A blog by Lawrence F Muscarella, PhD:
1. Fact? According to the Centers for Disease Control and Prevention’s (CDC) Morbidity and Mortality Weekly Report (MMWR; click here to read the CDC’s article) published in March 4, 2011: “Decreases in CLABSIs have been attributed to various factors, including increased financial and leadership support for CLABSI prevention, improved education and engagement of clinicians in prevention efforts, packaging of prevention recommendations into practice bundles, increased data monitoring and feedback on progress, improvement of the safety culture in health-care, and local and statewide collaborative prevention efforts.”(25)
Response: These factors reportedly have contributed to measured reductions in CLABSIs, but this list is incomplete. Other considerations, too—namely, biases and confounding factors, including measurement bias, feedback bias, publication bias, confounding bias, confirmatory bias, and changes in clinical behaviors—as much as these factors listed by the CDC, may be responsible for measured decreases in the number of CLABSIs in ICUs.
To read Dr. Muscarella’s detailed review of this CDC report, click here.
Moreover, discussed in another blog by Dr. Muscarella — click here — although feedback among staff about the intent, progress and effectiveness of an intervention is often lauded as a benefit of the study, such dialogue can create feedback bias, which can introduce error into the study’s results.(27) Indeed, like a blinded drug study, limiting the staff’s knowledge of the intervention and its intent is important to the study’s validity.
2. Fact? The infection data published by the CDC in its report about CLABSIs featured in the March 1, 2011, issue of MMWR, like most published rates of CLABSIs, are sound and have been independently validated for quality, accuracy and completeness.
Response: The majority of all reported HAI data have not been validated, including the CLABSI data used by the CDC in this report in MMWR. This is a potentially problematic finding, as independent audits of medical records have demonstrated under-reporting of the true incidence of CLABSIs (refer to my more detailed article on this topic by clicking here).(32,37)
Several biases and factors, including confounding, confirmatory, sampling, and feedback biases, may cause not only published CLABSI rates to under-report the true incidence of infection, but also prospective cohort studies (and retrospective comparisons of infection data) to over-exaggerate the effectiveness of an evaluated intervention and to misattribute to this intervention observed reductions in the CLABSI caused instead by one or more confounding factors (e.g., changes in behavior, such as reductions in the sensitivity of the surveillance methods used to detect, record and report CLABSIs; or, the more aggressive use of antibiotics).
Click here to read Dr. Muscarella’s blog “Factors that Might Cause an Intervention’s Effectiveness for Preventing CLABSIs to be (Inadvertently) Exaggerated.”
3. Fact? In 2009 the CDC published a study in the Journal of the American Medical Association (JAMA) that examined trends in the incidence of CLABSIs caused by methicillin-resistant Staphylococcus aureus, or “MRSA.”(42) Using for its analysis infection data that hospitals reported to the CDC between 1997 and 2007, this study reported that the incidence of CLABSIs due to MRSA decreased by at least 50% since 2001 in the 6 most common types of adult ICUs. The CDC concludes in this report that this finding “means that the risk of primary MRSA bloodstream infections among patients with central line in these ICUs has substantially decreased in recent years.”
Response: This conclusion may not be valid. The infection data used by the CDC for this analysis were self-reported and were not validated. Identifying the under-reporting of infection rates, audits have demonstrated that reported data suggesting reductions in infections do not necessarily reflect (or “mean”) actual infection reductions.(32,37) Due to a number of factors (that are discussed in more detail in another blog — click here), the CDC’s conclusions that the incidence of CLABSIs due to MRSA in these adult ICUs decreased by at least 50% may be in error, and the possibility cannot be ruled out that the risk of these infections actually increased in recent years—the CDC’s conclusions notwithstanding.
A number of biases and uncontrolled confounding factors—such as reductions in the sensitivity of surveillance methods used since 2001 to measure, record, and report infections—could result in significant under-reporting of CLABSIs due to MRSA in these 6 types of adult ICUs.
References: Click here.
Blog by: Lawrence F Muscarella PhD posted on 12-11-2011; updated 8-19-2013.