Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
In Reply We appreciate the concerns raised by these letters. Dr Fang points out that toxin immunoassays can be negative in patients with clinically significant Clostridium difficile infection (CDI), including rare patients with fulminant disease. We agree. However, we believe toxin immunoassays miss fewer CDI cases than reported in the literature, and the insensitivity of toxin tests has been exaggerated by the assumption that all patients with diarrhea and C difficile have CDI. For example, the toxin-negative fulminant CDI cases in the Pittsburgh study1 cited by Dr Fang represented just 0.3% of CDI cases over 12 years (6 of 2334). The Swiss study2 referenced by Dr Fang assumed all patients with toxin immunoassay-negative (Tox−) results and C difficile-positive diarrhea had CDI even though most had mild symptoms; 20% were not treated, and other causes of diarrhea were not excluded. Our assessment that most patients who are Tox− and polymerase chain reaction (PCR)-positive (Tox−/PCR+) do not have CDI agrees with previous studies.3,4 Dr Fang also suggests that our study was inadequate to conclude that patients who are Tox−/PCR+ do not benefit from treatment. We agree that a placebo-controlled randomized clinical trial would be helpful to determine the effect of treatment in patients with mild symptoms. Regarding the concern that inclusion of patients without diarrhea confounds our results, we note that our diarrhea outcome analyses compared the proportion of patients with symptoms. Finally, we agree with Dr Fang about the potential epidemiologic utility of detecting toxigenic C difficile carriage. However, our project was designed to determine the natural history and need for treatment of patients who were Tox−/PCR+, not assess the infection control benefit from detection of C difficile carriers.
Polage CR, Solnick JV, Cohen SH. Toxin Immunoassays and Clostridium difficile Infection—Reply. JAMA Intern Med. 2016;176(3):414-415. doi:10.1001/jamainternmed.2015.8539