Multilevel analysis of a hospital-wide hand hygiene program

Topic: Multilevel analysis of a hospital-wide hand hygiene program in Singapore
I am planning to write a research article to be submitted to the Journal of Hospital Infection
Full-length, original research articles
Should contain up to a maximum of 4000 words, which includes the structured summary, text, acknowledgements and references. Each figure and/or tables counts as 200 words towards the total. Separate Figures or Tables labelled 1A, 1B, 1C etc would count as three separate tables, not one.
For this assignment, I am only targeting around 3000 words + 4-5 figures/tables.
Please do the reference format in EndNote so that it would be easier for me to track and simplify changes.
-Why is hand hygiene (HH) important? Reduce hospital acquired infections that have detrimental effects
-HH compliance is low worldwide.
-Our hospital 50-60% compliance
-November 2013 our hospital started a campaign where each inpatient ward is allowed to select one or more interventions to be applied with the goal to improve HH.
-Aim: to determine which intervention combination works the best
-How? We compare the audit results before and after the program.
-We need to treat this as a cluster trial as groups of individuals rather than the individuals are randomized to different interventions (I have provided some papers on this). Need to elaborate.
-Please refer to PROPOSAL.pdf and RESULTS.pdf
-Audit – refer to PROPOSAL
-Audit – refer to PROPOSAL
-Wards and HH moments included in the study (see DATA.xls)
-Types of intervention adopted by the wards (brief description)
-Study design
-Statistical analysis: paired t-test to compare overall and individual wards improvement before and after intervention. Multilevel modelling.
-Overall improvement but some wards did better than others.
-Professions 1=nurses, 2=doctors, 3=allied health, 4=nursing students, 5=medical students. HH depends on profession. Why? Any other papers supporting this?
-If barriers to HH is removed so HH is easier to be done, HH compliance would improve.
-Interestingly, if combined together, they do display a strong effect.
What does this mean to healthcare workers: stop useless programs and use audio/visual/audit method – but phrase it cautiously as human behaviour is a dynamic interplay between individuals and their environment – what works in one ward may not work in another. No one size fits all.
-Strengths of the paper: made the best out of a real life scenario, multiple time points to assess, multiple wards, many assessed at each ward. Established cause and effect. Advantages of cluster design. One ward to the next may be different