MPH Master pieces

Sub-group analysis and race-specfic drugs

Interesting expose of the BiDil saga in this month’s Scientific American (Jonathan Kahn ‘Race in a Bottle’ Sci Am Aug 2007; 26 - 31)  – I was vaguely aware of this but didn’t know the details. BiDil is a drug that was approved in 2005 by the FDA for heart failure – but in African-Americans only.

The logic case was that African-Americans have different gene frequencies to white Americans and so race-specific drugs represent sensible genetic targeting. In principle this might just about be true for small tightly knit kinships but 'African-American' is a huge category.  Black Americans do seem to suffer a higher rate of stroke than Whites at any given blood pressure level, suggesting perhaps some relevant cardiovascular genetics, but the facts of this particular case turn out to be very different.

Firstly, and I hadn’t realised this, BiDil is just a combination pill of two very common drugs for heart disease: hydrallazine and nitrate. Nothing special there which is likely to be gene-specific. 

Secondly the effect of BiDil on African-Americans was noted only in post-hoc sub-group analysis of a much larger trial. This is well known to be an extremely dangerous thing to do. A trial had shown no overall survival benefit from BiDil but if you split the subjects by ethnic group there was a small survival advantage in the 49 Black patients. This then led to a trial which only recruited self identified African-Americans. This second trial demonstrated a large (43%) survival advantage in the treatment group versus the placebo group. Because the trial had only been conducted in Black patients, the FDA only licensed BiDil for Black patients. 

As Kahn points out, this is barmy – no drug has ever been licensed for whites only because the trial only included white patients. This is policy making ‘on the hoof’ – I feel some sympathy for the FDA asked to rule on a complex issue they didn’t have time to think through properly. We’ve all been there, albeit on a smaller scale. 

Take home message – beware retrospective sub-group analysis! and don't make policy on the hoof.

Tap water scalds

I enjoyed this paper from Canada in 'Injury Prevention'. Although this is an economic analysis, it's the underlying idea that attracts me - pass a law requiring new water heaters to be set at 49 degrees Centigrade (current regs in Canada allow a factory setting of 60degrees). They've already passed laws like this in 28 states of the USA. The authors cite a paper from 10 years ago that showed a 56% drop in paediatric admissions for tap water scalds in the 15 years after legislation in Washington DC.

The economic analysis is needed because the proposal goes a bit further than (no cost) legislation and requires utilities to send a notice every year to householders warning them about the dangers of high water temperatures.

In the UK we are all a bit paranoid about Legionnaire's disease, and after the outbreak in Stafford hospital which led to a public inquiry practically all public buildings have scaldingly hot water. But homes are of course different. The authors state (with a slightly iffy reference) that there has been no increase in the 28 states of the USA which passed tap water laws.  Not sure of the position in the UK.

Scalds account for 70% of burn hospitalisations, and about 5 - 10% are tap water scalds (I guess the rest are kettles, frying pans etc). So a smallish problem but even so the answer seems simple and doable.

Outbreak control the CDSC way

Some classic outbreak reports in this issue of the HPR.

The one I particularly admire is the E.coli 0157 report - note the use of active case finding (needs a case definition), descriptive epidemiology (map and outbreak curve) and analytical (case control) study to put things beyond doubt even though the microbiology was pretty convincing right from the start.

No more coriander and chicken for me. At least not from any national supermarket chains.

Fat friends

Another product from the Framingham study! It's difficult not to be wowed by the video attached to web version of this piece - but we need to check the underlying science is sound.  The basic message is that obesity is not just about you, it's about your friends and neighbours. Another reminder of why do public health - man is a social being and influenced by social processes.

Social processes are difficult to measure, which means the statistics can be incomprehensible: and incomprehensible means potentially seriously misleading.

I remember a great debate about the effects of unemployment on health which hinged on showing that cycles of unemployment matched cycles of deliberate self harm. The problem was that the statistics of matching cycles (time series analysis) is very tricky and in the end it was clear that the apparent correlation was spurious.

In this case, we have to take comfort in the fact that the analysis is published in the NEJM. It's the world's leading biomedical journal so I guess they can call on some seriously competent referees.

Adding folic acid to bread

Here's another classic - impressive data from Canada which shows that fortifiying cereal products with folate (mandated by law in 1998 in Canada) reduces neural tube defect.

There's still a residual concern about the effect of this on people who are deficient in vitamin B12, particularly perhaps residents of nursing homes. But even if this is an adverse effect it may be one we want to accept given the benefits of supplementation.

This is a rare example of a measure which reduces inequalities - in this case between provinces in Canada, because the effect was greater in areas with a high baseline rate of NTD. Adding folate to cereals is law in the USA and Canada but not yet in the UK.

The only other examples I know of how to reduce inequalities are :       

 fluoridating the mains water supply (ecological evidence)
        area traffic calming (can't cite anything here)
        eradicating disease completely (stands to reason)

Note that these are all things that the public have done to them. Almost anything requiring active participation will favour the rich and powerful - because power in society is the ability to act - and hence increase inequality.