Viewpoints in Infectious Diseases: Antibiotic Use and Clostridium difficile

John G. Bartlett, MD


March 05, 2009

Patterns of Antibiotic Use and Risk of Hospital Admission Because of

This Viewpoint takes a look at the link between certain antibiotics and development of C difficile infections, but also cautions about drawing conclusions based on pharmacy database records.

Dial S, Kezouh A, Dascal A, Barkun A, Suissa S.
CMAJ 2008;179:767-772

Article Summary

This study sought to determine the precipitating cause of Clostridium difficile infection (CDI). The investigators' emphasis was on the use or non-use of antibiotics and the specific agents used. The study linked health databases using the Canadian system for pharmacy records for persons over 65 years and the provincial hospital discharge summary database, which was screened for persons admitted with CDI. For each case, investigators randomly selected 10 controls from the study population and matched for time of hospital admission. The pharmacy records were reviewed for antibiotic prescriptions given during the 45-day period before the date of hospitalization.

The screening process found 836 cases of CDI and 442 of these (53%) had no history of antibiotic exposure during the previous 45 day period according to pharmacy records. Antibiotic exposure was associated with a risk for CDI of 10.6. With regard the individual agents, clindamycin was associated with the greatest risk; for tetracycline and trimethoprim-sulfamethoxazole the risk was virtually nil. These results are summarized for the antibiotics in the following table .

Of note, the risks specified in the table represent the relative risk for the specified agent compared with 8,360 controls. Of additional interest is the risk associated with proton pump inhibitors and antacids, which were 1.6 and 1.5, respectively.

The authors conclude that CDI is common without previous antibiotic exposure, thus emphasizing the need to accept this possibility in patients who deny the most commonly recognized risk factor.


I always worry about studies that match databases based on my assumption that many patients keep antibiotics in their cupboards, borrow, or have alternative methods that would escape detection by review of pharmacy records for the 45 days before hospital presentation. Nevertheless, this study makes some important observations:

  • The infection is relatively frequent in outpatients despite emphasis on the role of the hospital;

  • Clindamycin, cephalosporins, and fluoroquinolones continue to be the "big 3";

  • Of importance, this study was done in Quebec where an astonishing number of cases have involved the NAP-1 strain since 2004. This strain is resistant to fluoroquinolones, which may account for their relatively large numbers as inducing agents in this as well as other more recent reports; and

  • Although non-antibiotic use seems to be prominent, it should be noted that these cases clearly do occur and the NAP-1 strain appears to be the most common pathogen in such cases; my skepticism is simply based on the large numbers here because 50% seems high and the possibility of missing some cases with a 45-day review of pharmacy records may obviously miss some of the cases.



Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.