I was sitting in the audience at a recent nursing conference listening to a friend present her research results. A few minutes into telling the audience about her intervention, I became fascinated by the whispered conversation taking place one row in front of me. First, I was fascinated because they were millennials and had their own unique mannerism for communication and, second, because the thread of their discussion was basically "great idea, but it won't work at my hospital." This fairly familiar sentiment got me thinking about replication research.
In graduate school, I was taught to carefully define and describe my methodology. The mantra being that someone should be able to read my methods section and replicate my study. The methods need to be described in enough detail that another nurse at another hospital could replicate the study and, ideally, end up with the same results. But how often does this ever happen? It has been over a decade since Fahs et al1 called for resurgence in replication research. In that time, little has changed. Far too many studies published with small sample sizes from a single hospital are accepted as fact. The emphasis on new and innovative seems to have trumped the scientific methods of yore.
In an age when nursing research is on the rise, we now have the manpower to conduct replication research. As a scientist, I have had the good fortune to design or participate in the design of dozens of clinical research studies. One or two are simply brilliant, and looking back honestly, a handful of them are absolute trash. Most, however, boil down to good designs with various flaws. I can live with being imperfect. In fact, all studies (not only mine) have flaws. Usually, the honest author discusses the major flaws within their article in the section on limitations. One of the most common limitations I see is "[horizontal ellipsis]small sample size at a single institution[horizontal ellipsis]." This is often followed by "[horizontal ellipsis]supports the need for further research including replication[horizontal ellipsis]."
Hopefully, nobody really thinks that the science is settled after a single study. The methods, and even the instruments used in the study, need to be reexamined under various conditions.2 I often have nurses coming up to me asking how to get involved in research. Replication research is certainly a prime opportunity. The millennials in front of me may have been 100% correct; the intervention might not have worked in their unit. We don't know whether they would get the same results. Unless and until we have replication studies, we can't know.
I grew up being taught that the 3 Rs were "Reading, wRiting, and aRithmetic." There are a host of interventions, projects, and ideas that have been nearly universally adopted without adequate replication research. To advance the science of healthcare, perhaps, we should focus on the 3 Rs of "Replicating Research Results" a bit more. Just in case you're wondering, yes, the Journal of Neuroscience Nursing will publish well-written replication research results.
The Editor declares no conflicts of interest.
References