LAUSR

particularly important for low prevalence conditions, such as CMV colitis, which are subject to substantial case misclassification if the diagnostic codes are not accurate. We recently demonstrated that an accurate diagnosis of CMV reactivation in inflammatory bowel disease requires a tissue diagnosis, and is dependent on biopsy number, location, and the type of histologic test.2 The sensitivity of whole blood polymerase chain reaction was only 31% (95% confidence interval, 16-46) in patients with confirmed CMV reactivation by immunohistochemistry. Similarly, conventional histopathology by hematoxylin and eosin-stained tissue has poor sensitivity for detecting CMV reactivation (25%; 95% confidence interval, 15-36). In the current study, only 37% of patients without CMV underwent an endoscopic examination, and it is unknown if any of the individuals who underwent an endoscopic examination were adequately assessed for CMV. As a result, the prevalence of CMV in this population cannot be accurately ascertained, nor can the CMV negative group be regarded as an appropriate control. A second limitation of this study was the inability to assess for the use of antiviral therapy or the degree of viral burden. We recently demonstrated that antiviral therapy results in an 80% reduction in colectomy risk, among patients with corticosteroid-refractory UC.3 We also demonstrated that antiviral therapy seems to be most effective in patients with a high burden of CMV.4 Similarly, others have demonstrated that patients with a high-viral burden are more likely to be resistant to multiple immunosuppressive agents,5 suggesting that the viral burden may impact pathogenicity. On the basis of these studies and others, a new paradigm is emerging which suggests that the pathogenicity of CMV lies upon a continuum, dependent on the degree of viral burden. When low levels of virus are present, the virus likely exists as an innocent bystander, and when high levels are present the virus likely contributes a pathogenic role. Unfortunately, the current study was not able to assess these factors. Therefore, conclusions regarding the impact of CMV cannot be drawn from this work. In conclusion, we would like to stress that accurate case ascertainment through appropriate diagnostic testing, and an understanding of the degree of viral burden along with the use of antiviral therapy are crucial when assessing the impact of CMV in inflammatory bowel disease.