To the Editor:
We welcome the engagement and comments of Dikon on our recently published article describing the relationship between sharps disposal containers and Clostridium difficile infections in acute care hospitals.
1
The reader questioned several aspects of the methods used in the study, including the use of regression analysis, response rates, and sample representativeness that we would like to address.Dikon questioned the scientific credibility of the study based on its use of regression analysis stating that “if other variables are not removed, it is difficult to determine causality from mere coincidence.” However, the use of multivariable regression in fact allowed us to examine the relationship between the exposure and outcome of interest, after controlling for other variables (ie, removing the effect of other variables). In our study, we controlled for variables that we posited to be potential confounders of the relationship between the use of different sharps containers and C difficile rates, including hospital bed size, ownership status, geographic region, and urbanicity. As discussed by Dikon, we did not collect data and control for antibiotic use or the type laboratory testing used for C difficile. We agree that these are known risk factors or predictors of C difficile rates. However, in order to influence the relationship between different types of sharps containers and C difficile rates, a variable should be correlated with both the outcome and exposure. Because antibiotic use or the laboratory testing method is not likely to be correlated with the type of sharps container used by a hospital, it is unlikely to confound the relationship under study.
We agree that the response rate of 30% is a limitation of this study and that there is potential for selection bias. However, as described in the published study, we conducted a comparison of respondents and nonrespondents to assess whether participating and nonparticipating hospitals differed on potential confounders and found no differences in terms of urbanicity, ownership type, and C difficile rates.
In her comments on sample representativeness, Dikon brings up a point that most of the respondents were environmental services personnel and not infection preventionists and that infection preventionists, and not environmental services personnel, are the most knowledgeable personnel regarding infection rates in the hospital. We completely agree with this statement but want to clarify that the survey was used to collect data on the use of sharps disposal containers only; therefore, the survey respondent from each hospital was the person who was most knowledgeable about the type of sharps containers used in their facility. C difficile infection rates were not collected through the survey and instead were ascertained from the Medicare Provider Analysis and Review data set and linked to the survey. In terms of the validity of the survey itself, this survey was developed specifically for this research project in collaboration between the author and a survey research company. The survey did not elicit opinion but asked specific questions about the type and brand of sharps containers used and was piloted prior to launch. In addition, the staff of the survey research company who conducted the telephone survey were blinded as to the overall aim of the study.
As mentioned by Dikon, the funding for this study was provided by BD, a maker of single-use sharps containers, and this funding was fully disclosed at the time of peer review and publication of the study. Although funding was provided by BD, the study was conducted independently, starting from study design through to data acquisition, analysis, and publication. Industry funding does not automatically make the results of a study biased.
Finally, this study found a statistically significant correlation between the use of single-use sharps disposal containers and C difficile rates. We agree with Dikon that there is not enough scientific evidence to prove that reusable sharps containers are a source of C difficile transmission because correlation does not necessarily mean causation. However, we do believe that our data support the need for further research to explore the role of sharps containers as a potential factor in the transmission of pathogens in the hospital setting.
Reference
- Relationship between sharps disposal containers and Clostridium difficile infections in acute care hospitals.Am J Infect Control. 2015; 43: 1081-1085
Article info
Publication history
Published online: February 02, 2016
Footnotes
Conflicts of Interest: Pogorzelska-Maziarz reports that she served as a consultant to Becton, Dickinson and Company on this project.
Identification
Copyright
© 2016 Published by Elsevier Inc. on behalf of Association for Professionals in Infection Control and Epidemiology, Inc.
