ID Academic Digest · Infectious Disease
Antimicrobial Pearls, Resistance & Stewardship
Consolidated from ID board-review and case-conference teaching · reviewed 21 July 2026
Educational study digest pitched at a new ID fellow. Clinical recommendations trace to the cited guidelines and trials; the enrichment box is reviewer synthesis of additional evidence, each claim cited. Not medical advice, and not a substitute for the primary guidelines.
In one line
The stewardship traps that recur on rounds and on the boards are all pattern recognition: an Enterobacter that de-represses its chromosomal AmpC and shreds a “ceftriaxone-susceptible” report mid-course, an “encephalopathy” that is really the cefepime trough, a linezolid optic neuropathy you can never take back, and a C. difficile you under-treat because you skipped the fulminant criteria. Learn the pattern, and the drug choice follows.◆Which Enterobacterales will de-repress AmpC and betray a “susceptible” cephalosporin report?
The organisms that carry an inducible chromosomal ampC gene with a high likelihood of clinically significant de-repression are the ones to fear when a report reads “susceptible.” The classically taught roster of the most relevant de-repressors is Hafnia spp., Enterobacter cloacae complex, Citrobacter freundii, Klebsiella aerogenes, and Yersinia enterocolitica. The subset with the best supporting data — the trio you should never second-guess — is E. cloacae, K. aerogenes, and C. freundii. A second, less-studied group carries the same gene but has thin evidence for how often it matters clinically: Hafnia alvei, Citrobacter youngae, Yersinia enterocolitica, and by convention Serratia marcescens, Morganella morganii, and Providencia spp.
The danger is mechanistic. A third-generation cephalosporin (ceftriaxone, cefotaxime, ceftazidime) is a strong inducer, and even when the isolate first tests susceptible, a de-repressed mutant can be selected during therapy — so the bug that looked treatable on Monday is resistant by Thursday. That is why you do not trust a ceftriaxone-S result for the high-risk trio: you avoid third-generation cephalosporins and treat with cefepime (stable to AmpC hydrolysis and a poor inducer) or a carbapenem. For the less-studied organisms it is reasonable to let susceptibility testing guide you — give ceftriaxone if the isolate is susceptible — except when the bacterial burden is high and source control is limited (endocarditis, central nervous system infection), where the same cefepime-or-carbapenem caution applies.
◆Which antibiotic toxicities do you have to recognise on sight?
Board vignettes rarely name the toxicity; they describe it and expect you to name the drug. The highest-yield moves are the ones that are irreversible or that change the agent: linezolid’s optic neuropathy does not come back, and daptomycin is chemically dead in the lung. Hold the signatures in one grid.
| Agent | Signature toxicities (what the vignette hangs on) |
|---|---|
| Tetracyclines | Photosensitivity; hepatotoxicity in pregnancy; permanent teeth discoloration in children (avoid in pregnancy); vertigo (minocycline); blue-grey skin discoloration with long-term use |
| Linezolid / tedizolid | Reversible cytopenias / marrow suppression (typically after ~2 weeks); lactic acidosis; peripheral neuropathy; IRREVERSIBLE optic neuropathy; serotonin syndrome (weak MAO inhibitor). Tedizolid: lower MRSA MICs (may work if linezolid-resistant), less MAOi effect, higher Vd and protein binding |
| Clindamycin | C. difficile risk; inhibits bacterial toxin synthesis (an asset in toxic-shock states); inducible S. aureus resistance |
| Fluoroquinolones | QTc prolongation; tendinopathy / rupture; C. difficile risk; multivalent-cation chelation (absorption); cartilage & joint effects; CNS toxicity; photosensitivity; dysglycemia (hyper- or hypoglycemia); hypokalemia; S. aureus resistance on monotherapy |
| Rifampin | Hepatitis; orange urine and secretions; potent enzyme-inducer interactions (classically ART, azoles, other quinolones, macrolides); flu-like syndrome with interrupted dosing; low resistance barrier (use only in combination) |
| TMP-SMX | Rash, SJS; photosensitivity; pancytopenia; renal effects (rise in creatinine / reduced clearance, hyperkalemia, acute interstitial nephritis); pregnancy risk (folate / neural-tube development) |
| Vancomycin (IV) | Vancomycin-flushing (“red-person”) syndrome; hypersensitivity / DRESS (T-cell mediated); thrombocytopenia; nephrotoxicity; requires drug monitoring |
| Daptomycin | Inactivated by pulmonary surfactant — never for pneumonia; rhabdomyolysis with CPK rise; pneumonitis; cross-resistance with VISA; large molecule with high protein binding |
◆When can a gram-negative bloodstream infection step down to an oral β-lactam, and how do you pick one?
The oral switch is earned, not automatic. Convert once the patient is clinically improved, the source is controlled, the isolate is susceptible on testing, and the gut is intact and functional. Duration is roughly 7 days for an uncomplicated bacteraemia and 10–14 days otherwise. Which oral agent you choose turns on pharmacokinetics — bioavailability, protein binding, and above all the fraction of the dosing interval that free drug stays above the MIC (fT>MIC): aim for >50% for penicillins and >60–70% for cephalosporins, which is exactly why an under-dosed oral β-lactam fails.
Be honest about the comparative data: fluoroquinolone or TMP-SMX step-downs recur less often than oral β-lactams (recurrence odds ratio ~2.06 favouring FQ/TMP-SMX in one cohort, and ~3.43 for urinary-source bacteraemia), even though all-cause mortality is comparable — and notably the oral β-lactams in those studies were not dosed to maximum. Finally, mind the susceptibility surrogate: cefazolin over-calls resistance to cefpodoxime (it can label a treatable isolate resistant), while ceftriaxone under-calls it (it can pass an isolate that is actually less susceptible), even though ceftriaxone agrees with cefpodoxime more often than cefazolin does. Read whichever marker your lab reports with that bias in mind.
◆Your patient on cefepime is confused on day 3 — sepsis, or the drug?
Cefepime-induced neurotoxicity typically appears 2–4 days after the drug is started and spans a spectrum: acute confusion is the most common presentation (~41%), followed by a depressed level of consciousness (~26%), myoclonus (~10%), convulsive seizures (~7%), and frank encephalopathy (~5.4%). The discriminator is that it is exposure-related: a cefepime trough >20 mg/L raises the risk, which is the rationale for therapeutic drug monitoring in patients who accumulate the drug — classically those with renal impairment on unadjusted doses. The confirmatory, reassuring feature is reversibility: it resolves within 48 hours of stopping cefepime. So in the encephalopathic patient on cefepime, especially with renal dysfunction or a high measured trough, the answer is to stop the drug and expect improvement within two days — not to escalate the sepsis workup reflexively.
◆How do you grade C. difficile, and what do you start in 2026?
The definitions are the testable part, because they drive escalation. Severe disease is a WBC >15,000 or a serum creatinine >1.5 mg/dL (or a >50% rise from baseline). Fulminant disease is pancolitis, toxic megacolon, ileus, or hypotension/shock attributable to the infection. A recurrence is resolution of symptoms followed by a fresh episode within 2–8 weeks. The recurrence risk factors worth naming on the spot are age ≥65, prior severe disease, a second episode within 6 months, compromised immunity (IBD, transplant, severe immunosuppression), and the hypervirulent ribotypes 027/078/244.
| Category | Defining features |
|---|---|
| Severe | WBC >15,000/µL, or serum creatinine >1.5 mg/dL (or >50% above baseline) |
| Fulminant | Pancolitis and/or toxic megacolon, ileus, or hypotension/shock from the infection |
| Recurrence | Symptom resolution, then a repeat episode within 2–8 weeks |
On treatment, the older reflex to metronidazole has given way: current first-line therapy for an initial episode is fidaxomicin or oral vancomycin. For a recurrence, the two evidence-based routes are a tapered-and-pulsed vancomycin course or fidaxomicin — and in the head-to-head framing presented, tapered-pulsed vancomycin produced the highest response rate, reaching statistical significance versus standard vancomycin. Fulminant disease is the tier that mandates escalation of therapy and urgent surgical evaluation.
◆Which antimicrobials are contraindicated in myasthenia gravis — and what treats chronic bacterial prostatitis?
Three antibiotic classes can unmask or worsen myasthenia gravis by impairing neuromuscular transmission and are therefore contraindicated (or used only with great caution): aminoglycosides, fluoroquinolones, and macrolides. The clean crossover to remember is that those same fluoroquinolones are the oral treatment of choice for chronic bacterial prostatitis when the organism is susceptible, precisely because they achieve the prostatic penetration that most β-lactams cannot. The drug that rescues the prostate is the drug you must not reach for in the myasthenic — that single overlap is the pearl a vignette will exploit.
◆Is there finally an oral drug for gonorrhoea as ceftriaxone resistance climbs?
Gepotidacin is a first-in-class oral triazaacenaphthylene bacterial topoisomerase inhibitor — a novel mechanism, which is the point, because it sidesteps the resistance eroding existing agents. In EAGLE-1, a phase 3, randomised, open-label, non-inferiority, multicentre study, oral gepotidacin was compared with the standard of ceftriaxone plus azithromycin for uncomplicated urogenital gonorrhoea, with culture-confirmed eradication of N. gonorrhoeae at test-of-cure (days 4–8) as the primary endpoint. Gepotidacin achieved eradication in 92.6% (187 of 202) versus 91.2% (186 of 204) for ceftriaxone plus azithromycin — meeting non-inferiority and putting a genuinely oral option on the table for a pathogen whose injectable last line is under pressure.
Enrichment — how the IDSA 2024 AMR guidance actually treats AmpC
Reviewer addition (beyond the talk). Grounded in current guidance + recent data.
The talk named the organisms; the IDSA 2024 guidance on antimicrobial-resistant gram-negative infections supplies the algorithm. For infections due to organisms at moderate-to-high risk of clinically significant AmpC production — E. cloacae, K. aerogenes, and C. freundii — the guidance prefers cefepime over a carbapenem when the cefepime MIC is in the susceptible range (≤2 µg/mL), a deliberately carbapenem-sparing choice; it reserves a carbapenem for the critically ill or when the cefepime MIC is susceptible-dose-dependent (4–8 µg/mL), a value that can also flag a co-produced ESBL. It recommends against third-generation cephalosporins even when the isolate tests susceptible, because of on-treatment de-repression, and it does not endorse piperacillin-tazobactam for these infections; for the less-studied organisms (S. marcescens, M. morganii, Providencia spp.), susceptibility testing may guide therapy. Observational bloodstream-infection cohorts support the approach: comparisons of high-dose cefepime with carbapenems for AmpC-risk Enterobacterales bacteraemia found broadly comparable outcomes when isolates were cefepime-susceptible, while a companion analysis cautioned that AmpC mutation and emergence remain real — reinforcing source control and follow-up cultures rather than reflexive de-escalation.
Sources: IDSA 2024 Guidance on the Treatment of Antimicrobial-Resistant Gram-Negative Infections (Clin Infect Dis 2024; doi:10.1093/cid/ciae403; PMID 39108079); High-dose Cefepime vs Carbapenems for Bacteremia Caused by Enterobacterales With Moderate-to-High Risk of AmpC Production (Open Forum Infect Dis 2023; doi:10.1093/ofid/ofad034; PMID 36968970); Mutation Rate of AmpC β-Lactamase–Producing Enterobacterales — A Word of Caution (Clin Infect Dis 2024; doi:10.1093/cid/ciae160; PMID 38527853); ACG Clinical Guidelines: Clostridioides difficile Infections (Am J Gastroenterol 2021; doi:10.14309/ajg.0000000000001278); Michigan Medicine CPG, Treatment of Clostridioides difficile in Adults (2025); Fidaxomicin vs Vancomycin for C. difficile Infection (N Engl J Med 2011; doi:10.1056/NEJMoa0910812; PMID 21288078); EAGLE-1, oral gepotidacin vs ceftriaxone plus azithromycin for uncomplicated urogenital gonorrhoea (Lancet 2025).