In my job, one of the trickiest things to consider in outbreak management is the secondary effects of any restrictions. They are often difficult to predict, let alone quantify, yet they can drastically alter the balance of risks.

Imagine, for example, a batch of ready meals which are suspected of being contaminated with a bacterial pathogen yet are destined to be heated and served to hospital patients. I think everyone’s first instinct would be to withdraw the meals to reduce the risk of giving patients food poisoning.

But this decision would not be as straightforward as it first appears. The risk of illness from the meals might well be mitigated—but not necessarily eliminated—by the need to heat them, which will usually kill the bacteria. If the patients are not to have the hot meals, then they will have to be served something else. If that ‘something else’ is, for example, pre-packed sandwiches, then the risk of illness for vulnerable patients from pathogens such as listeria, which are reasonably common in pre-packed sandwiches, might be higher than the risk associated with the hot meals.

In this scenario, the fact that patients must be fed and that there will, therefore, be a secondary effect is obvious and predictable. Sometimes, secondary effects are entirely unpredictable, and you can do no more than take an informed, professional guess. And then, crucially, keep an eye on the impact of interventions so they can be tweaked if necessary.

The same is true in clinical medicine: adding one extra tablet to a patient’s regimen might reduce their risk of developing a particular illness. But it might also increase their side effects, interact with another medication, or be the extra tablet that tips them over the edge into non-compliance with the whole regimen. Well-intentioned decisions can have unexpected secondary consequences.

Kelsey Piper of Vox gave a great example of the importance of considering secondary effects last week. She describes the fact that the US Federal Aviation Authority has done extensive research which has found that it is considerably safer for very young children to fly in their own secured seat, rather than travelling in someone’s lap. They give clear public advice on this, yet—despite all the evidence—choose not to mandate it.

Their rationale is that requiring the purchase of a seat for very young children would make flying unaffordable for many families, and a proportion of those families would choose to drive to their destination instead. Driving is considerably more dangerous than flying, to the point where the FAA calculates that for every life saved by a ‘separate seat’ mandate, sixty lives would be lost on the road.

Sometimes, we can all get drawn into looking only at our piece of any given puzzle, but it’s essential to keep a broad view: you never know when a secondary effect might undermine your action.

The image at the top of this post was generated by DALL·E 3.