DOES DOXYCYCLINE REDUCE THE RISK OF CLOSTRIDIUM DIFFICILE INFECTION IN PATIENTS WITH CELLULITIS?

Monday, October 21, 2013
Key Ballroom Foyer (Hilton Baltimore)
Poster Board # P2-9
Decision Psychology and Shared Decision Making (DEC)
Candidate for the Lee B. Lusted Student Prize Competition

Poonam Mathur, DO, MPH, Doug Leslie, PhD, Amy Welch, MD, MSN, MSc, John J. Zurlo, MD and Cynthia H. Chuang, MD, MSc, Penn State/Milton S. Hershey Medical Center, Hershey, PA
Purpose: A previous study has suggested that using doxycycline for community-acquired pneumonia reduces the likelihood of subsequent Clostridium difficile infection (CDI). It is unknown whether using doxycycline for the outpatient treatment of cellulitis reduces the risk of subsequent CDI compared with other antibiotics.

Method: The MarketScan Database captures clinical utilization and prescription drug claims from private insurance companies. Using MarketScan data, we identified patients 18-65 years who had an outpatient diagnosis for cellulitis in 2009-2010 and were treated with antibiotics. The outcome was CDI within 30 days of treatment for cellulitis, defined as 1) an outpatient visit for diarrhea with concomitant prescription for either oral vancomycin or oral metronidazole, or 2) an inpatient hospitalization with a CDI diagnosis. We modeled the effect of different antibiotics on risk of CDI, controlling for history of inflammatory bowel disease, diabetes, hypertension, coronary artery disease, hospitalization in the past year, and previous CDI in the past year.

Result: There were 574,274 adult patients with a diagnosis of cellulitis in 2009-2010 who filled an antibiotic prescription within 7 days. Trimethoprim (with or without sulfamethoxazole) and cephalosporins were the most commonly used antibiotics, accounting for 34% and 30%, respectively. Doxycycline was used 8.3% of the time. When compared to other antibiotics, doxycycline conferred the same risk for 30-day CDI (adjusted OR 1.10, 95% CI 0.99-1.22).  Antibiotics associated with significantly higher risk of CDI compared with doxycycline included clindamycin and linezolid, by 22% and 55%, respectively. Trimethoprim (with or without sulfamethoxazole) and cephalosporins had a 16% and 17% (respectively) decreased risk of CDI compared with doxycycline.  The greatest predictor of CDI after antibiotic treatment for cellulitis was having a history of CDI in the preceding year, which was associated with a 115-fold increased risk. Other predictors of CDI were female sex, ulcerative colitis, Crohn’s disease, diabetes, hypertension, coronary artery disease, current PPI use, current H2B use, or hospitalization in the past year. Age was not predictive of developing CDI in these analyses.

Conclusion: In this very large national cohort of patients receiving antibiotics for outpatient treatment of cellulitis, doxycycline does not reduce the risk of subsequent CDI compared with other antibiotics.