation from initial antimicrobials (29% vs 21%, aOR 1.8 with p=0.02) and were more likely to already be on the narrowest appropriate antimicrobial when cultures finalized (55$ vs 44%; aOR 2.0 with p=0.003).  However, time to adequate therapy was no different between groups, and patient centered outcomes like LOS, mortality, readmission, CDI, etc were not investigated. PMID: 32640028

Which mask decontamination and re-use protocols are actually backed by data?  This systematic review identified 15 studies that examined decontamination of N95s and reported outcomes related to eradication of pathogens, preservation of filtration efficiency, and/or physical structural integrity.  The authors identified four strategies supported by these studies.  The strategies included UV treatment (n=9), moist heat (n=5), microwave generated steam (n=4), and hydrogen peroxide vapor (n=4).  Specifically, UV irradiation was shown to inactivate influenza and Bacillus and not to affect filtration efficiency, but did damage the mask’s structure at higher doses.  Moist inactivated influenza when used in a “pre-warmed sealable container” (academic speak for a dutch oven with a rack inside it) but not when masks were just put in a lab incubator; filtration efficiency was not affected but again, this method caused structural damage to the mask in several studies.  Microwave generated steam (academic speak for microwaving a mask inside a pipette box with some water for 2min on ‘high’) and hydrogren peroxide vapor both inactivated influenza without decreasing filtration efficiency, and had minimal effect on mask structure.  PMID: 32729444

How often do Staphylococcus aureus infections complicate elective surgery? In a retrospective study of nearly 885,000 elective inpatient surgery discharges in the US between 2010 and 2015, the incidence of S. aureus bacteremia or surgical site infection within six months after surgery was 1% (0.7% surgical site and 0.3% bloodstream).  While this may not sound too bad, the authors estimate that the US’s estimated 4.2 million elective surgical admissions result in approximately 55,764 S. aureus infections. So, we ID docs have some work to do, and job security in the meantime. PMID: 32634829

What challenges do we face in convincing clinicians to stop reaching for quinolones and late-generation cephalosporins? In this study, the authors interviewed 64 clinicians across 4 Eastern US hospitals, including physicians, pharmacists, APPs, and nurses.  The interviewees cited easy dosing/adherence, broad coverage, and being ‘easy to make decisions about’ in the context of time pressure.  They also described reluctance to stop these drugs once started, citing inertia and fear, and in general felt that the risk of antibiotic undertreatment was usually greater than the risk of C.difficile colitis. PMID: 32468967