by Andreas Voss and Eli Perencevich
This response to the recent JAMA Viewpoint by Morgan, Wenzel and Bearman is cross-posted on the “Reflections” and “Controversies” blogs. Both blogs were recently ranked “top 10” ID blogs by this recent ICHE study, so we might as well collaborate 😉
During the recent ICPIC 2017 and a pre-meeting think tank, the sense and non-sense of RCTs looking at various infection control measures was a major point of discussion during many sessions. Data from well-designed quasi-experimental studies, epidemiological evidence and logic seems to vanish whenever a new RCT is published, even if the results are not applicable to situations that are non-endemic, have higher or lower compliance with the preventive measures in question, or whether the intended measures were actually applied within the intended patient group. Some studies seem to assume that the transmission during the first days of admission are of no consequence. Others assume that given endemicity and a high patient load, the intended measures such as single-room isolation can’t be applied, even if a patient was randomized to receive those measures.
We do know that Morgan, Wenzel and Bearman are very well aware of the fine differences, but we fear that their cited research along with editorials like their viewpoint will be misinterpreted by many within the infection control community and those adjacent to infection control, such as the nurses, clinicians from other specialties, administrators, guideline makers, and legislators.
Our question would be, how many more RCTs do we need to slice and dice needed infection control measures? We believe that for as long as one can surmise an acceptable level of endemicity, deduce an allowable lack of compliance with the intended infection control measure, as well as justify the impossibility of applying the intended measures to all assigned patients, any infection control measure can be proven “unnecessary” in an RCT (or a poorly designed quasi-experimental study), and later on, in a Cochrane review. Is this what we need?
Let’s be clear about one thing. We believe that good studies in our subspecialty are needed and that evidence should be stronger than opinions. At a time when costs are critically important we need to know what to invest in. Infection control therefore needs to be effective and cost-efficient. Still we can’t help but feel that at present the “scientific pendulum” is swinging in the wrong direction.
At present, MRSA and VRE are endemic in many countries, ESBL is everywhere, and we are at a decisive moment before the same is true for CRE. Consequently, now more than ever, it is of up-most importance that we acknowledge the limitations of the existing RCTs as well as include all other available evidence in our evaluation of infection control measures. If we are forced to ignore strong epidemiological evidence supporting transmission in healthcare settings and methods for halting the spread of MDR-pathogens while waiting for the perfect study, are we really providing the safest care for our patients?