MPH Master pieces

Molecular epidemiology

Identifying precisely the organism which is causing an outbreak is essential - and nowadays that means molecular typing of the strain. Here is a good account of this process. Like all the best outbreak investigations, it reads like a detective story.

Overweight: India vs China

Here is a straightforward piece of descriptive epidemiology with a most intriguing result. In short, in Indian and Bangladeshi populations, but not Chinese, Korean and Japanese there was no evidence that a high body mass index is associated with increased mortality. But we must think very carefully before deciding that overweight is ok if you are from the Indian subcontinent! One possibility is that high BMI in India means you are rich, and that this outweighs (as it were) the disadvantage of obesity.

One moral of this tale is not to take anything for granted - you might have thought this study not worth doing because it is so well established from so many studies in so many places that high BMI means higher mortality.

Safer hospitals

You don't often see evaluations in the UK of organisational change but here is a good example. It is probably relevant that the initiative was funded by a private organisation: they wanted a decent evaluation. The evaluation consisted of a mix of quantitative and qualitative: statistical measures and interviews with staff.

The aim was to improve patient safety in four study hospitals (later disseminated more widely). Great use was made of statistical process control charts.

The number work showed no detectable effect of the intervention, but this as the authors emphasise was against a high baseline and a rising secular (i.e. NHS wide) trend in the control hospitals. The ethnography was interesting - enthusiasm at the top but mystification at the bottom. The authors point out that the hospitals didn't really have an organistional strategy for delivering the change.

Organisational change is really difficult, or as one of my favourite book titles has it 'Hope is not a strategy'.

Effective control of blood pressure

Community trials of cardiovascular prevention have generally found it difficult to prove positive effects: either the effort was not big enough, or there was equally large change in the control communities from nation-wide campaigns. So here is that rare thing - a community trial of cardiovascular prevention which worked.

It was based on a substantial effort to mobilise communities in small towns in Ontario. It's not clear whether you could sustain this type of campaign long term, of whether the benefits are just a one-off.

The outcome of interest was a reduction in hospital admissions. One small niggle here - always remember, when looking at hospitalisation data, that admission is a decision not an event: the campaign may just possibly get local doctors to think differently about the need for admission.

I liked the comments at the end of the discussion comparing the effects of this intervention (a 10% reduction in admission from heart attack, stroke and heart failure) to other possible interventions (e.g. a 2 gram / day reduction in salt intake, 40% reduction in tobacco use etc). The authors say an economic evaluation of their project is in preparation. I look forward to reading it.

All in all, an interesting project.