Fellow's quick read · Infectious Diseases
Diabetic foot infection
IWGDF/IDSA 2023 · reviewed 2026-06-27
Personal study digest for a new ID fellow. Recommendations are from the IWGDF/IDSA 2023 guideline (Senneville et al., dual-published in Clinical Infectious Diseases and Diabetes/Metabolism Research and Reviews). The “What’s changed since 2023” section is reviewer synthesis of newer evidence, each claim cited. Not a substitute for the full guideline.
In one line
DFI is a clinical diagnosis — you cannot diagnose it from a swab or a culture. Three fellow-owned moves drive the outcome: grade severity with the IWGDF/IDSA scheme, decide whether bone is involved, and match antibiotic spectrum and duration to severity and surgery — not to the wound’s colonizers.◆When to suspect / diagnose
- Diagnose infection clinically: ≥2 local or systemic signs of inflammation (swelling/induration, erythema, tenderness, warmth, purulence) AND no other cause (gout, acute Charcot, trauma, venous stasis). A positive culture without inflammation is colonization, not infection.
- Grade every infected foot with the IWGDF/IDSA scheme: mild (erythema 0.5–2 cm, skin/subcutaneous only) → moderate (erythema ≥2 cm OR deeper than subcutaneous: tendon/muscle/joint/bone) → severe (the above PLUS SIRS ≥2). Append “(O)” when bone is involved.
- Order CRP / ESR / PCT only when the exam is equivocal (Best Practice) — biomarkers support, they don’t make, the diagnosis.
- Suspect osteomyelitis (DFO) in a deep, large, or long-standing ulcer; a diffusely swollen “sausage toe” is osteo until proven otherwise (see below).
◆Workup the fellow drives
- Suspected DFO → start with probe-to-bone + plain X-ray + ESR (or CRP/PCT) (Conditional). PTB sensitivity 0.87 / specificity 0.83 — a positive PTB in a high-risk foot rules in; a negative PTB in a low-risk foot helps rule out.
- MRI when DFO is still in doubt after exam, X-ray, and labs (Strong; Moderate). PET, leukocyte scintigraphy, or SPECT if MRI is unavailable/contraindicated (Conditional).
- Send a tissue specimen (curettage or biopsy), not a superficial swab, for culture (Conditional; Moderate). For DFO, culture bone — not soft tissue — percutaneously or intraoperatively (Conditional; Moderate).
- Conventional culture, not molecular/16S, for first-line pathogen ID (Strong) — molecular methods can’t separate live from dead organisms and drive over-broad therapy.
◆Empiric & definitive therapy
| Scenario | Drug / route | Duration | Strength |
|---|---|---|---|
| Mild, no recent abx (temperate climate) | Gram-positives only — cephalexin or cloxacillin (oral) | 1–2 wk | BPS / Strong |
| Mild + MRSA risk | TMP-SMX, doxycycline, clindamycin, or linezolid | 1–2 wk | BPS |
| Moderate / severe | GPC ± GNR — amox-clav or amp-sulbactam, or 2nd/3rd-gen cephalosporin; pip-tazo or ertapenem if recent abx | 2–4 wk (10 d post-debridement) | Strong |
| + MRSA risk | Add vancomycin, linezolid, or daptomycin | — | BPS |
| Pseudomonas | Don’t cover empirically in temperate climates; add anti-pseudomonal (pip-tazo, ceftazidime, cipro) only if recently isolated, or mod/severe with macerated wound (Asia/N. Africa) | — | BPS |
- No antibiotic class is superior — choose by stewardship (narrowest, shortest, safest, cheapest), then step down to oral by culture. Avoid tigecycline (inferior to ertapenem).
- DFO may need upper-range doses for bone penetration; high-bioavailability oral agents are acceptable. New BL/BLI combos and dalbavancin/oritavancin: insufficient DFI data.
◆Duration & stopping
- Soft tissue: mild 1–2 wk; moderate/severe 2–4 wk (10 d if surgically debrided).
- Bone (DFO): 6 wk if no surgery / dead bone remains · 3 wk if resected with positive bone-margin culture · 2–5 days if fully resected with negative margins (residual soft-tissue only).
- Not improving by 4 weeks of apparently appropriate therapy → re-evaluate and re-image; reconsider the diagnosis rather than simply extending (Strong).
- Judge DFO remission (not “cure”) at a minimum of 6 months after stopping antibiotics (Best Practice).
◆Key decisions a fellow owns
- Do NOT give antibiotics for a clinically uninfected ulcer — they neither promote healing nor prevent infection, and they breed resistance (Best Practice). The most common stewardship miss in the foot.
- Antibiotics alone (no surgery) are acceptable for selected forefoot DFO: no need for I&D, no severe PAD, no exposed bone (Conditional). Otherwise combine bone resection + systemic antibiotics.
- Urgent surgical (± vascular) consult for severe infection, or moderate DFI with gangrene, necrotizing infection, deep abscess, compartment syndrome, or critical ischemia (BPS); aim for early (24–48 h) debridement in moderate/severe DFI (Conditional).
- PAD + infected ulcer/gangrene → urgent surgical AND vascular consult to time drainage/revascularization (BPS).
- Skip the unproven add-ons for treating the infection: G-CSF, topical antiseptics/silver/honey/bacteriophage, NPWT, topical antibiotics, and hyperbaric/topical oxygen are not recommended (Conditional).
What’s changed since 2023
Reviewer synthesis of newer evidence — not the guideline. This is a recent guideline (literature search to mid-2022), so most of it stands; these are the moving pieces.
- Rifampin is still unsettled — the trial that would settle it isn’t out yet. The guideline made no rifampin recommendation. VA INTREPID (6-wk adjunctive rifampin 600 mg/d vs placebo, n=880, primary endpoint amputation-free survival) is built to answer this; no results published as of June 2026. Rifampin is routine for DFO in Europe, rare in the US. Recency-sensitive — confirm whether INTREPID has reported. Bessesen, BMC Infect Dis 2020;20:23 (PMID 31914940); NCT03012529
- Duration keeps shrinking and is becoming bone-margin–tailored. The 3-wk (positive margin) / 6-wk (no surgery) split rests on Gariani’s noninferiority pilot (3 wk ≈ 6 wk post-debridement; remission 84% vs 73%, P=.21). Since: a 2024 post hoc of 2 RCTs found patients with NO residual osteo (negative margins) do well with shorter courses; 2024–25 meta-analyses converge on ≤6 wk; a 4-vs-6-wk nonsurgical RCT is enrolling (results ~2029). Gariani, CID 2021;73:e1539 (PMID 33242083); Lavery, Int Wound J 2024;21:e70072 (PMID 39375181)
- Oral-first for DFO is consolidating (OVIVA-aligned). The guideline already permits high-bioavailability oral step-down; a 2025 narrative review (Senneville group) argues oral ≈ IV for DFO with bone-penetrating agents, cutting length of stay and line complications. Gachet/Senneville, J Clin Med 2025;14:1405 (PMID 40094798)
- “Sausage toe” named as a DFO sign. First case series: ~30% of diabetic toe infections were sausage toes, and MRI showed osteomyelitis in 87.5%. Low threshold to image and treat as forefoot osteo. Yammine, Int Wound J 2024;21:e70002 (PMID 39041186)
- Fluoroquinolones lower on the list. Table 4 still lists FQs, but FDA/MHRA safety actions (aortic aneurysm/dissection, tendon rupture, dysglycemia, neuropathy) argue for reserving them when alternatives exist. FDA/MHRA drug safety communications
- Companion vascular guideline. The intersocietal IWGDF/ESVS/SVS 2023 PAD guideline standardizes PAD work-up (WIfI) and revascularization in the ischemic infected foot — pairs with Recs 18/20. Fitridge, Eur J Vasc Endovasc Surg / J Vasc Surg 2023
◆Anki cards minted this run
- DFO antibiotic duration by surgical / bone-margin status (6 wk · 3 wk · 2–5 d).
- Empiric coverage for mild DFI = gram-positives only; when to add Pseudomonas.
- Antibiotics alone (no surgery) acceptable for select forefoot DFO.
- “Sausage toe” = forefoot osteomyelitis until proven otherwise (Yammine 2024).
- Held — already in deck: IWGDF/IDSA severity classification (moderate & severe DFI); MRI for osteo in a non-healing ulcer.
- Held — under 4-card cap: no antibiotics for clinically uninfected ulcers; rifampin/INTREPID still pending.
Sources: IWGDF/IDSA 2023 (Senneville, Clin Infect Dis 2023, doi:10.1093/cid/ciad527; Diabetes Metab Res Rev 2023;40(3):e3687, doi:10.1002/dmrr.3687); Gariani 3-vs-6 wk (Clin Infect Dis 2021;73:e1539, PMID 33242083); Lavery residual-osteo post hoc (Int Wound J 2024;21:e70072, PMID 39375181); VA INTREPID protocol (BMC Infect Dis 2020;20:23, PMID 31914940; NCT03012529); oral antibiotics for DFO review (J Clin Med 2025;14:1405, PMID 40094798); sausage toe (Int Wound J 2024;21:e70002, PMID 39041186); IWGDF/ESVS/SVS PAD 2023 (Fitridge, Eur J Vasc Endovasc Surg / J Vasc Surg 2023).