Feeds:
Posts
Comments

Posts Tagged ‘Study Design’

Currently I am practicing good Infection Control and am in quarantine – which has thrown a spanner in the works with regard to my active research at UHCW, but on the positive has given me time to work on planning my next phase of work.

Following the emerging themes produced from Study 1 and 2 (including interviews with Healthcare Professionals) the design for Study 3 has had to change.  It is no longer feasible to carry out the Inherent/Elective categorisation task as originally planned, as there is too much ‘noise’ around the concepts of Hand Hygiene triggers such as “Patient Contact” and “Contaminated Items”.  Therefore, instead of using existing audit data for a retrospective analysis, the new design involves an in-situ observational study collecting fresh Hand Hygiene data relating to specific activities deemed “Inherent” or “Elective” by current Healthcare Professionals.

From a conceptual point of view, this seems simple.  Indeed, when discussing the idea with others the idea – to see whether people perform Hand Hygiene more at certain points than others, and then whether this behaviour is stable – seems both clear and interesting; however when delving slightly deeper, the actual quantitative analysis required to investigate this theory appears much more complex.  Issues such as the use of binary (dichotomous) data, and a within-subjects design, makes the study actually quite abnormal.   However, after much design/re-design and discussions with some great sources of help, I think we may be getting close to a solution.  Today I thus have to just confirm that  the McNemar and Cochran Q are potentially solutions, and then I can firm up the design…

Whilst statistics are difficult in such cases, as has been said to me recently, the good part is once you have found the correct design, the actual carrying out is relatively simple. I cannot say how much I am looking forward to that part…!  The redesign of Study 3 has really shown how hard it can be to decide what is possible in research – having to compromise what you find interesting with what is possible to quantify or work into a design.  Patience, logic and the understanding that future work is always an option.  Lessons learnt…!

Read Full Post »

%d bloggers like this: