Lymphoedema Evidence Library
Lymphoedema Evidence Library › Topics › Wounds, skin and cellulitis

Wounds, skin and cellulitis

Wounds, skin and cellulitis

21 sources
All 21 sourcesReferences (RIS)Spreadsheet (CSV)

Evidence in one minute

Update 2026-10-08 — downloaded evidence

  • Recurrent cellulitis: penicillin V 250 mg twice daily for 12 months halves recurrence while taken (22% vs 37%, NNT 5); effect stops when stopped; leg lymphoedema, BMI ≥33 and multiple prior episodes predicted a poorer response — treat the oedema as well (PATCH I — double-blind, low risk of bias: high certainty for recurrent cellulitis, moderate in lymphoedema; PATCH II after mostly first episodes inconclusive, low). Thomas 2013 · UK Dermatology Clinical Trials Network’s PATCH Trial Team 2012
  • Hygiene cuts filarial acute attacks by ~two-thirds. Stocks 2015
  • Stewart–Treves angiosarcoma mimics bullous erysipelas — biopsy atypical purple papules. da Costa Vieira RA 2019

Appraisal update 2026-10-09

  • Do not follow the Alavi/Sibbald chapter on drugs: it calls warfarin “coumarin, a benzophenone”, describes benzopyrones as low-toxicity (coumarin was withdrawn for hepatotoxicity) and advises doxycycline/trimethoprim/erythromycin prophylaxis instead of penicillin V. Its skin-care and compression-by-ABPI sections are usable. Alavi unknown
  • Biopsy the non-healing lymphoedematous ulcer: a basal cell carcinoma was treated with dressings for 3 months (patient account); Stewart–Treves and Kaposi sarcoma (HIV, Durban case) are the other malignancies to exclude. Hasznosy 2012 · Ebrahim 2009 · da Costa Vieira RA 2019
  • Industry pathway sheets (Wounds UK 2014 with BSN, 2022 with medi) are brand-led and unreferenced — use their structure, not their product choices. Single cases (van Zanten 2013, Vivian 2019) show what inelastic compression does to ulcers and oedema but have no comparator. Wounds 2014 · Wounds 2022 · van Zanten M 2013

Gaps

No trials of wound outcomes with vs without lymphoedema treatment are held; Stewart–Treves now covered (da Costa Vieira 2019).

How the evidence points

HighMod.LowV. lowNone
Positive10100
Null or negative00100

Sources on this page with a comparison, by the result of their primary outcome and the GRADE certainty of their main finding (Mod. = Moderate, V. low = Very low, None = not graded). Context sources are not counted. How direction is decided.

All 21 sources

More filters
Loading 21 sources…