Skip to 0 minutes and 7 seconds This week we’ve been looking at risk, how we identify it, and the limitations of research designed to do this. These research designs are subject to confounding other explanations and to bias. Generally, they identify possible relationships, associations, and correlations, rather than define cause and effect. When talking to researchers, correlation is not causation is a bit of a mantra. But sometimes the evidence is just so overwhelming that we have to accept there’s a clear link between doing something and the likelihood of a particular result. Our final case study looks probably the best known example of this. Smoking and lung cancer. The research we’re considering was a pivotal study conducted over 50 years. Why do we take risks?
Skip to 1 minute and 8 seconds We all know that certain behaviours produce a greater risk. In some cases it’s denial that this could happen to us. When we’re young we can feel immortal. In others it’s because we really enjoy it. Or it could just be that we’re sceptics who don’t believe or trust what we’ve been told. In the case of smoking, when we’re young, we might not see the harms. We find it enjoyable, can become addicted to it. Or we don’t believe the evidence that say it could happen to us. They’re always going to be those who do risky things and get away with it.
Skip to 1 minute and 46 seconds For example, both my grandfathers who lived to a ripe old age one to 86, the other to 91, smoked all their lives. They didn’t die of lung cancer. That will happen because risk is calculated at population level. So not everyone who smokes will get lung cancer. However, according to the research we’re looking at in this case study, the risk of death from lung cancer is about 15 times greater in current smokers compared to those who never smoked. However, lung cancer is not exclusively the result of smoking. About 14% of lung cancers are not linked to smoking. So how was the link between smoking and lung cancer established? A clinical trial isn’t possible in circumstances like this.
Skip to 2 minutes and 40 seconds It’s simply not ethical to ask people to do something that could cause them harm just so we can see how it affects them. Remember the Declaration of Helsinki? This says that patient safety is a primary concern in health research. Now research looking at the increased risk of lung cancer cases was already identifying a possible association with smoking as far back as 1950. Cigarette advertising is now banned in much of the world. But in the 1940s and ’50s, smoking was commonplace. And advertising messages were all about acceptability, family, purity, and healthy living. Advertisements using images like these of babies and doctors were quite common. So where did things began to change? Well, in 1950, a physician, Dr.
Skip to 3 minutes and 37 seconds Richard Doll, and an emminent statistician, Dr. Austin Bradford Hill, published a retrospective study that looked back at the habits and environments of patients that had been diagnosed with lung cancer. Their paper highlighted the need for prospective research that would follow a group of people forward and see which of them developed lung cancer. This is the 1950 research published in the British Medical Journal. In it, Doll and Bradford Hill say that when looking at the increase in lung cancer, two main causes have from time to time been put forward. One, a general atmospheric pollution, and two, the smoking of tobacco.
Skip to 4 minutes and 24 seconds So having found funding for their research, in 1951, Doll and Bradford Hill commenced their study which ran for over 50 years. They wrote to every doctor in the UK, and signed up 34,000 male and 6,000 female doctors. As doctors, they all had to be signed up to the medical register, so it was easy to keep track of them. In the next section, you’ll see a discussion of how the research developed and listen to Richard Doll talking about his research.
Smoking and lung cancer: lifestyle choices and risk
The case study this week brings together many of the themes that we have explored throughout the course.
The topic of smoking and lung cancer encompasses concepts of risk, the nature of causation and correlation, the ethics of clinical trials and the relationship between vested interests and the dissemination of health research.
In this first video, we examine smoking and risk and the context to Richard Doll’s and Bradford Hill’s landmark 1950 study on smoking and lung cancer.