May 2009

Document Type


Degree Name



Dept. of Public Health and Preventive Medicine


Oregon Health & Science University


Clostridium difficile associated disease (CDAD) consists of severe diarrhea, fever, lower abdominal pain, anorexia, nausea, malaise and leukocytosis. Over the last few years, increasing incidence and severity of Clostridium difficile infections have been documented in hospitalized settings. The bone marrow transplant (BMT) population is one assumed to be at increased risk for Clostridium difficile infection due to the high levels of immunosuppression, in addition to prolonged and frequent hospitalizations. We undertook this study to see if incidence of CDAD has increased in our high risk BMT population and identify potential risk factors that may be modified to prevent disease in the future. The objectives of the current study are to analyze clinical data collected from OHSU’s BMT population to 1) describe the burden and outcomes of CDAD in this BMT population; 2) to evaluate patient and epidemiologic factors associated with CDAD. Methods Using existing data from electronic medical records and various databases at Oregon Health and Science University from 2002-2008, we identified CDAD cases among BMT recipients along with controls matched for transplant year and time to infection. We used conditional logistic regression to identify risk factors for developing Clostridium difficile infection in the first year following bone marrow transplantation. Univariate and multivariate analyses were completed on subsets of the data, including early infections (< 40 days after transplantation), late infections (> 40 days after transplantation) and allogeneic transplant recipients. Results Cases and controls were similar with regard to age, gender, race and ethnicity. The study population was predominantly white (96.1% cases, 98.0% controls). The median age was 53.3 years in controls and 54.6 years in cases, p=0.82. Cases included a greater number of allogeneic transplant recipients (73.5%) compared to the control group (50.0%). Approximately 14.7% of patients undergoing BMT experienced at least one episode of CDAD. A steady increase in the proportion of patients with CDAD was observed between 2002 and 2007. Donor relation was the most prominent risk factor identified in multivariate analysis. Forty nine percent of cases (50) had unrelated donors compared to 22% (22) of the controls, matched OR 4.30, 95% confidence interval 2.04-9.07. There was no statistically significant difference between those with related donors and autologous transplant recipients (p=0.21). After controlling for donor relation, prednisone use and glycopeptide use, patients had an estimated 8.5% increased odds of CDAD for every additional day hospitalized (95% confidence interval: 1.6% increase to 15.8% increase, p=0.015). Glycopeptide exposure in the 30 days preceding the index date occurred in 60% (61) of cases and only 30% of controls (31), matched OR 3.82, 95% confident interval 1.357-10.771, p=0.01. The proportion of patients with active GI GVHD was much higher in the cases (20%, 20 patients) compared to the control group (2%, 2 patients). The odds of exposure to GI GVHD in cases is estimated to be 19 times the odds of exposure to GI GVHD in the controls (95% CI 3-789, p<0.0001). In addition, a larger proportion (85%) of autologous transplant recipients experienced CDAD early after transplantation as opposed to late. Ninety two percent of late infections occurred in allogeneic transplant recipients. Conclusion The increasing trend in proportion of patients with CDAD after transplantation is an important one. Notable increases were observed between 2003-2004 and 2004-2005, which may be due to more sensitive methods of detection and the emergence of the NAP 1 strain in Oregon as opposed to increases in patient level risk factors. The low power of this study limited the number of variables that could be examined using multivariate analysis. Many of the identified variables, specifically GVHD, use of immunosuppressive agents and length of hospitalization, are also strongly linked to transplant type (autologous vs. allogeneic). Transplant type, or the complications linked to transplant type, may be the underlying factors driving CDAD differences in this population. Utilizing a population of patients receiving only allogeneic transplants may lead to larger cell counts for some of these variables and the ability to perform the additional analyses required to identify pertinent associations. Differing risk factors between early and late infections indicate that time from transplantation is an important factor to consider in future studies. Late infections had a higher proportion of allogeneic transplant recipients (91.5% of cases) whereas early infections showed little difference between transplant types. In addition, GI GVHD diagnosis around the time of CDAD diagnosis is important to consider in this population. Since grade 2 GI GVHD and CDAD have similar symptoms, it is reasonable to assume that GVHD may be the underlying cause of persistent CDAD symptoms in this population.




School of Medicine



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.