Doctors work in the real world so although in principle all medical decisions should be made on the basis of empirical evidence, some interventions—such as deciding whether to admit a patient to a stroke unit—are complex. This week a paper in the BMJ presents a model (p 694) with which researchers can clarify the stage of evaluation that a particular complex intervention is at and discusses which methodologies are most appropriate at each stage. The examples given mainly come from interventions in which a social or educational component is important.
On the web, readers have a choice of two papers: the BMJ version or the original version published by the Medical Research Council in PDF format (www.mrc.ac.uk/complex_packages.html), which runs to 19 pages and uses most of the space to explain its message on complex evaluations in more detail and with more examples. Aspiring grant holders will certainly want it.
At www.hta.nhsweb.nhs.uk/index.htm the publications page of the NHS Health Technology Assessment programme has a long list of methodological reports as well as several examples of evaluations of complex interventions. The problem with both sites is that only PDF files are available for downloading. Although PDF is an open standard and free readers for the files are readily available, using PDF format for everything can be burdensome and wastes bandwidth, as it requires you to download the whole document before you can read any of it. That said, making a PDF from a single word processed document and putting it on the web is a trivial task, whereas modifying information so that it can easily be read on the web is difficult and time consuming.