Helping people avoid getting sick or helping them navigate it when they are is one of the main goals of patient risk stratification. But a key question was raised by British epidemiologist, Jeffrey Rose back in the 80s. Is it possible that there are sick populations? And if there are, how do they differ from sick individuals, and what can be done to help? In this module, we discuss the conundrum of whether or not illness is due to sick individuals or sick populations. And we're going to demonstrate the inherent paradox of trying to separate people from their communities or populations. I think most of us intuitively know that disease and population arises as an interplay between people and where they live and what they do, their place. Over the last 20 years, we've seen what I'll call an epidemic rise in obesity, suicide, social anxiety, vaping-related pulmonary conditions, autism, dementia, and many more chronic diseases and acute diseases. So the question is are people inherently getting sicker, or is it where they're living and what they're doing, their community and their culture, that's actually making them sicker? When Jeffrey Rose put forth this concept of the dichotomy or really the difference between sick individuals and sick populations, it came from an understanding that for many diseases a person who is sick is part of a larger integrated continuum. So here I've got a graph of the systolic blood pressure, and what you can see is that there's a single line at 140 millimeters mercury. Where if you're above it, your hypertensive, and if you're below it, your normotensive. So somebody could have a systolic blood pressure of 142 and be hypertensive, but their neighbor could have one of 138 and be considered normotensive. So here's an example that Jeffrey Rose presented in his paper, two histograms representing blood pressure of two populations, Kenyan Nomads and London Civil Servants. What do you see? Well, the general shape of the distributions are the same, but the means are quite different. It's almost as if one got shifted to the right and one to the left. What do you think is going on? Why do you think they're so different? Again, if we use 140 millimeters mercury as the cut point, you can ask the simple question, well, what percentage of Kenyans and London Civil Servants have hypertension? And just looking underneath kind of the area underneath of the curve, you can see that London Civil Servants probably have something like 30%, or maybe a little bit more, in the hypertensive range. Whereas, the Kenyan Nomads only have about 5% maybe that are in this hypertension range. Overall, if you look at how the shape of the distributions are so similar, evidence is pointing to the fact that the difference in disease is probably generated by a person's population setting. How do we know this? Well, if we think about the question, what is the cause of high blood pressure in London Civil Servants? And is it different than Kenya Nomads? We have to grapple with the fact that, in general, we're not pinpointing the cause of high blood pressure in an individual when we identify it. We're simply treating it, and we're keeping track of it to make sure it doesn't escalate. But because the shape of the distribution is so similar between both the Kenyan and the London populations, well, that is indicating to us that there's not something just with a few individuals. And we think, well, the mean probably is the biggest shift here that we're seeing, that's probably due to powerful population forces. We know farewell that actual processes in somebody's body that gives rise to hypertension, the physiology, the genetics of it, is likely to be the same across all human beings. So in this case, what we see when we look at these distributions is that the Kenyan Nomads are closer to a healthy underlying population pattern than the Londoner's distribution, which seems to be shifted upwards. And probably represents some unhealthy lifestyle background that they're living in, and it's eventually getting inside and actually causing the high blood pressure. So is it a sick population or a sick individual? Another way to think about this is to ask are the causes of blood pressure in the population different than the causes of high blood pressure in individuals? And basically, the causes of the rightward shift probably indicate social norms about diet and exercise and stress. Those are our major risk factors for hypertension, and they're probably acting on the entire population, sort of uniformly, and that's what gives rise to the the shift in the mean. So the causes of sickness in the population of London Civil Servants are conceptually different than the causes of sickness in the individuals from the Kenya Nomad population. To be honest, this now presents us with a conundrum. I think we all know that the population distribution of risk factors, and that shape of the relationship between a risk factor in a disease are things that population health and public health use in order to develop prevention programs. So the question now turns into, should we be focusing on high-risk individuals and preventing them from being high-risk, or should we be focusing on reducing the entire population's risk? These two approaches are radically different because one is focused on the individual and the other is focused on the population of causes. To better illustrate this conundrum and controversy, let's look at using these two approaches to solve alcohol abuse in a population. So here we have three distributions sort of demonstrating a high-risk approach. And in the next slide, we're going to talk about a population-based approach. From a sick individual's perspective, an individual may be an alcoholic or abuse alcohol because of their response to their own personal anxiety or depression, unemployment, or simply a deeper fondness for the alcoholic beverages that lead to addiction. So the actual distribution of alcohol, sort of units per week that's shown on the left, would be specific to individuals in a particular population. And we might focus on the high-risk individuals at the tail of the distribution, shown in red here. They have an increased health risk in a number of different domains and those map onto their excessive use of alcohol. So a high-risk approach would actually try to work on those high-risk individuals and push them back into the distribution. Realistically, that's not going to happen. But those high-risk individuals, some of them might move into the more middle part of the distribution, and then other ones might come out onto the tail. The desired truncation of the distribution is not likely to happen unless they're very severe laws or regulations about certain amount of drinking, which is currently not used any place in the world that I know of. In contrast, a population-based approach is thinking about what are the elements inside of the population, inside of local communities, for instance, that might have norms about high or low consumption. And there may be different reasons for that. There may be norms around drinking and religion that both pull up at different times drinking frequency or actually have sort of norms against drinking. There are also ways in which the sort of community norms around hospitality or even the available income or access to alcohol could be influencing an overall population mean. And so rather than thinking about the addressing of high-risk individuals and truncating the distribution, the population-based approach is trying to take the whole population and shift it down. So that there's a new mean that is much lower. And one of the key things in this kind of differential approach is that you're much more likely to actually have a beneficial effect on more people this way, not just the high-risk drinkers, but everybody in the population. A handy metaphor that was put out, I believe by the World Health Organization or one of its members, was the idea that this is illustrated by population approaches influencing kind of downstream consequences by turning off the tap. So if you have a leaky system that is constantly giving rise to disease and somebody downstream, namely the healthcare system, has to sort of help address all of these issues that are emerging from this leaky system. Well, why not focus on turning off that tap? And reality, both of these jobs are very important. But it sort of juxtaposes what we might gain by focusing on population-based approaches as one of the ways to stop downstream disease. In reality, anyone who's interested in improving population health has to be invested in both of these, both understanding upstream causes and addressing them, but then also being a good steward of the downstream consequences. Because when somebody is ill, they need help. They need it from the healthcare providers or from their community immediately. But if you want to make a long-standing contribution, potentially to millions of people, than focusing on these upstream culprits is definitely a step in the right direction.