(edited for clarity, corrections 7th Aug 2015)
Medicine is hard. Nobody has ever been particularly certain of the best thing to do, we've all just been making it up as we go along. In the age before mass communication medics relied on anecdotes, personal experiences and texts of dubious provenance to decide what to do. Until relatively recently this amounted to little more than trusting luck and most of the time we would have been better off if they had left us alone. Nowadays we have randomised controlled trials, meta-analyses, epidemiology and sophisticated statistical methods. But it's still not always obvious what we should do.
|
A picture of some hard medicine |
Every so often there is a tragedy that leads people to decide on some obvious course of action. Recently there's been a lot of attention on social media about a case of a newborn dying as a result of infection from group B streptococcus (strep). This is an infection caused by a very common type of bacteria that can lead to severe, life-threatening complications like septicaemia. It is transmitted from mothers to children during childbirth. The obvious course of action is to eliminate the bacteria in question from the mothers or to treat newborns with antibiotics to ensure that any possible infection doesn't get a foothold.
Unfortunately, reality is not so simple. While there is no doubt that screening and antibiotics would have (probably) saved the lives of the newborns who die from this infection we have no way of knowing in advance which ones are going to die. This is the root of the problem. We see that a child died and we can say, with some confidence, that if only we'd had a mass screening programme combined with preventative antibiotics then that child would have been saved, Many campaigns are founded on this sort of over-simplification. What these campaigns always fail to take into account is the number of people who would be harmed or killed by the intervention they recommend. The proposed solution is always 'do more medicine' but we're not yet at the point where medicine does no harm.
As a very brief run-through of this example - we'd need a screening programme that tests mothers for group B strep before birth. A screening programme is pointless unless is informs real clinical decisions - finding out somebody is a carrying group B strep isn't useful information unless there's something we plan to do about it. Assuming we have a perfect test (we have no perfect tests for anything) we'd identify a certain proportion of women who are at risk of transmitting group B strep to their newborns. So all we'd have to do is load them up with antibiotics before birth and do the same for the newborns. Bingo, no strep B deaths.
But there's complications. Tests for group B strep will identify somewhere between 25% and 40% of women as carriers. Not all of those women will even still be carrying strep B at the time of birth. Their status as carriers is unlikely to even be relevant in cases of caesarean section. Even during conventional births not all women will transmit the bacteria to the newborns. Ones that do probably won't get infected. Those that do get infected probably won't come to any sort of lasting harm. Some will. Some will die or will suffer permanent ill health as a result of doing nothing to prevent the infection.
The screening would identify at least a quarter of women as carriers of group B strep. There were around 700,000 births during 2014 in England and Wales. If we're conservative, we can say at least 150,000 women will be identified as carriers.
So that would be 150,000 women recommended to receive intravenous antibiotics in a hospital setting during the births of their children and about the same number of babies receiving preventative antibiotics. The overwhelming majority of these would never have suffered any health problems as a result of carrying group B strep. As previously mentioned, however, some lives would be saved as a result.
But that's not the whole story.
In addition to the lives saved, some will die or will suffer permanent ill health as a result of trying to prevent the infection. There's various risks of antibiotic use - discovery of latent allergies in the mother, unexpected toxic effects and all sorts. There's strong evidence that use of antibiotics in very young children increases the risk of antibiotic allergies as well as conditions such as asthma. There are risks associated with increased medicalisation of birth - does a positive strep B result rule out certain types of birth choices, such as midwife-led units, leading to more c-sections, worse overall outcomes? Mass use of antibiotics is widely recognised as a bad move due to the development of resistant bacteria yet screening for such a common bacterium demands it. We'd reach a point where we'd be forced to use more and more toxic antibiotics to treat new, resistant forms of strep. This would have the effect of increasing the death rates in the rare cases where a newborn gets infected.
That's only the start, there's all sorts of unintended consequences of any sort of mass health intervention. This is why we need really strong evidence of the benefits of a screening programme before we can introduce it. Medicine is a balance of risk and benefit - medicine does harm as well as good and any intervention needs to demonstrably do more good than harm. The mass treatment of healthy people is a very difficult as the majority of people will only receive the risks with no benefits. Until we're (reasonably) sure we're (probably) better off doing nothing. We're already pretty certain of the easy interventions, after all.
All of the outstanding problems in medicine are extremely complicated and we're only very slowly figuring out the best things to do. We have to balance risks against benefits and in most cases that's difficult to work out. There are no easy answers any more. Everything left in medicine is hard.