
Have you heard the story about the one person who beat cancer with positive thinking? Or that Winston Churchill smoked cigars and drank alcohol from morning to evening and even while in bed, lived to be 90 so we can all smoke and drink without harm? Or that out of the top 200 Australian companies on the ASX, 19 are led by female Chief Executive Officers, and because these women made it the “merit based” system is working and is fair? Can these individual stories help us to understand what is really going on?
The problem here is when we focus on an individual story, rather than understanding what is happening across a population. Epidemiology is the accompanying science to public health, because it helps us to understand what is happening across a population. It works on the basis of summary measures like average (means) and medians, it generally does not concern itself with the experience of one.
If we take an individual view, e.g. my uncle smoked every day and lived to 84 years old, we miss the bigger picture about the effects of smoking on life expectancy as a whole.
For example, while some people do not have their life cut short by smoking, at least 50% of people who smoke will. When we study the link between tobacco smoking and early death, we look at what happens on average in a population. If we took a group of 100 smokers, we would see how many died at different ages and summarise these as averages across the whole group. While some people might live a long life, around 50 of them will not.
A study published in BMC Medicine looked at the risk of death at different ages for smokers and non-smokers (Banks et al). They showed there is a risk of death at any age, and that risk increases steadily from age 45 to 75 years old. Men who were non-smokers at age 65 had a 6.7% chance of death, and smokers had a 17.8% chance of death to that point. Smokers have a more than double risk of death at age 65 years. This gap in chance of death increases with age.
Most importantly, if we take an individual rather than population lens, we lose critical information. We don’t just lose what happens on average in a population, we also lose how the health issue is distributed in the population. That is, that the health issue is more common in people who have a lower income or live in a lower income area, or that the health issue occurs more often in one community when compared to another. We lose the ability to see that inequity plays a pervasive and important role in most causes of ill health.
When epidemiology allows us to think at a population level, the sorts of interventions we have expand to those available at the population level. No longer are we limited to advising individual people to stop smoking and offering quit smoking aids. We can now act across the population with legislation, policy, communications and education (among other tools), to assist all people to quit, and even better, reduce the uptake of smoking in the first place. Critically, these population approaches are aimed at everyone, not just the one person sitting in front of a clinician.
How good is that?
Comments