Personal study digest for a new ID fellow. Recommendations are from the IDSA 2026 AMR Gram-Negative Treatment Guidance (DOI: 10.1093/cid/ciag481). The "What's changed since March 2026" section is reviewer synthesis of newer evidence, each claim cited. Not a substitute for the full guidance document.
Match the antibiotic to the resistance mechanism, not the organism name: ESBL → carbapenems (invasive) or FQ/TMP-SMX (UTI); AmpC → cefepime; KPC → CZA/MVB/I-R; NDM → ATM-AVI/cefiderocol; OXA-48 → CZA; DTR-PsA → C-T (pneumonia) or CZA/I-R; CRAB → SUL-DUR + carbapenem; Steno → cefiderocol monotherapy.
How to use this guidance
- This guidance covers 6 organism groups: ESBL-E, AmpC-E, CRE (by carbapenemase), DTR P. aeruginosa, CRAB, and S. maltophilia.
- It answers question-by-question for each — preferred vs alternative agents, by site of infection (uUTI, cUTI, non-UTI invasive).
- Phenotype first: obtain AST; use CLSI breakpoints (Table 2 in guidance); send reflex carbapenemase testing for all CRE.
- Extended-infusion dosing for all β-lactams capable of EI (Table 1 in guidance) — this is cross-cutting.
- Reserve newer agents (CZA, C-T, I-R, MVB, cefiderocol, SUL-DUR) for documented resistant organisms; empiric CRO agents waste them.
▶ Table 1 — Suggested adult dosing (normal renal/hepatic function)
Dosing per IDSA 2026 AMR Guidance Table 1. Some doses differ from FDA labeling. See guidance for renal adjustment, pediatric dosing, and CNS infection dosing.
| Agent | Dosing |
|---|
| Amikacin | uUTI: 15 mg/kg IV × 1. cUTI: 15 mg/kg IV once, then PK-guided. |
| Ampicillin-sulbactam | Sulbactam 9 g/day: amp-sulb 9 g IV q8h over 4 h, OR 27 g CI over 24 h. |
| Aztreonam-avibactam | Load: 2.67 g (ATM 2 g / AVI 0.67 g) IV over 3 h, then 2 g (1.5/0.5) IV q6h over 3 h. |
| Cefepime | uUTI: 1 g IV q8h over 30 min. Other: load 2 g over 30 min, then 2 g IV q8h EI over 3 h. |
| Cefepime-enmetazobactam | 2.5 g (2/0.5) IV q8h over 2 h. CrCL ≥130: infuse over 4 h. |
| Cefiderocol | 2 g IV q8h over 3 h. CrCL ≥120: 2 g IV q6h over 3 h. |
| Ceftazidime-avibactam | 2.5 g (2/0.5) IV q8h over 3 h. |
| CZA + aztreonam | CZA 2.5 g IV q8h over 3 h + ATM 2 g IV q8h over 3 h (simultaneous Y-site). |
| Ceftolozane-tazobactam | uUTI: 1.5 g (1/0.5) IV q8h over 1 h. Other: 3 g (2/1) IV q8h over 3 h. |
| Ciprofloxacin | uUTI: 400 mg IV q12h or 500 mg PO q12h. Other: 400 mg IV q8h or 750 mg PO q12h. |
| Colistin / Polymyxin B | Per international consensus guidelines (Tsuji et al., Pharmacotherapy 2019). |
| Eravacycline | 1 mg/kg IV q12h. |
| Ertapenem | 1 g IV q24h over 30 min. |
| Fosfomycin | uUTI: 3 g PO × 1. cUTI: 6 g IV q8h over 1 h. |
| Gentamicin | uUTI: 5 mg/kg IV × 1. cUTI: 7 mg/kg IV once, then PK-guided. |
| Gepotidacin | uUTI: 1.5 g PO q12h. |
| Imipenem-cilastatin | uUTI: 500 mg IV q6h over 30 min. Other: 500 mg IV q6h EI over 3 h, or 1 g IV q8h over 1 h. CrCL ≥90: 1 g IV q6h over 1 h. |
| Imipenem-cilastatin-relebactam | 1.25 g (500/500/250) IV q6h over 30 min. |
| Levofloxacin | 750 mg IV/PO q24h. |
| Meropenem | uUTI: 1 g IV q8h over 30 min. Other: 1–2 g IV q8h EI over 3 h. 2 g preferred for obesity, CNS, ECMO, or CrCL ≥130. |
| Meropenem-vaborbactam | 4 g (2/2) IV q8h over 3 h. |
| Minocycline | 200 mg IV/PO q12h. |
| Nitrofurantoin | Macrobid: 100 mg PO q12h. Suspension: 50 mg PO q6h. |
| Pivmecillinam | uUTI: 370 mg PO q8h. |
| Plazomicin | uUTI: 15 mg/kg IV × 1. cUTI: 15 mg/kg IV once, then PK-guided. |
| Sulbactam-durlobactam | 2 g (1/1) IV q6h over 3 h. CrCL ≥130: 2 g IV q4h over 3 h. |
| Sulopenem-probenecid | 500/500 mg PO q12h. |
| Tigecycline | Load 200 mg IV, then 100 mg IV q12h. |
| Tobramycin | uUTI: 5 mg/kg IV × 1. cUTI: 7 mg/kg IV once, then PK-guided. |
| TMP-SMX | uUTI: TMP 160 mg IV/PO q12h. Other: TMP 8–15 mg/kg/day IV/PO divided q8–12h. |
Empiric & definitive therapy — ESBL-E
| Scenario | Drug / dose / route | Notes |
|---|
| Uncomplicated UTI |
Nitrofurantoin 100 mg PO BID × 5 d OR TMP-SMX DS PO BID × 3 d (preferred). Alt: single-dose aminoglycoside, FQ (ciprofloxacin 250 mg PO BID or levofloxacin 750 mg PO × 1). |
Avoid oral cephalosporins. Pip-tazo acceptable as empiric pending AST. |
| Complicated UTI / pyelonephritis |
FQ or TMP-SMX (preferred if susceptible, stable). Unstable: ertapenem 1 g IV daily or meropenem 1–2 g IV q8h EI. |
Cefepime-enmetazobactam 2 g/0.5 g IV q8h EI is an alternative (non-carbapenem). Avoid ceftriaxone. |
| Non-UTI invasive (e.g., bacteremia, IAI, pneumonia) |
Ertapenem 1 g IV daily (preferred) or meropenem/imipenem EI. Oral step-down: FQ or TMP-SMX if source controlled + susceptible. |
Pip-tazo NOT suggested for invasive ESBL-E (MERINO trial: 3× higher mortality). Cefepime not suggested. Cephamycins (cefoxitin/cefotetan) not suggested. |
Empiric & definitive therapy — AmpC-E
| Scenario | Drug / dose / route | Notes |
|---|
| Risk stratification |
(Not a regimen — a framework) |
Moderate risk: E. cloacae, K. aerogenes, C. freundii, H. alvei. Lower risk: M. morganii, Providencia, S. marcescens. Avoid SPACE/SPICE mnemonics — they misclassify. |
| Non-UTI invasive (moderate-risk species) |
Cefepime 2 g IV q8h EI (preferred, if MIC ≤8 μg/mL). Alt: carbapenems. |
~20% ceftriaxone resistance emergence for invasive infections — avoid. Pip-tazo not suggested. C-T not suggested for AmpC-E. |
| Lower-risk species |
Standard AST-guided therapy; ceftriaxone acceptable if susceptible. |
Non-β-lactam agents (FQ, TMP-SMX, aminoglycosides) acceptable per AST. |
Empiric & definitive therapy — CRE by carbapenemase
KPC-producing Enterobacterales
| Scenario | Drug / dose / route | Notes |
|---|
| cUTI |
FQ, TMP-SMX, or CZA/MVB/I-R (all preferred if susceptible). |
uUTI: aminoglycoside, FQ, nitrofurantoin, TMP-SMX. |
| Non-UTI invasive |
CZA 2.5 g IV q8h EI, OR meropenem-vaborbactam (MVB) 4 g IV q8h EI, OR imipenem-relebactam (I-R) 1.25 g IV q6h. Panel slightly favors MVB > CZA > I-R. |
CZA resistance emergence ~10% vs <3% for MVB/I-R. Cefiderocol is an alternative. No routine combination therapy. |
NDM-producing Enterobacterales
| Scenario | Drug / dose / route | Notes |
|---|
| Non-UTI invasive |
Aztreonam-avibactam (ATM-AVI) preferred. Cefiderocol preferred. Alt: CZA + aztreonam (commercially available now). |
CZA alone NOT active against NDM. PBP3 insertions + CMY → ~30% mid-Atlantic NDM E. coli resistant to ATM-AVI. |
OXA-48-like-producing Enterobacterales
| Scenario | Drug / dose / route | Notes |
|---|
| Non-UTI invasive |
CZA 2.5 g IV q8h EI (preferred). Alt: ATM-AVI, cefiderocol. |
MVB and I-R are NOT suggested for OXA-48-like producers. EI meropenem NOT reliable — MICs often near breakpoint. |
Empiric & definitive therapy — DTR P. aeruginosa
| Scenario | Drug / dose / route | Notes |
|---|
| cUTI |
CZA, C-T, I-R, or cefiderocol (all equally preferred). |
For MDR (not yet DTR) PsA, prefer traditional β-lactams (cefepime, pip-tazo) with EI dosing. |
| Non-UTI invasive (non-pneumonia) |
CZA 2.5 g IV q8h EI, C-T 3 g IV q8h EI, OR I-R 1.25 g IV q6h (all preferred). Cefiderocol alternative. |
~20% treatment-emergent resistance for all newer agents. Cefiderocol <10%. |
| Pneumonia (HABP/VABP) |
C-T 3 g IV q8h EI (preferred over CZA/I-R for pneumonia). Cefiderocol alternative. |
C-T ELF penetration ~50% vs ~30% for CZA. C-T NOT for KPC-producing PsA. MBL PsA → cefiderocol only. |
Empiric & definitive therapy — CRAB
| Scenario | Drug / dose / route | Notes |
|---|
| All sites (preferred) |
SUL-DUR (sulbactam-durlobactam) 1 g/1 g IV q6h + a carbapenem (preferred). >95% US susceptibility. |
ATTACK trial: 81% vs 68% 28-d all-cause mortality (vs colistin + imipenem). This is the only agent with a dedicated CRAB RCT. |
| Bridging (SUL-DUR unavailable) |
High-dose ampicillin-sulbactam (sulbactam 9 g/day IV) + a second active agent. |
Bridging only — inferior to SUL-DUR. |
| Alternatives (in combination) |
Cefiderocol + second agent (alternative; concerning CREDIBLE-CR data: 51% vs 82% survival). Minocycline + second agent (<50% susceptible at 2025 CLSI breakpoint ≤1). Polymyxin B + second agent. |
All alternatives are used IN COMBINATION, never monotherapy. Nebulized antibiotics not routinely suggested. |
Empiric & definitive therapy — S. maltophilia
| Scenario | Drug / dose / route | Notes |
|---|
| All sites (preferred) |
Cefiderocol 2 g IV q8h EI — MONOTHERAPY (preferred). ~100% susceptibility. |
Rabbit pneumonia model: 88% vs 25% survival (vs TMP-SMX). Only agent preferred as monotherapy. |
| Alternatives (in combination) |
ATM-AVI + second agent (>90% susceptible). Levofloxacin + second agent (~90% susceptible; ~20% resistance emergence). Minocycline + second agent (~90% susceptible). TMP-SMX + second agent (>90% susceptible; bacteriostatic only). |
All alternatives require combination. Ceftazidime NOT suggested. Tigecycline inferior to minocycline (higher protein binding). |
Duration & stopping
- No fixed duration specified in this guidance — defer to site-of-infection guidelines (uUTI 3–5 d, cUTI 5–7 d, bacteremia 7–14 d, pneumonia 7 d).
- Source control is paramount for IAI, abscess, prosthetic material.
- Oral step-down acceptable for ESBL-E and AmpC-E when source controlled + susceptible oral agent available + clinically improving.
- Repeat blood cultures for bacteremia to document clearance.
Key decisions a fellow owns
- Send carbapenemase testing for every CRE isolate — the therapy depends on KPC vs NDM vs OXA-48, not just "CRE."
- Check MICs, not just S/I/R — cefepime for AmpC-E requires MIC ≤8; meropenem for OXA-48 is unreliable near breakpoint.
- Extended-infusion dosing for meropenem, cefepime, pip-tazo, CZA, C-T, cefiderocol — standard infusions leave you below PK/PD targets.
- Reserve newer agents for confirmed resistance — empiric CZA for a UTI that might be ESBL wastes stewardship capital and breeds resistance.
- De-escalation: narrow once AST returns. A patient started on meropenem for empiric coverage who grows ESBL-E susceptible to TMP-SMX should switch.
- C-T for PsA pneumonia: know when to reach for C-T over CZA — ELF penetration drives this choice.
- Combination therapy is NOT routine for any organism group in this guidance, with the exception of CRAB and Steno alternatives (where a second agent is always added).
What's changed since March 2026
Reviewer synthesis — not the guidance document.
- Cefepime-zidebactam (Zaynich) FDA approved May 29, 2026 — novel β-lactam/β-lactam enhancer for cUTI (E. coli, K. pneumoniae, P. mirabilis, E. cloacae complex, P. aeruginosa). Phase 3 ENHANCE-1 trial. Adds a non-carbapenem IV option for resistant cUTI not covered in the March 2026 guidance. FDA label NDA 220787
- Tebipenem pivoxil (Utebzi) FDA approved June 17, 2026 — first oral carbapenem, approved for cUTI/pyelonephritis. PIVOT-PO trial (non-inferior to ertapenem IV). Changes ESBL-E cUTI landscape: an oral carbapenem option now exists for patients who are stable but need carbapenem-level coverage. FDA; PIVOT-PO trial
- C-T vs CZA for DTR PsA: meta-analysis confirms C-T advantage — 6 observational studies, N=1,161. C-T showed higher clinical cure (aOR 1.82, 95% CI 1.10–2.99) and lower recurrence (aOR 0.46, 95% CI 0.26–0.78) vs CZA. No mortality difference. Supports the guidance's preference for C-T in PsA pneumonia. Gatti et al., CID 2026; PMID 41844194
- Oral step-down for GN bacteremia: meta-analysis supports safety — 13 studies, N=10,000. Lower all-cause mortality with oral step-down (RR 0.39, 95% CI 0.16–0.93), mainly with FQ. Sensitivity analyses (propensity-adjusted) showed no difference — supports safety, not yet superiority. Onorato et al., EJCI 2026; PMID 42593719
- SUL-DUR treatment-emergent resistance reported — first clinical case of CRAB sulbactam-durlobactam MIC increase (8-fold) after 14 days of therapy. Mechanism: AdeFGH efflux pump mutations with 166-fold increase in adeG expression. Isolated case, but signals a surveillance need. Sharara et al., CID 2026; PMID 42442753
- Pipeline: zosurabalpin entering phase 3 for CRAB — Roche's novel tethered macrocyclic peptide (blocks LPS transport) entering pivotal testing (est. ~400 patients). Would be first new mechanism class for CRAB if approved. Roche/CIDRAP 2025
- Still current — the guidance's core architecture (match-to-mechanism, CZA/MVB/I-R for KPC, ATM-AVI/cefiderocol for NDM, CZA for OXA-48, SUL-DUR for CRAB, cefiderocol monotherapy for Steno) remains well-supported 5 months post-publication. No contradictory RCTs or withdrawals.
Anki cards minted this run
- Tebipenem pivoxil (PIVOT-PO) — first oral carbapenem for cUTI (noteId 1787343452823; prior run)
- Cefepime-zidebactam (ENHANCE-1) — novel BL/BLE for resistant cUTI (noteId 1787343452869; prior run)
- Steno cefiderocol monotherapy — preferred agent, ~100% susceptibility (noteId 1787343452898; prior run)
- DTR-PA pneumonia — C-T preferred over CZA (ELF penetration) (noteId 1787343452948; prior run)
All 4 cards from prior run (2026-08-21) confirmed present in deck. No new cards added — cap met.
Sources:
- Tamma PD, Bonomo RA, Heil EL, Justo JA, Satlin MJ, Mathers AJ. IDSA 2026 Guidance on the Treatment of Antimicrobial Resistant Gram-Negative Infections. Clin Infect Dis. 2026. DOI: 10.1093/cid/ciag481. PMID: 42570093.
- Gatti M et al. Comparison of C-T and CZA in MDR/DTR P. aeruginosa: systematic review and meta-analysis. Clin Infect Dis. 2026. DOI: 10.1093/cid/ciag194. PMID: 41844194.
- Onorato L et al. Oral stepdown therapy for GN BSI: systematic review and meta-analysis. Eur J Clin Invest. 2026;56(8):e70244. DOI: 10.1111/eci.70244. PMID: 42593719.
- Sharara SL et al. Treatment-emergent A. baumannii non-susceptibility to sulbactam-durlobactam. Clin Infect Dis. 2026. DOI: 10.1093/cid/ciag436. PMID: 42442753.