Sensitivity of superficial cultures in lower extremity wounds

J Hosp Med. 2010 Sep;5(7):415-20. doi: 10.1002/jhm.688.


Background: Superficial wound cultures are routinely used to guide therapy, despite a lack of clear supporting evidence.

Purpose: To conduct a systematic review of the correlation between superficial wound cultures and the etiology of skin and soft tissue infections.

Data sources: Medline, EMBASE, CINAHL, Scopus.

Study selection: Articles published between January 1960 and August 2009 involving superficial wound cultures and deeper comparison cultures.

Data extraction: Two reviewers independently searched for abstracted information pertaining to the microbiology of lower extremity wounds sufficient to calculate the sensitivity and specificity of superficial wound cultures versus comparison cultures.

Data synthesis: Data pooled using a random-effects meta-analysis model.

Results: Of 9032 unique citations, 8 studies met all inclusion criteria. Inter-rater reliability was substantial (Kappa = 0.78). Pooled test sensitivity for superficial wound swabs was 49% (95% confidence interval [CI], 37-61%], and specificity was 62% (95% CI, 51-74%). The pooled positive and negative likelihood ratios (LRs) were 1.1 (95% CI, 0.71-1.5) and 0.67 (95% CI, 0.52-0.82). The median number of isolates for surface cultures (2.7, interquartile range [IQR] 1.8-3.2) was not significantly different than that for comparison cultures, (2.2, IQR 1.7-2.9) (P = 0.75).

Conclusion: Few studies show a strong relationship between superficial wound swabs and deep tissue cultures, and the current data demonstrate poor overall sensitivity and specificity. The positive and negative LRs were found to provide minimal utility in influencing pretest probabilities. Results of this analysis show that wound cultures should not be used in lieu of local antibiograms to guide initial antibiotic therapies.

Publication types

  • Meta-Analysis
  • Review
  • Systematic Review

MeSH terms

  • Colony Count, Microbial
  • Confidence Intervals
  • Humans
  • Leg Ulcer / microbiology*
  • Lower Extremity / microbiology*
  • Predictive Value of Tests
  • Reproducibility of Results
  • Sensitivity and Specificity
  • Statistics as Topic
  • Statistics, Nonparametric
  • Wound Healing*