Surgical Antibiotic Prophylaxis and Clostridium difficile Toxin Positivity

Daniel Kreisel, Thomas G. Savel, Alan L. Silver, John D. Cunningham

Research output: Contribution to journalArticlepeer-review

51 Scopus citations

Abstract

To examine a possible relationship between prophylactic antibiotic therapy (PAT) and the development of Clostridium difficile toxin (CDT) positivity. Retrospective case-control study. Tertiary care medical center in New York, NY. A total of 357 patients, admitted from November 1992 to April 1994, with positive CDT assays. Of these, 23 patients (6%) received only PAT for elective surgical procedures. Thirty-nine patients were matched as controls for age, sex, and surgical procedure. Both CDT positivity and inappropriate use of PAT. Appropriate PAT was used in 26 (42%) of 62 patients (17% cases, 56% controls). The Mantel-Haenszel estimator for the summary odds ratio for the development of CDT positivity from inappropriate use of PAT was 5.1 (95% confidence interval, 1.10 to 23.64). Mean duration between the operation and the final antibiotic dose was significantly longer in the CDT-positive group compared with the control group (3.1 vs 1.7 days, P<.05). The length of hospital stay was significantly longer in the CDT-positive group compared with the control group (16.5 vs 10.2 days, P<.05). The prolonged use of PAT in elective surgical cases increases the risk of developing CDT positivity. The appropriate use of PAT could significantly reduce health costs by eliminating unnecessary doses of antibiotics, by decreasing the rate of CDT positivity, and by shortening the hospital stay. Restrictive policies may need to be implemented to prevent further antibiotic misuse.

Original languageEnglish
Pages (from-to)989-993
Number of pages5
JournalArchives of Surgery
Volume130
Issue number9
DOIs
StatePublished - Sep 1995
Externally publishedYes

Fingerprint Dive into the research topics of 'Surgical Antibiotic Prophylaxis and Clostridium difficile Toxin Positivity'. Together they form a unique fingerprint.

Cite this